The Madras High Court’s Judgment in Sunitha vs Additional Chief Secretary – II: Waking Up to the Abuse of Preventive Detention Law [Guest Post]

[This is a guest post by Varun Ahuja.]

On 14th November 2022, a Division Bench of the High Court of Madras delivered a scathing judgement indicting the State for misusing the preventive detention law. Titled Sunitha v. State [HCP(MD) No. 1710 of 2022], the Court quashed two preventive detention orders passed under the Tamil Nadu Goondas Act, 1982. Both the Orders were quashed on the ground that the activities alleged did not fall under the purview of disturbance to ‘public order’ (Para 41 & 56). Notably, the Court recognized that repeated illegal detention orders would amount to a constitutional tort (Para 26) and awarded a compensation of Rs. 25,000 in one of the cases to be paid by the Additional Chief Secretary, who was the detaining authority in the case. This was not the first time that the Madras High Court awarded compensation for illegal preventive detention. In September 2022, the Court awarded 5 lakhs compensation to the detenu who was kept in illegal custody for 128 days after the Advisory Board opined that there was no sufficient cause for their detention (Manokaran v. State, HCP. No.297 of 2022)

The judgement has been discussed on this blog before. It has been correctly lauded for recognizing a pattern of abuse by the executive and for providing a remedy (compensation) in cases of illegal preventive detention orders. This post will primarily deal with the use of data by the Court to reach the conclusion that the State is misusing the law of preventive detention. Additionally, it will compare another judgment from the Manipur High Court, which conducted a similar exercise, but stopped short of providing substantial remedy to the detenu.

Dearth of Data

It is settled law that the powers of preventive detention are extraordinary in nature and cannot be used in place of ordinary criminal procedure. But how does one prove that the law is being misused by the State? The power to detain a person preventively has been in operation ever since India became an independent Country (and even before). However, there has always been a vacuum with respect to data on the number of persons being detained. Earlier, fragments of the data could have been accessed through Questions in Parliament (See here and here) or through Parliamentary debates (See Columns 9981-82). It was only in 2017 that the National Crime Records Bureau (NCRB) started publishing data on the number of persons detained every year.

In 2017, NCRB started publishing statistics on the persons detained under the various preventive detention laws in its Crime in India Report. A plain reading of the total number of persons detained show the rise in the use of preventive powers of the State. On an average, the State detains close to ninety-five thousand people under various preventive detention laws every year.

YearNo. of persons Detained

Recently the Supreme Court, in Mallada K Sriram v. State of Telangana, 2022 SCC OnLine SC 424, highlighted the abuse by the State of Telangana of its Goondas Act, 1986 and termed its repeated use ‘a callous exercise of the exceptional power of preventive detention.’ In that case, too, the Court made use of data and noted that:

It is also relevant to note, that in the last five years, this Court has quashed over five detention orders under the Telangana Act of 1986 for inter alia incorrectly applying the standard for maintenance of public order and relying on stale materials while passing the orders of detention. At least ten detention orders under the Telangana Act of 1986 have been set aside by the High Court of Telangana in the last one year itself. [Emphasis Supplied]

The High Court refera to this judgement (Para 8) and goes much further in conducting a granular exercise to point out the abuse of power by the State.

High Court’s Analysis

The Court relies upon the Prison Statistics India Reports published by the National Crime Records Bureau (Para 5) to show that the State of Tamil Nadu has had the highest numbers of detenus since 2011. However, the Prison Statistics only show how many detenus were in jail as on 31st December of the year the Report is being published and not the number of detention orders being passed each year. To buttress their point further and show the frequent use of the law, the Court relies upon the ‘Crime Review, 2020’, a Report published by the Tamil Nadu State Crime Records Bureau which indicated that 84.3% of all detentions in the State (2457 out of 2913 orders) were passed under the State Goondas Act (Para 10).

A source of dismay for the Court seems to be the statistics with respect to habeas corpus petitions disposed by the Madurai Bench from January 2022 till 31 October 2022 (Para 17). The results from this dataset is strikingly similar to a larger study conducted of the habeas corpus petitions decided by the Madras High Court (both the Benches) from the year 2000 till January 2022 (total of 7,448 cases over 22 years) under all preventive detention laws operational in the State. The study revealed that an overwhelming 95% of the cases (7,096 out of 7,448) emanated from the Goondas Act; in the Court’s dataset this proportion was 72% (961 of 1,332 cases filed). The Study states that 87.9% of the petitions (6,547 out of 7,448) were ‘successful’ i.e. the detention order was rendered illegal; the Court has a similar finding of 84% of the petitions decided (598 out of 708). In the Court’s data for the year 2022 there was not a single detention order which was upheld on merits; the Study paints a similar picture showing only 1.8% of the cases (136 out of 7,448) which are allowed on merits.

An important objective of the Study was to find how much time the High Court took in deciding a habeas corpus petition – arguing that the blame for violating personal liberty lies on the doors of the Executive as well as the Judiciary. Under the Goondas Act, the maximum period of detention is prescribed as 12 months (Section 13). The study showed that calculating from the date of filing of the petition, the Court, on an average, took 141 days (a little less than 5 months) to decide a case and in its dataset, at least 30 cases were dismissed as infructuous due to the expiry of the detention period and 722 cases were dismissed as the detention order was revoked by the Government (collectively amounting to 10% of the cases). On the other hand, the Court’s dataset shows that at least 15.5% of all the cases (110 out of 708) were disposed as ‘closed’ i.e. infructuous owing to time and revoked by the Advisory Board (Para 18).

The departure in Sunitha’s casefrom the other judgements challenging detention orders has been to not overlook the significance of this delay but to acknowledge it and provide for practical remedy. The Court begins the judgement in Para 3 with the lament that orders in such cases take ‘at least 4-6 months’ to be decided ‘owing to the existing backlog of cases’ (reiterated in Para 15). This acknowledgement is important because a speedy disposal of such petitions will help keep the State in check and prevent wilful abuse of its extraordinary powers.

In a Similar Vein

It is not the first time a Court has used data in a challenge preventive detention order to prove indifference of the State. While the Supreme Court in Mallada provided numbers of detention order quashed, another judgement from the High Court of Manipur had conducted a detailed analysis on why detention orders were being routinely quashed.

In Ningthoujam Yamba Singh v. State of Manipur, 2018 Cri LJ 466, the Manipur High Court was considering a detention order passed by the District Magistrate under the National Security Act, 1980. The detenu in this case was a Rifleman of Army Supply Corps (ASC) and was arrested in a case of kidnapping and extortion when the detention order was issued. The Order was quashed because even though it recorded that there is a likelihood of release of detenu on bail, no material was placed before the detaining authority that would indicate the same (the Court relied upon Union of India v. Paul Manickam, (2003) 8 SCC 342)

The Court went on to tabulate details of 39 cases (Para 16) where it had quashed the detention order on the identical ground – non application of mind as regards the satisfaction of the detaining authority that the detenu is likely to be released on bail in near future. It also pointed out that in spite of at least 4 previous judgements (W.P. (Cril). No. 37 of 2013, W.P. (Cril) No. 37 of 2016, W.P. (Cril) No. 35 of 2016 & W.P. (Cril) No. 36 of 2016), where the Court had passed directions to the Chief Secretary of the State to circulate the judgement laying down the correct law, the officials were still passing illegal orders (Para 21-23). The Court even registered a suo-moto case since the detenu in that case was affiliated with the Army and continued to be in its employment even after detention orders were issued against him. The case was titled In Re Detention Under National Security Act, 1980, PIL 28/2017 and the decision date as shown on ecourts is 20-06-2018; however, no judgment is uploaded by the Court, nor is it available on any other website.

The Manipur High Court, however, left the Sate with a warning:

28. It may be also noted that this Court may be constrained to pass appropriate orders in future against those persons responsible for issuing defective detention orders by way of imposing exemplary costs who continue to issue such defective orders as these defective orders may amount to illegally detaining persons and violating their Fundamental Right as guaranteed under Article 21 of the Constitution. [Emphasis Supplied]

However, this warning has been reiterated by the same the Court again and again quoting the same observations from Ningthoujam Yamba Singh but no action has been taken. Therefore, it is in this context that the Madras High Court’s judgement in Sunitha does not pay lip-service to the cherished right of personal liberty but ensures that the State will have to suffer consequences for its wilful disregard of the same. Only time will tell if compensation is deterrent enough to prevent abuse of the powers but it is a welcome change from the current status quo of absolute impunity.

Preventive Detention, State Abuse, and an Unconstitutional State of Affairs: The Madras High Court’s Judgment in Sunitha vs Additional Chief Secretary

On 14th November 2022, a division bench of the Madras High Court handed down an important judgment on the subject of preventive detention (Sunitha vs Additional Chief Secretary). The High Court examined – and set aside – two preventive detention orders passed by the State, under the Tamil Nadu Goondas Act. There are three significant features of the judgment, that merit careful study.

A Pattern of Abuse

The Court began by recounting its experience of “scores of habeas corpus petitions” against preventive detention orders being filed every day, which “are inevitably allowed”, albeit after 4-6 months, because of the backlog of cases (paragraph 2). The Court elected to go deeper into the issue, and consider statistics. Consulting prison statistics, it found that Tamil Nadu had the maximum number of people in detention, for every single year between 2011 and 2021 (rising to 51% of the total number of detentions in the country). The Court observed that this could either mean that the state was descending into lawlessness, or that the state was abusing its ‘jurisdiction of suspicion’ (paragraph 6). Given that an overwhelming number of preventive detention orders (86%) were eventually set aside, and indeed, not even one was confirmed, option 2 was evidently more likely. The Court therefore observed that, enabled by the vague provisions of the Goondas Act, “preventive detention has become an instrument of convenience whereby such elements are dealt with on the sure knowledge that once a detention order is passed, such persons are bound to be jailed for at least 3-6 months, pending reference to the Advisory Board or a challenge before this Court by way of a petition for habeas corpus.” (paragraph 13)


Having established this “pattern of abuse,” the Court came to the issue of remedies. It noted that “the time has now come for this Court to explore the possibility of awarding damages in cases where detention orders are set aside after finding that it was invoked on wholly extraneous and irrelevant grounds, which amount to a conscious abuse of power.” (paragraph 24) The Court held that such “indifferent” detention orders would amount to a “constitutional tort”, and therefore give rise to a claim for damages (paragraph 26). Following the judgment in Rudul Shah, the Court characterised this as a public law remedy, flowing from Article 21 of the Constitution.


The Court then went on to find that in the first of the two cases before it, the trigger for compensation had clearly been met. The Preventive Detention Order had been issued on the basis that the detenue had abused a public servant – but given that the said abuse had taken place in a private space, it was the evident that the threshold requirement for the application of the Goondas Act – a threat to public order – had not been met, given the constitutional definition of “public order” (paragraph 35). The Court held that this clearly amounted to a “casual invocation” of the preventive detention law would violate Article 21 of the Constitution (paragraph 38). And because it was a constitutional violation, it overrode the Goondas Act “good faith” immunity for agents of the State. Crucially, the Court noted that this would be necessary in cases of “misuse of the preventive detention law against a detenue who could have been proceeded against under the available penal laws.” (paragraph 40) This is important, because the Court’s focus was not on proving ill-intent or malice against State officials (an almost impossible task), but the fact that preventive detention had been invoked in a case that could evidently have been dealt with under ordinary law. Indeed, this was the basis on which the Court set aside the second detention order as well (although it did not impose costs): in the case, the detenue had abused and bitten the finger of a police constable. The Court noted that while this was a heinous act, there was no reason why it could not be dealt with under ordinary penal law, without the need to invoke preventive detention provisions. The order was, therefore, set aside.


Compensation for wrongful arrest and detention is a very under-developed concept in India, with courts almost never holding the State to account (indeed, the Supreme Court notoriously refused a compensation claim from people who had been imprisoned for a decade until they were acquitted), on the basis that doing so would “demoralise” the police and the security agencies. In that sense, the High Court’s judgment is welcome, as are its strong remarks about the abuse of the preventive detention law. Also welcome is its focus on asking whether the alleged offence in question can be dealt with under ordinary law. It is worth remembering that preventive detention is meant to be an exception to the normal rules of criminal law, which frown upon imprisoning a person not for something they have done, but for something they might do. That principle, however, has long been lost sight of; the High Court’s judgment reiterates that crucial distinction between ordinary criminal law and exceptional preventive detention law, and warns against blurring the lines between the two.

That said, a few points arise. The first is that the High Court invoked the doctrine of “constitutional tort”, but it could do with more fleshing out. The contours of the doctrine vary in different jurisdictions (for instance, US constitutional tort is distinct from Irish constitutional tort); thus, the nature of the doctrine, and the legal standards that it encompasses, require some development; perhaps that is for a case for another day.

Secondly, there is the question of future enforcement. As noted above, our constitutional culture is yet to normalise the concept of compensation against the State for wrongful arrest; consequently, it is doubtful how effective the High Court’s judgment will be as setting out an effective and continuing deterrent. For this reason, I believe that the High Court could have gone further than it did. Once it had found – as it did – that there was a statewide pattern of abuse, it need not have restricted its finding to the case before it. Previously, on this blog, I have discussed the doctrine of an unconstitutional state of affairs, defined in the following way:

The unconstitutional state of affairs is a legal ruling that allows the Constitutional Court to acknowledge the failure of both the Legislative and Executive branches of government to enforce public policies against widespread and systemic violation of fundamental rights, thus justifying a judicial intervention in order to combat the structural causes of the violations and to put everything back in order with the Constitution. 

It is evident that the High Court’s own analysis pointed to the existence of an unconstitutional state of affairs: it found that preventive detention was being repeatedly invoked for frivolous purposes, and effectively to keep people detained for four to six months, until the order would be set aside by the advisory board or by the court. This was evidently a “widespread and systemic violation of fundamental rights.” The declaration of an unconstitutional state of affairs, therefore, would have enabled the Court to address structural causes through tools such as the continuing mandamus, where the bench in question could have commenced continuing oversight over the State’s use of preventive detention, with the possibility of more compensatory orders for breach. This, I submit, would be a more effective deterrent against State impunity.

And thirdly, it is questionable whether the actual remedy is “punitive” or deterrent in any meaningful way. The High Court ordered compensation of Rs 25,000, which – in practical terms – is nothing at all. In order to prove an effective deterrent, it is evident that the compensation amount must be increased by several degrees.


In sum, therefore, the High Court’s judgment breaks important new ground, and is a laudable first step towards combating State impunity in the invocation of preventive detention laws, and the normalisation of preventive detention laws. However, future judgments can now build upon this platform, and go further.

Guest Post: Should the Doctrine of Repugnancy Apply to Delegated Legislation?

[This is a guest post by Ashwin Vardarajan.]


Two recent judgments of the Supreme Court – Gambhirdan K. Gadhvi v. State of Gujarat (March 2022) and Sreejith P.S v. Rajasree M.S. (October 2022) – have brought to the fore an important issue of constitutional interpretation: does the doctrine of repugnancy, under Article 254 of the Constitution of India, allow a law made by a State Legislature to be subordinated to delegated legislation promulgated by the Union Government under a law of the Parliament? While the SC’s answer was a resounding ‘yes’, the reasoning employed by it is a continuation of a pre-existing constitutional anomaly which has existed under Article 254 for quite some time. Without questioning the correctness of the outcomes in Gambhirdan and Sreejit on merits, this essay argues that the reasoning employed by the SC in answering the aforesaid question is incorrect and deserves to be reconsidered.

The Decisions in Gambhirdan and Sreejit, and a Historical Background

Briefly put, Gambhirdan and Sreejit concerned the appointment of two Vice Chancellors (‘VC’) under state laws of Gujarat and Kerala, respectively, challenged for being in contravention of the method for choosing a VC prescribed under the University Grants Commission Regulations, 2018 (‘UGC Regulations’).

In Gambhirdan, the VC was chosen on the basis of a criteria prescribed by the selection committee established under the Sardar Patel University Act, 1955 (‘SPU Act’) as opposed to the UGC Regulations. The SC, in this regard, held that the SPU Act’s provisions were “contrary to the [UGC Regulations], which…are binding on the State Government and the universities thereunder”. While setting aside the concerned VC’s appointment for being contrary to the UGC Regulations, they held that “being a subordinate legislation, UGC Regulations becomes part of the [UGC] Act”, and that “any conflict between State legislation and Central legislation, Central legislation shall prevail by applying the rule/principle of repugnancy as enunciated in Article 254 of the Constitution as the subject ‘education’ is in the Concurrent List (List III) of the Seventh Schedule of the Constitution.”

In Sreejith, a similar question arose in respect of the appointment of the VC at the APJ Abdul Kalam Technological University (‘AKTU’). The challenge against the appointment of the VC was advanced again on lines of it being in violation of the UGC Regulations. Following Gambhirdan, the SC set aside the appointment of the VC at AKTU, noting that:

8.2…to the extent the State legislation is in conflict with the Central legislation including subordinate legislation made by the Central legislation under Entry 25 of the Concurrent List, the same shall be repugnant to the Central legislation and would be inoperative.” (Emphasis added)

In effect, the two decisions discussed above rule that a subordinate/delegated legislation promulgated by the Union Government under a law made by Parliament will have an overriding effect over a law made by the legislature of a State in accordance with Article 254. These decisions, however, are not the first of their kind.

One of the earliest instances encountering this issue was the SC’s decision in State of Jammu & Kashmir v. M.S. Farooqi (1972), where the All-India Services (Discipline and Appeal) Rules, 1951 were found to be in conflict with the State of Jammu & Kashmir Government Servants (Prevention of Corruption) Act, 1962. While relying on a plethora of decisions dealing with the doctrine of repugnancy, the SC held the State law to be repugnant to the 1951 Rules in accordance with Article 254. However, it must be noted that none of the decisions which the SC relied on to substantiate its ruling dealt with a case where a Central delegated legislation superseded a (repugnant) law enacted by the legislature of a State. Rather, all of these decisions concerned repugnancy between laws enacted by State legislatures and the Parliament.

Farooqi was also a deviation from the language of Article 254(1), which reads that if:

“…any provision of a law made by the Legislature of a State is repugnant to any provision of a law made by Parliament which Parliament is competent to enact, or to any provision of an existing law with respect to one of the matters enumerated in the Concurrent List, then…the law made by Parliament…shall prevail.” (Emphasis added)

Clearly, the wording emphasises on the enactment of a competent State legislature and the Parliament, and does not provide any room for delegated legislation made by the Union Government in its language. Naturally, thus, a delegated legislation must not be included within the fold of the doctrine of repugnancy. Farooqi also does not offer any explanation or reasoning as to why a delegated legislation superseded an enacted law of a State legislature by virtue of Article 254, despite the language indicating to the contrary. Later, in Annamalai University v. Information & Tourism Deptt. (2009), while dealing with the question of repugnancy between regulations made under the UGC Act and a State law, that the SC categorically ruled that “[subordinate] legislation as is well known when validly made becomes part of the Act (sic).”

This observation became the basis for courts, in subsequent decisions, to conclude that if a central delegated legislation promulgated under a Parliamentary law must be considered an extension of the latter, and thus supersede a law made by the State legislature under Article 254. For instance, in Kalyani Mathivanan v. K.V. Jeyaraj (2015), the SC relied on the decision in Annamalai and concluded that the State Madurai Kamaraj University Act, 1965 was repugnant to the UGC Regulations, 2010 (a delegated legislation under the UGC Act).

Thus, Farooqi, Annamalai and Kalyani collectively allowed courts to hold that delegated legislation promulgated under a Parliamentary law is an extension of the latter, and deserves to supersede a repugnant law enacted duly by the State legislature in accordance with Article 254. Gambhirdan and Sreejit are a continuation of this line of judgments.

Constitutional Concerns

It is submitted that the SC’s reasoning suffers from several infirmities.

Firstly, as noted above, and pointed out elsewhere too, the language of Article 254(1) itself does not permit a Central delegated legislation to supersede a law enacted by a State legislature. Notwithstanding that, Article 254 resides in Part XI, Chapter I of Constitution, which deals with ‘Legislative Relations’ between the Union and the States–a domain dealing squarely with relations between State legislatures and the Parliament, and not the State legislatures and the Union Government. To substantiate, we also see that the provisions under Part XI, Chapter I deal only with the distribution of legislative powers between the Parliament and State legislatures, whereas Chapter II deals with ‘Administrative Relations’ (i.e., relations between the Union and State Executives).

Second, a potential fallout of the SC’s interpretation is that the will of a legislature comprising of the elected representatives in the concerned Indian State is made subservient to delegated laws promulgated by the Union Government (which may comprise of ministers indirectly elected to the Rajya Sabha and/or possessing a lesser magnitude public confidence and trust). Furthermore, we may also read the per incuriam–but, it is submitted, correct–decision of the Calcutta High Court (‘HC’) in Medical Council of India v. State of WB (2012). Dealing with the question of repugnancy between the Bengal Medical Act, 1914, and rules and regulations under the Medical Council Act, 1956, the HC held that a “delegated legislation cannot run counter to supreme legislation” of a State by virtue of being below an enactment of a legislature in the hierarchy of laws. In other words, a State law can only be repugnant to a Central law’s provisions. The Bombay HC, in Suresh Patilkhede v. Chancellor, Universities of Maharashtra (2012), too subscribed to this line of reasoning, but was overruled in Kalyani.

Thirdly, Parliamentary laws frequently grant extensive authority to the Union Government to promulgate delegated legislations under vaguely worded provisions often devoid of any determining scope within which the Government is permitted to act. Resultantly, the Government often exercises essentially legislative functions–i.e., functions primarily performed by a competent legislature–without adequate checks and balances–a criticism noted by the SC as well. Therefore, to rule in a blanket way that the delegated legislation is a part, or an extension, of the Central law overlooks the problems it causes in the context of the doctrine of repugnancy under Article 254. In the absence of any safeguards to tackle it, the SC’s interpretation undermines the federal compact between the Parliament and State legislatures manifested in the Constitution.

Fourthly, the SC in Farooqi, Annamalai, Kalyani, Sreejit and Gambhirdan did not consider its decisions’ impact on Article 245(2). An exception to the doctrine of repugnancy, Article 254(2) permits a State law to prevail over a Central law if the same has been placed for consideration by, and received an assent to that effect from, the President of India. By including delegated legislations within the fold of Article 254(1), the SC impliedly has subjected State legislatures to undergo the tedious process–a process often influenced by political motivations–of obtaining the assent of the President in circumstances where the Union Government promulgates a delegated legislation which is inconsistent with the impugned State law. It is also well known that State or Parliamentary laws go through a much more rigorous process to get passed by the concerned legislature, as opposed to a delegated law promulgated often without any legislative sanction or scrutiny. Thus, despite there being no discussion on this front by the SC, ripple effects of their interpretation will inevitably impact the working of Article 254(2).


The doctrine of repugnancy under Article 254 has a direct bearing on the rights of States to enact legislatures under the Concurrent List, whereby the Parliament even may occupy the concerned legislative field even without prior consultation with the States–a concern also flagged by the Sarkaria Commission (here, ¶2.14.01). By (historically) allowing Central delegated legislations to prevail over laws enacted by State legislatures, the SC creates fissures in the balance of legislative powers distributed between the Union and States under the Constitution. In an appropriate case, therefore, the SC must reconsider and reverse its stance on this issue.

Proportionality Under Article 25?: A Response to Kartik Kalra [Guest Post]

[This is a guest post by Priyansh Dixit.]

In India Young Lawyers’ Association v State of Kerala (‘Sabrimala’), Justice Chandrachud observed the possibility of using proportionality as a limitation analysis in Article 25(1), as opposed to the Essential Religious Practice (‘ERP’) test. The case is currently pending before a nine-judge bench under the Court’s review jurisdiction. In line with Justice Chandrachud’s observations, it has been more elaborately argued elsewhere that proportionality should be applied to the limitation analysis of religious rights, as a superior alternative to the contemporary ERP test.

Such arguments have taken on new significance in light of the recent split division bench of the Supreme Court in Aishat Shifa v State of Karnataka, regarding the (in)famous Hijab issue. The case has been put before the Chief Justice of India and is likely to raise the issue of limitation analysis in religious rights, which would be echoed in the Sabrimala review. The important issue of applying a proportionality-based review can arise. Nevertheless, because of the critique that the ERP test is subjected to, and because of the importance that limitation analysis based on proportionality is given, the question of applying proportionality to Article 25(1) becomes important.

In assessing the application of proportionality to Article 25(1), the first step of the inquiry, before a normative analysis of the desirability of using proportionality in the adjudication of religious rights, would be to ascertain whether a legal foundation exists that justifies limitations analysis based on proportionality. Recently, in a guest post, Kartik Kalra, while criticising Justice Dhulia for sidestepping the ERP test in his judgement in the Hijab case, argued that a cleaner, judicially sound alternative would have been to use proportionality in Article 25(1) of the Indian Constitution, effectively contending that such a legal basis exists. His central premise was that because it has been held that no state action can be arbitrary, unreasonable, or fanciful (Maneka Gandhi v Union of India, Mithu v State of Punjab), and because no right can be abridged more than what is necessary to achieve a state’s legitimate objectives (K.S. Puttaswamy v Union of India), proportionality can be applied.

In this article, I take a stance contrary to Kalra. I argue that proportionality, as a structured four-prong structured test, cannot be applied to Article 25(1). Firstly, I will argue that standards of reasonableness, even though they have been read broadly into the constitutional scheme, do not justify reading proportionality as a limitation analysis in Article 25(1). This argument is not a contribution toward my actual inquiry into whether proportionality can be applied to Article 25(1). Rather, it acts as a rebuttal to Kalra’s argument that mere reasonability can justify the use of proportionality. To make such a rebuttal is particularly important because otherwise, any need for a specific inquiry into Article 25(1) would be rendered meaningless. Secondly, I will assess the threshold for establishing proportionality in a constitutional provision by analysing the case of Modern Dental College and Research Centre v State of Madhya Pradesh (‘Modern Dental’), where proportionality as a structured doctrine was established in Article 19’s context. I would further argue that the threshold is not met in the context of Article 25.

Note that I will not deal with normative issues of whether proportionality should be applied. Rather, I will limit myself to a formal inquiry in analysing if proportionality can be applied if required.

Also note that across the article, I use proportionality to refer to the structured four-prong test – where the four prongs are: (i) Legitimate Purpose (ii) Rational Connection (iii) Necessity (iv) Balancing – which is used to, as Aharon Barak explains, balance ‘constitutional’ and ‘sub-constitutional’ norms. The former are principles/rules enshrined in the Constitution (such as Article 25(1)), while the latter include legislations that limit the scope of such principles/rules. For the latter to be valid, they must conjunctively satisfy each prong of the test.


Let me briefly extend the argument that I am dealing with. What is being proposed is that any elaborate analysis of whether proportionality cannot be applied to Article 25(1), or any other provision for that matter, is unnecessary since the same has been done by Courts, which have read reasonableness as a requirement in every legislative action. This reasonableness is understood to be a requirement that runs across the constitution, including Article 25(1) by extension. If this contention is indeed true, then there is no point in further analysis of this article. But there are two reasons why this is not the case and why a more specific inquiry into the question of applying proportionality is needed.

Firstly, it must be noted that despite Courts reading reasonableness broadly into the constitutional structure, specific limitation analysis of any constitutional provision has still been considered important. Professor Aparna Chandra has argued that while the Supreme Court has stated that the requirement of reasonableness runs through the entire constitution, it does not consider the same to imply that there should be a single limitation test. This is evident by the multiple approaches courts have taken to analyse limitations on fundamental rights. Thus, whether standards of reasonableness can be used in Article 25(1) at all is a contested issue. To determine the same, the limitation of Article 25(1) and its interpretation by the Court must be specifically analysed. It will be argued later that any balancing exercise, like reasonableness or proportionality, cannot be read into Article 25(1) because of its limitation, and the judicial history of the same.

Secondly, even if we assume that reasonableness can be read into Article 25(1), the case is not necessarily the same for proportionality. This is because the two standards – reasonableness and proportionality – are different. They are often mistakenly equated because both entail balancing competing values, but two fundamental differences between the two illustrate otherwise.

One, the substantive content of both is different. While the proportionality test requires an independent review of the four prongs, there is no need for the same in assessing reasonableness. The mandate to analyse each step can considerably affect the adjudication process. A judge might reach a different conclusion when using reasonableness than while using proportionality when she is looking at the same set of facts. It can be argued that in cases like Chinataman Rao v State of MP and V.G. Row v State of Madras, where reasonableness was first introduced, strands of proportionality can be found. For instance, in V.G. Row, it was observed that the “disproportion of imposition” needs to be analysed while assessing the validity of right-restricting measures. Similarly, in Chintaman, it was held that:

“Legislation which arbitrarily or excessively invades the right cannot be said to contain the quality of reasonableness and unless it strikes a proper balance between the freedom guaranteed in article 19 (1)(g) and the social control permitted by clause (6) of article 19, it must be held to be wanting in that quality.” (para 6) [emphasis mine]

However, while using reasonableness, the approach of Courts has largely been deferential to the legislative wisdom. Consider the following observation from the Directorate of Film Festivals v Gaurav Ashwin Jain, which illustrate the approach this approach:

“Courts do not and cannot act as appellate authorities examining the correctness, suitability and appropriateness of a policy…Courts cannot interfere with policy either on the ground that it is erroneous or on the ground that a better, fairer or wiser alternative is available. Legality of the policy, and not the wisdom or soundness of the policy, is the subject of judicial review…”

The conceptualisation of judicial review here is that Courts are not supposed to explicitly look at whether a policy is necessary for its purported aim. The lack of the necessity prong makes a substantial difference since the prong gives Courts leeway to inquire into the appropriateness of policy which is otherwise exempted from judicial review. Professor Chandra has argued here that many cases, where proportionality has been ostentatiously applied, have deemed the alleged right-restriction measure to be valid, only because they improperly analysed the necessity prong. If courts had applied reasonableness as the standard of review, Professor Chandra’s critique would have had no legal basis at all. The point is that the content of both the tests and the results they produce are different. Thus, it cannot be argued that the mere presence of reasonableness warrants applying proportionality.  

Two, the evidentiary standards of both are different. Evidentiary standards comprise the burden of proof, the standard of proof, and the quality of evidence. The higher these standards, the greater the justificatory burden on the State to justify its right infringement. For instance, take burden of proof. In applying reasonableness, the burden of proof to establish that a certain restriction invalidly infringes rights and is unconstitutional is on the party alleging unconstitutionality. This is based on the well-established doctrine of presumption of constitutionality, which (i) places the burden of proving the unconstitutionality on the petitioner and (ii) raises the degree of evidence required to prove unconstitutionality, since presumption of constitutionality entails a strong assumption that the impugned legislation is compliant with part III rights.  In proportionality, however, the burden of proof is reversed. Justice Chandrachud observed in Puttaswamy (II) v Union of India, interpreting his plurality judgement in Puttaswamy (I) v Union of India:

“Proportionality requires the State to justify that the means which are adopted by the legislature would encroach upon the right to privacy only to the minimum degree necessary to achieve its legitimate aim” (para 32) [emphasis mine]

Reasonableness, as compared to proportionality, has lower evidentiary standards, as has been illustrated above. The point that I am again making is that the impact of a judicial review done using reasonableness would be different than if it is done using proportionality. Therefore, even if it is assumed that the standards of reasonableness can be applied to Article 25(1), proportionality cannot be read into it.

In summation, proportionality and reasonableness are two different kinds of judicial review. If limitation analysis of a certain provision requires the application of reasonableness, the same cannot justify applying the structured four-prong proportionality test. To apply the same, a separate interpretative process is required. What such a process would entail and what would be the result of applying it to Article 25(1) are issues that the next section deals with.


In the previous section, I proved that reasonableness does not justify using proportionality in Article 25(1). What is required is a more specific inquiry. In this section, I do exactly that through two steps. First, I will explain how Indian Courts sourced proportionality into Article 19 and distil the appropriate threshold that ought to be used to apply proportionality to other rights. Second, I will show why this threshold is not met in Article 25(1)’s context.


Proportionality as a structured doctrine was explicitly established for the first time by Indian Courts in Modern Dental in the context of Article 19. The Court’s reasoning to reach its conclusion regarding proportionality had two levels. Firstly, and on a more abstract note, the Court cited Aharon Barak, a renowned constitutional scholar, and the former Chief Justice of the Israeli Supreme Court, who identifies two sources of proportionality – democracy and rule of law – that warrant balancing constitutional rights against the public interest (para 62).  The Court worked with the premise that these sources are intrinsic features of India’s constitutionalism and thus reasoned that the said balancing is justified (para 62).  Secondly, the Court noted that the doctrine of proportionality is already built into the Article, using the term ‘reasonable restrictions’ in its text (para 65).  The court analysed case laws that have interpreted this scope of reasonableness and argued that it has been interpreted in a manner that would warrant a conjunctive use of all prongs of the structured doctrine of proportionality (para 65).

Here, a question arises: Can either of these conditions justify using proportionality as a limitation analysis for a provision or do both need to exist? It is submitted that the latter is the correct interpretation. If the contrary is correct, the implication would be that proportionality can be applied to any right since ‘democracy’ and ‘rule of law’, which forms the first condition, is a part of the basic structure of the Indian constitution and is thus a part of every provision. Such an interpretation has two issues. Firstly, Modern Dental’s analysis of Article 19’s limitation clause would be rendered useless. This fails to make sense since the Court imputed sufficient weight to the analysis. Moreover, this would also violate the point that specific limitation analysis of different rights ought to be done and that broad limitations cannot be read across rights, which was established earlier. Secondly, the deliberate differences made by the constituent assembly in different limitation clauses would be ignored. This would render the drafting exercise a futile endeavour, which would be contrary to constitutional interpretation.

Thus, to assess if proportionality can be applied to Article 25(1), an analysis of the limitation clause presents a valid method.  In the next sub-section, I undertake the analysis and conclude that Article 25(1)’s limitation clause disallows applying proportionality.


The limitation clause of Article 25(1) states that the right provided under the article (to freely profess, practice, and propagate religion) are “subject to public order, morality, and health and the other provisions of this part.”  This section, first, analyses the text of Article 25’s limitation clause and argues that because of how it is hierarchically sewn with other provisions of Part III, it does not warrant balancing exercises like proportionality and, secondly, argues that if the judicial history of Article 25 is analysed, as was done for Article 19 in Modern Dental, one concludes that it does not warrant applying proportionality.

Firstly, ‘Subject to… other provisions of this part’ leads to the conclusion that Article 25(1) is hierarchically inferior to other articles in Part III. Two arguments, regarding the interpretation of the term ‘part’, can be made against this interpretation. One, that the term ‘part’ refers to Article 25(2), and two, that it refers to Article 26. However, as has been argued elsewhere, these arguments do not hold. The former does not hold, because if the framers wanted to do so, they could have retained “this Article” instead of “this part” as was done in the earlier drafts of the current Article 19. The latter does not hold because to do that, the framers could have simply stated “subject to Article 26” as was done with some other provisions (Art 62(2), 68(2), 81(1), etc.). The specific wording, which was not used by the constituent assembly for other provisions, signifies that Article 25(1) was meant to be sewn into a hierarchy with other fundamental rights.

This argument also gains some force from Justice Chandrachud’s decision in Sabrimala. He had noted while holding that values like dignity and equality form overriding considerations against which religious rights would have to be sacrificed: “These constitutional values stand above everything else as a principle which brooks no exceptions, even when confronted with a claim of religious belief” (para 54). This is illustrative of the idea that some other constitutional values are in a supreme position vis-à-vis individual religious freedom.

In Sahara India Real Estate Corporation. Ltd v SEBI, the Court had held that the conceptual premise of any balancing exercise, such as proportionality, is the equality between these rights: “When rights of equal weight clash, Courts have to evolve balancing techniques or measures based on re-calibration under which both the rights are given equal space in the Constitutional Scheme” (para 42). Thus, Article 25(1), whose limitation clause puts it in an inferior position, cannot be balanced using proportionality.

Secondly, the specificities that were found in Article 19’s limitation clause are absent in Article 25(1). No standard of reasonableness has been provided in the text as has been done with Article 19 with the use of the word ‘reasonable’. Moreover, notions of reasonableness find no mention in the jurisprudence of limitations of Article 25(1). Instead, the history of its limitation analysis primarily consists of the ERP test, which simply justified state interventions where the concerned religious practices are not essential. What concerns us is how this essentiality has been analysed.

This essentiality, evident by court judgements, is demonstrably devoid of the requirement of reasonableness. In its nascent form, the ERP test evaluated the essentiality of practices as per the beliefs of the followers (See Commissioner v Shirur Mutt). Later, the jurisprudence flipped on its head, with the Court deciding the essentiality of a practice based on extraneous considerations as opposed to evaluating the faith of the believer. Such considerations include the optionality of the practice (See Ismail Farooqui v Union of India, Hanif Quareshi v State of Bihar) and the recency of the practice (See Acharya Jagdishwaranand Avadhuta v Comm. Of Police Calcutta). Evidently, the inquiries under the ERP test assess the concerned religious practice. No assessment of the state restriction is done, which is what an inquiry into reasonableness would entail. Therefore, the reasonableness of state restrictions is left out of the picture while applying the ERP test.

At this juncture, one might argue that if proportionality and reasonableness are different standards, why does it matter if the latter has not been used to adjudicate Article 25(1) cases to analyse if the former can be applied? At the cost of reiteration, I will have to restate my argument regarding reasonableness and proportionality to rebut this argument. Reasonableness and proportionality are closely related, and the existence of the former can lead to the conclusion that the latter might be justified. Therefore, I argue the lack of reasonableness in the adjudication of rights becomes a reason why the application of proportionality is not justified. But the two are still different in the sense that their application requires different lines of inquiry and can lead to different results. So, while related, the use of proportionality and a separate interpretative exercise is required to apply it. If the contrary happens, the character of reasonableness as a standard of judicial review would be drastically changed. Therefore, it is conceptually and legally consistent to argue that while a lack of reasonableness can be used to justify not applying proportionality, the two standards are different in the sense that one does not lead to the presumption that the other can apply.

In summation, proportionality cannot apply to Article 25(1). The standards that are required for the same cannot be established owing to the requirements of balancing, which are not met in Article 25(1)’s context, and because of the jurisprudence of Article 25(1), which has refrained from applying standards of reasonableness.


Whether proportionality should be applied to Article 25(1) is a normative question, answering which would require much research. In this article, I limited myself to a much-limited inquiry of whether proportionality can be applied to Article 25(1) if needed and answered it in the negative. I began with the argument that standards of reasonableness cannot be used to justify proportionality in India. Then, I argued that to establish proportionality, the limitation clause of Article 25(1), and its judicial interpretation, needed to be analysed and that on such an analysis, it is found that proportionality could not be applied to the article.

Notes From a Foreign Field: The US Supreme Court’s Latest “Gun Rights” Decision [Guest Post]

[This is a guest post by Anant Prakash Mishra.]

On June 23, 2022 the American Supreme Court gave its ruling in the case of N.Y State Rifle and Pistol Assn. vs. Bruen, which left the political spectrum of the nation divided. Conservatives largely hailed the court’s verdict as championing their right to ‘bear firearms’ as enshrined in the constitution. Most liberals, on the other hand, protested against it with misgivings about the aftermath of the verdict and viewed the state of affairs with skepticism. Even among constitutional law scholars, there has never been a consensus as to what qualifies as a correct interpretation of the Second Amendment. Interestingly, the second amendment of the US Constitution mentions that “a well regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.”

In order to understand the root of the debate, let me first sketch out the facts which led to the present dispute before the US Supreme Court.

The State of New York had a law that in order to carry a fire-arm ‘inside’ or ‘outside’ of one’s home; the concerned individual would have to procure a license by satisfying a ‘proper cause’ requirement under the statute. An individual who satisfied that a ‘proper cause existed’, ‘may’ at the discretion of the authorities be issued an unrestricted license to carry such weapon in a concealed manner outside of their homes. The petitioners in the instant case applied for a license and were denied the same. Hence, they challenged this law, which had been in place for about a period of 109 years. They were unsuccessful in the inferior courts but ultimately succeeded at the US Supreme Court. The Supreme Court in a 6-3 verdict, held that the New York state law ran afoul of the Fourteenth Amendment of the American Constitution as it created an impediment for a citizen in the exercise of their Second Amendment right. The verdict was partisan in nature as the Court’s Conservative majority leaned on the side of relaxed gun ownership while the liberal minority joined the dissenting camp led by Justice Stephen Breyer.

Why the majority has erred in the judgement

The language of the Second Amendment is rather uncanny and poses a significant question as to whether the ‘right to bear arms’ was mentioned only in the context of a militia? The usage of the word ‘militia’ and ‘bear arms’ in the same vein does hint towards this interpretation. If this premise were to be accepted, then no individual will have a constitutional right to bear arms.

However, the position was clarified by the US Supreme Court in its 2008 decision in the case of District of Columbia vs Heller. The Court ruled that there is a constitutionally protected right to keep fire-arms in one’s homes for self defence. It was again re-affirmed in 2010 in the case of McDonald vs Chicago.

Yale law Prof. Akhil Amar, in his book ‘The Constitution Today’ has argued that keeping  firearms is ingrained in the American way of life and the same has a constitutional basis rooted in history. Looking from this perspective, it appears that the Supreme Court’s ratio decidendi in the Heller case is tenable and judicially sound. However, the present decision is stretching the relaxation too far and the same appears to be constitutionally unfounded. ‘Keeping a gun in one’s house’ and ‘carrying a gun in public’ cannot be put on the same pedestal because the consequences of both lie on different tangents.

If in fact, the state had a propensity to exercise its administrative discretion on license arbitrarily; the same could have been reviewed judicially on a case-to-case basis. The Supreme Court could have avoided the one-scoop sweeping approach by which it struck down the New York statute’s ‘proper cause’ requirement.

The obligation to show ‘proper cause’ in furtherance of issuing an unrestricted gun license for outdoors was a very reasonable condition formulated by the state of New York. In actual practice, the courts of New York while determining the ‘proper cause’ requirement judged  whether ‘the person in need of an unrestricted license has in fact a greater degree of danger than the general community’. Although, the Supreme Court majority frowns upon this standard and terms it as ‘demanding’; in my view, this is a perfectly balanced approach. It is to be understood that the state has a legitimate interest in regulating the said right because it is trying to ensure the safety of its own people. Additionally, not a single word in the American Constitution or any historical material on record has ever indicated that the ‘right to bear arms’ is an absolute and unconditional right. Imposing a sound and justified criterion for issuing a gun license cannot per se be deemed to have an ‘infringing’ effect on the right in question.

The majority in this case rejects the ‘means-end approach’ and instead adopts an amorphous ‘historical approach’ to adjudicate upon the dispute in question. What seems interesting is that the court appears more absorbed in the history of the second amendment combined with the framers’ will to unravel the proper context of the right. The court almost exclusively relies on the ‘Heller case’ (aforementioned) to dig into the American tradition with regards to the gun laws. The obvious irony here is that the law in scrutiny itself is a 109 years old legislation.  It is to be understood that there is only a limit to what history can teach us; under no circumstance can it serve as an exclusive guidebook for posterity. These shortcomings/limitations of an originalist interpretation admittedly involve a long-running debate and have now been understood by multiple jurisdictions. The United States is an outlier in this regard as no other jurisdiction across the globe is strictly wedded to originalism.

What does the verdict mean

While America is still in mourning and acutely wounded from the incessant episodes of harrowing mass shootings at schools and other public places, this judgement in my view further aggravates the brunt. Simply put; with the ‘proper cause’ requirement revoked, it would be well understood that the second amendment enables the citizens to bear arms in public. The question that still remains unanswered is, “In a civilized-modern state are guns necessarily the need of the hour for the citizenry?”  The court’s indifference in realizing the perils of the verdict is something that needs to be discussed and debated. The fact that more than 45000 Americans died in gun violence in the year 2020 is enough to indicate that the issue at hand is grave. In parting, I wish to express that the ‘second amendment’ or the ‘gun control’ debate is much more than a discussion on constitutional interpretation; it is a human rights issue and should be perceived that way!

Guest Post: Pension Cessation and the Rekindled Framework under Article 15(3) [Part II]

[This is a guest post by Kartik Kalra. This piece is the second of a two-part series on the evaluation and application of a rekindled framework to test protective discrimination under Article 15(3). Part One can be accessed here.]

In the previous post of this two-part series, I have proposed a four-pronged test that is capable of navigating the thin line between recognizing and entrenching stereotypes when evaluating the saving of discriminatory laws under Article 15(3). To recall, this test consists of the following prongs:

  1. the law must be based on stereotypes that are quantitatively verifiable;
  2. the stereotype must be perpetuated in pursuance of remedying historical disadvantages faced by women;
  3. there must be no availability of a lesser-discriminatory alternative than the impugned law; and
  4. the advantage conferred on the historically disadvantaged group (i.e., women) must be greater than the disadvantage generated via the perpetuation of stereotypes.

In this piece, I apply this framework to the pensionary regime in general and remarriage-related Pension Cessation Rules (“PCR”) in particular, demonstrating the test’s ability to dismantle stereotypes without causing tangible disadvantages to women as a class. In doing so, I first introduce the judgement of the Queen’s Bench Division in Green v. Metropolitan Police Commissioner, also discussing the nature of Indian PCRs and the type of constitutional challenge they are susceptible to. Following this, I evaluate PCRs within Article 15(1), showing their indirectly discriminatory nature. Finally, I apply the four-pronged test to PCRs and propose a possible narrowing of the sex-based class to a dependency-based one, preventing the active entrenchment of stereotypes using the test’s third prong.

Remarriage Causing Pension Cessation

Earlier this year, the Queen’s Bench Division of the United Kingdom delivered its judgement in Green v. Metropolitan Police Commissioner, where it upheld a rule that denied pensionary benefits to widows on remarriage. Sharon Green, the lead petitioner, was the wife of a police officer who died during service. After twelve years of staying single and surviving off the pension she received following her husband’s death, she found a partner near her street with whom she envisioned a future. Regulation C9 of the Police Pension Scheme, 1987, however, stood as an enormous obstacle, which mandated a cessation of pensionary benefits on her remarriage or on continued cohabitation with a partner. Green was presented with two options: either continue to survive off the pension and abjure from entering into a romantic relationship, or to enter into one but sacrifice her livelihood.

The condition of remarriage leading to a cessation of pensionary benefits is omnipresent in Indian law, expressly mentioned u/ss. 50(8)(a) of the Central Civil Services (Pension) Rules, 2021, 49(6)(i) of the Tamil Nadu Pension Rules, 1978 and 6.17(4)(i)(a) of the Punjab Civil Services Rules, 1984. This condition is also present, inter alia, in the service rules formulated by Haryana, Jammu and Kashmir, Madhya Pradesh and Rajasthan. Judicial review of such rules along with the regime of pensionary benefits presents a perplexing challenge that I have discussed in the previous post, given that the law’s validation perpetuates sex-based stereotypes and its invalidation causes immediate, tangible disadvantages to women.

The challenge to rules that have the effect of perpetuating sex-based stereotypes occurs within Article 15(1), which prohibits discrimination (both direct and indirect) on grounds of religion, race, caste, sex and place of birth. Such rules can be saved under Article 15(3) if they constitute a “special provision for women”. Presently, under Article 15(1), the constitutional assessment of laws perpetuating stereotypes via direct discrimination yields a direct conclusion of unconstitutionality as held in Navtej (¶438-9), while those doing the same via indirect discrimination is subject to a proportionality-backed justification by the state as held in Nitisha (¶87).

PCRs and Article 15(1)

At this stage, it would be important to note that Indian PCRs, as opposed to the unamended Regulation C9 of the United Kingdom, do not generally discriminate directly on the basis of sex. An exception to this could be ss. 87 of the Pension Regulations for the Air Force, 1961 and 133 of the Pension Regulations for the Army, 2008, both of which deny pensionary benefits to a widower. By and large, however, most states recognize the eligibility of both widows and widowers to obtain pensionary benefits that cease at remarriage. The constitutional challenge, therefore, lies in indirect instead of direct discrimination.

Indirect discrimination under Article 15 concerns disparate impact: neutral rules affect particular groups differently due to pre-existing structural disparities. It serves as a mode to confirm the denial of equality of opportunity from the disproportion of outcomes. Nitisha adopted the test in Fraser v. Canada to address indirect discrimination, which asks whether a neutral rule disproportionately affects a particular group, and whether the impugned rule has the effect of reinforcing, perpetuating, or exacerbating the disadvantage faced by that group (¶86). There is no defined standard of empirical evidence necessary to establish indirect discrimination, and the Court is empowered to make use of its common sense in assessing disparate impact (¶82). Who would be affected disproportionately by a law on the cessation of pensionary benefits on remarriage, men or women?

Data shows that men occupy a disproportionate number of jobs in state employment, and therefore widows are the primary recipients of pensionary benefits until they remarry. PCRs act as a deterrent in their remarriage, for the law demands that they choose between a financially secure life of solitude or an impoverished life of love. Either way, women as a class are disadvantaged: socially and psychologically if they choose the former, and economically if they choose the latter. It would be important to note that Green held that PCRs cannot be characterized as infringing privacy, for the effect of the law is not determinative of the engagement of constitutional rights (¶78). In Indian jurisprudence, however, it is solely the effect of a limitation on the individual that determines the engagement of a constitutional right. PCRs have the effect of preventing women from exercising their rights of intimate decision-making in choosing their marital (or otherwise) partners, which has been repeatedly held to constitute a facet of their right to privacy. Shakti Vahini v. Union of India held the choice to marry a person of one’s choice to be a facet of individual autonomy; Shafin Jahan v. Ashokan KM held the “choice of a partner whether within or outside marriage” to lie “within a core zone of privacy, which is inviolable”; Arunkumar v. Inspector General of Police held the decision of a transgender woman to marry a man to be a facet of her privacy; and Indra Sarma v. VKV Sarma held the decision to cohabit without marriage to be an intimate and personal decision.

The state, in effect, makes its citizens choose between a fundamental right and a compelling financial necessity, which leaves them with no real choice. In literature and case law, a state demand of the citizen’s relinquishment of a fundamental right is navigated within the framework of the doctrine of unconstitutional conditions. Under this doctrine, the way it has been developed in India, the engagement of a fundamental right is triggered whenever the affected individual can show that the state’s demand is “unreasonable in the special sense that it takes away or abridges the exercise of a right protected by an explicit provision of the Constitution” (St. Xaviers’ College v. State of Guajrat, ¶158). This is undoubtedly the case with PCRs, for the state demands that an individual choose between their rights of intimate decision-making or their financial well-being. There is, therefore, a prima facie engagement of their fundamental rights. The evaluation of the reasonableness of the restriction via PCRs, however, is separate, and would occur within an Article 21 enquiry.

In sum, PCRs disproportionately affect women and have the effect of depriving them of their freedom of intimate decision-making. On this basis, I submit that PCRs, as they currently stand in India, are indirectly discriminatory against women.

PCRs and the Four-Pronged Test under Article 15(3)

At this stage, it must be noted that a failure in meeting the Fraser criteria does result in a direct conclusion of the law’s unconstitutionality, for the law can be saved by a proportionality-based justification offered by the state. While this justification can be offered to save the law from being indirectly discriminatory in the first place, a separate exercise will have to be carried out to save it using Article 15(3). The proposed four-pronged test is of a higher scrutiny as compared to the traditional proportionality test, for it additionally requires the elimination of over-inclusiveness in the law’s application. Given the higher scrutiny of the proposed test, I shall directly examine the possibility of saving the law under Article 15(3) instead of evaluating its justifiability under Article 15(1). Lastly, it must be noted that PCRs would be saved by Article 15(3) only insofar as they function to cease the pension of widows, not of widowers.

  1. Quantitative Verifiability

The stereotype used and perpetuated by PCRs pertains to the economic dependence of women: they were first dependent on their earlier husband, then on the state, and now on their new husband. The verification of this stereotype must consider, inter alia, the proportion of women’s representation in state employment (around 20%), sex-based division of the total recipients of pensions, and the proportion of unmarried widows dependant on the state (around 90%). The empirical standards pronounced in Nitisha may also be deemed to apply here, meaning that the use of common-sense in the task of determining the quantitative verifiability of stereotypes shall be encouraged (¶82). In the South African Constitutional Court’s judgement in President v. Hugo, the judges concur that the stereotype of women as primary care-givers is true in fact without necessitating intense fact-finding endeavours (¶37, 77, 113). Economic dependence of women, therefore, is a verifiable stereotype and PCRs would satisfy the first prong.

2. Remedying Historical Disadvantage

The economic exclusion of women is a historical process of the solidification of the public-private divide, via which women have been relegated to the domain of the household where the “writ of the king doesn’t run”, and the man becomes the king, the sole breadwinner of the household. It is this historical subordination of women that pensionary benefits function to allay: they recognize that structural factors cause the present economic dependency of women, and that the state takes on the task of eliminating vagrancy that may be caused due to unfulfilled dependency. Correspondingly, PCRs also acknowledge this historical disadvantage by ceasing pensionary benefits when the woman has another partner. Fundamentally, this prong ensures that the vice of essentialism is not committed by portraying women as subordinate or inferior, instead acknowledging the historical and structural reasons that necessitate protective measures.

3. Availability of Lesser-Discriminatory Measures

The test of finding lesser-discriminatory alternatives requires, therefore, the investigation of a narrower class, the application of the law on whom doesn’t jeopardise stereotyping an entire sex-based (or otherwise) class a certain way. There exist such alternatives that are presently operational in India. The class subject to PCRs is the same as that receiving pensions, the possible narrowing of which must be evaluated to determine the availability of lesser-discriminatory alternatives. Consider, for example, Section 133(b) of the Pension Regulations of the Army, 2008, which introduces an additional variable to determine the eligibility to receive a widow’s pension: the maintenance of children. This law holds that if the widow is supporting the children borne out of her previous marriage, she shall be entitled to a full pension; and if she isn’t supporting her children after remarriage, she will be entitled to an amount between 30 and 60% of her original pension. The class eligible for the receipt of a full pension becomes smaller and reflects the variables affecting true dependency, veering the classification away from a sex-based one to a dependency-based one. In addition, the Indira Gandhi National Widow Pension Scheme applies to women who earn below a prescribed level, and the Swadhar Greh Scheme applies to women who are victims of difficult circumstances and require institutional support. There exist, therefore, alternatives measures that are able to narrow the classification, avoiding overinclusiveness and refraining from perpetuating the stereotype of women as dependent. PCRs can, therefore, be remodelled to cease benefits on the attainment of economic independence, not on a change in marital status.

4. Weighing of Advantages and Disadvantages

Given the availability of lesser-discriminatory alternatives that are able to meet the state’s eventual goal of preventing vagrancy, the law would have to be re-evaluated to make it conform to the third prong. For the fourth prong, we must weigh the advantages that women as a class receive under the pensionary regime with the disadvantages caused to them due to the perpetuation of sex-based stereotypes. Judges would arrive at different outcomes for this prong, as the contrasting reasonings offered by O’Regan and Krigler JJ. in President v. Hugo show. While O’Regan J. held that a law cognizant of stereotypes can confer meaningful advantages on women in the present (¶112), Krigler J. held that the perpetuation of sex-based stereotypes is inevitable in their recognition and must be avoided to the greatest extent (¶80).

I consider that an unwavering commitment to constitutional ideals at the cost of livelihoods is also bound to produce sub-optimal outcomes. A reading of this prong must ordinarily be construed in favour of the impugned stereotypical law, given that it reaches this stage after a confirmation of its application to the narrowest possible class. The pensionary regime confers an obvious advantage on women as a class in the form of tangible benefits of pensions, and a declaration of its unconstitutionality due to cessation on remarriage is bound to produce significant disadvantages for women in the present.


Through this two-part series, I have attempted to demonstrate the degree of perplexity one encounters in navigating the conflict between the pursuit of constitutional ideals and the immediate concern of addressing vagrancy via the recognition of verifiable stereotypes. The present framework under Article 15(3) proves inadequate to navigate this question, for it lacks a determining principle to save some laws and reject others. In response to this, I developed a four-pronged test that requires that the investigation of a narrower class within the sex-based classification to whom the impugned law applies. The four-pronged test, while easily susceptible to a critique of the kind levelled in Frontiero v. Richardson concerning the greatest possible avoidance of all sex-based classifications, bears fidelity to social iniquity in fact instead of a dogmatic commitment to constitutional ideals. This, I propose, best balances the twin considerations of dismantling stereotypes and eliminating vagrancy.

The dilemma between the recognition and entrenchment of stereotypes can be vividly observed in the functioning of pensionary benefits, a site for the four-pronged test’s application. The application of this test to the pensionary regime yields a general conclusion of the law’s constitutionality, provided that the class to whom they apply is narrowed, and the classification veered away from a sex-based one to a truly dependency-based one. The process of dismantling stereotypes is long and sustained, one where the law both follows and guides social progress. Immediate declarations of unconstitutionality benefit none, eliminating stereotypes only in law while they continue to thrive in fact. The four-pronged test, therefore, serves as a meaningful instrument in navigating this dilemma, upholding advantageous stereotype-based laws at the present while guiding the stereotypes’ ultimate dismantling.

Same-Sex Marriage Before the Supreme Court: Three Issues

Today, the Supreme Court issued notice on a batch of petitions seeking the recognition of same-sex marriages under the Special Marriage Act of 1954 [“SMA”]. Section 4 of the SMA refers to a marriage between “two persons.” Section 4(c) states, however, that a marriage may only be solemnised if “the male has has completed the age of twenty-one years and the female the age of eighteen years”, implying that an SMA marriage must necessarily be between a man and a woman. The PIL petitions challenge this implicit exclusion of same-sex couples from solemnising their marriages under the SMA.

This post will flag three issues that arise in these petitions.

The SMA as Secular Law

The scope of the challenge makes this an extremely straightforward case – arguably even more straightforward than the Supreme Court’s decriminalisation of same-sex relations in 2018. There is little doubt that the provisions of the SMA discriminate on grounds of sexual orientation, and therefore violate Articles 14 (equality before law) and 15(1) (non-discrimination on grounds of sex) of the Constitution (as held in Navtej Johar). Additionally, they infringe the Petitioners’ right to privacy and decisional autonomy under Article 21. To save the SMA, then, it must be shown that the restriction of constitutional rights is for a legitimate State purpose, and is proportionate.

If we look at same-sex marriage cases around the world, we find that there is essentially one argument that is open to the State: that marriage, properly defined, must be between a man and a woman is a deeply embedded cultural (and religious) norm, which is articulated through law. This – according to the State – is the basis of legally defining the institution of marriage in this way, and justifies denying same-sex couples the right to participate in it.

Whatever the merits of this argument in general, it is inapplicable to the SMA. This is because the SMA is explicitly framed as a secular law, which is open to people who do not want to get married under applicable personal laws (including in the case of inter-faith marriages). Not only is this evident from the history of the SMA, but the long title itself stipulates that it is “an Act to provide a special form of marriage in certain cases.” Religious or cultural conceptions of marriage, therefore, do not constitute the bases of the SMA; indeed, the SMA establishes a legal form of marriage that is an exception to marriage grounded in personal laws. Once this is understood, it is clear that the only argument open to the State – that marriage needs to be defined a certain way because it is a cultural/religious institution – is immediately undermined. Indeed, if the Court was to grant the petitioners’ prayers, it would be legalising same-sex marriage only under the aegis of the SMA; under personal law, marriage would continue to be defined as it has been.

For this reason, the case – so long as it is restricted to the SMA – presents a very straightforward, interpretive issue before the Court, which can be resolved in an equally straightforward manner.

The SMA as a Double-Edged Sword

The narrow scope of the challenge, however, raises certain questions about efficacy. Notoriously, the SMA places serious procedural hurdles before couples who want to get married under its aegis. Section 5 of the SMA requires a “notice” of an intended marriage to be given thirty days in advance, to the Marriage Officer. According to Section 6, this notice is to be published in a Marriage Notice Book (open to inspection), and to be “affixed in a conspicuous place” by the Marriage Officer. Section 7 authorises “any person” to object to the marriage within thirty days on the ground that the Act’s preconditions have been violated, following which, the Marriage Officer can refuse to solemnise the marriage if they deem fit (section 8).

This schematic description of the SMA should be enough to show that its viability as an option is inversely proportional to your social power. Indeed, studies of the SMA have repeatedly shown that the thirty-day notice period and the “affixation” of the notice in a “prominent place” facilitates social pressure upon vulnerable couples, especially in the context of inter-caste and inter-faith marriages. Indeed, the perversity of the SMA lies in the fact that those who will need to use it will – often – be precisely the kinds of couples who will need the shelter of anonymity: couples marrying against the wishes of their families, especially under the shadow of threats of violence and ostracism. It is to these couples that the SMA denies what they most need.

Thus, were the Supreme Court only to interpret the SMA to allow same-sex marriages, it would no doubt be an important move, but also, its benefits would be unequally distributed. Just like the SMA has failed inter-caste and inter-faith couples in the past, it will fail queer couples who circumstances make them vulnerable to social pressure and stigma.

The challenge, thus, is incomplete: in order to guarantee the full and effective enjoyment of constitutional rights, the Supreme Court ought also to interrogate the structure and design of the SMA, and to ensure that any consequential orders it passes are meaningful for the entire spectrum of the community. While the petitions may not categorically raise this point, it is of course open to the Court – in a constitutional case – to craft its own remedy.

A Note About Procedure

Finally, a quick point about procedure: cases asking for the recognition of same-sex marriages under the SMA have been pending before the Kerala and Delhi High Courts for a while. Today, when the Chief Justice pointed this out, it was argued that the state government intended to file an application for the transfer of all pending cases to the Supreme Court. The Chief Justice then issued notice.

However, whether or not the government was going to file for a transfer to the Supreme Court, in my view, there is merit to allow constitutional courts that have already been seized of an issue, to hear it; not only does this respect the comity of courts and the structure of the judiciary, but also, allows the Supreme Court to have the benefit of (one or more) reasoned High Court judgments when it finally considers the issue. Previously on this blog, I have criticised the Supreme Court’s tendency to pick up ongoing challenges before High Courts and hear them itself – effectively bypassing a crucial constitutional forum. Admittedly, this critique has come in the context of the Supreme Court stopping High Courts from hearing cases and then delivering judgments in favour of the State; at the same time, though, the force of the critique does not depend upon outcomes. Indeed, if we’re to look at this in a principled way, we have to critique the Supreme Court for bypassing procedure, especially in cases where the outcome is one that we agree with.

In my view, therefore, it would have perhaps been more appropriate for the Court to stay its hand, while requesting the Kerala and Delhi High Courts to decide the matter expeditiously (as it does raise important constitutional issues). Nonetheless, that said, we are here now; and one may look forward to a thoughtful and reasoned judgment in this case in the next few months.

Once Again, the Basics: The Bombay High Court’s Bail Order in Anand Teltumbde’s Case

On 18th November 2022, a Division Bench of the Bombay High Court passed an order granting bail to Anand Teltumbde, one of the accused in the “Bhima Koregaon Case.” The order is significant, because it is the first time that bail has been granted on merits to one of the “BK-16”, all of whom have been arrested and imprisoned under the Unlawful Activities Prevention Act [“UAPA”]. There have been cases of default bail (Sudha Bharadwaj) and medical bail (Varavara Rao), but so far, all bail applications on merits have been rejected (for example, Hany Babu and Jyoti Jagtap).

The Bombay High Court’s grant of bail to Anand Teltumbde continues the battles around personal liberty on the terrain of the UAPA. This blog has covered these battles in some detail. To recap, section 43(D)(5) of the UAPA statutorily bars a judge from granting bail if there are “reasonable grounds” to believe that the allegations against the accused are prima facie true. As, at the time of bail, the judge will only be considering the prosecution’s version, section 43(D)(5) effectively makes the defence fight with one arm tied behind its back. Furthermore, in the Watali case, the Supreme Court held that courts were not expected to scrutinise the prosecution’s material in great detail, and only make their assessment on the basis of “broad probabilities.” This was akin to tying both of the defence’s arms behind its back, throwing it into the sea, and ordering it to swim for the shore.

Notwithstanding the text of Section 43(D)(5) and the judgment in Watali, some judges have nonetheless been working through the interstices of the law to craft what I have called a “jurisprudence of liberty” in the shadow of the UAPA (see the discussion on this blog in “Back to the Basics” [Delhi High Court], “Staying With the Basics” [Bombay High Court] and “Entrenching the Basics” [Supreme Court]). To recap, this jurisprudence of liberty has two core features:

Principle 1: The definitional clauses of the UAPA must be given a strict and narrow construction. This was what the Delhi High Court did with respect to the meaning of “terrorism” in Asif Iqbal Tanha, what the Bombay High Court did with respect to Section 20 of the UAPA in Iqbal Ahmed Kabir Ahmed, and what the Supreme Court did in Thwaha Faisal.

Principle 2: The allegations in the chargesheet must be individualised, factual, and particularistic. The gap between what an individual is accused of, and the actual events, cannot be filled by inferences or speculation.

The jurisprudence of liberty, however, is an idea under siege. Other courts and other judges have refused to follow it, preferring instead a jurisprudence of State impunity, where the terms of the UAPA are interpreted extremely broadly, the Court itself plays the role of filling in the gaps in the Prosecution’s cases through inferences against the accused, and cites Watali as an overarching “stay in jail” card (see “Forgetting the Basics“). Indeed, not only is this conflicting jurisprudence present within the same Court and the same case, but at times the same judge has signed on to judgments that reflect both the jurisprudence of liberty, and (one year later) the jurisprudence of State impunity.

Anand Teltumbde’s bail order is important not just for its outcome, but also because, in this ongoing battle, it strikes an important blow for the jurisprudence of liberty. In the High Court’s judgment, one can find the deployment of both Principles 1 and 2, in painstaking detail.

As the judgment records, the NIA’s case against Anand Teltumbde was not that he participated in acts of terrorism, but that he was a high-level intellectual authority within the banned Communist Party of India (Maoist) [“CPI-Maoist”]. The Prosecution argued that Teltumbde was funded by the CPI-Maoist to give speeches abroad and that he brought back incendiary material that he then passed on to the CPI-Maoist through his brother, the (now deceased) Milind Teltumbdge; that he was one of the organising minds behind the Bhima Koregaon event, and therefore complicit in the “larger conspiracy”; and that he regularly provided intellectual succour and support to the activities of the CPI-Maoist. The Prosecution’s evidence for this consisted of (a) a set of documents (recovered from a co-accused, Rona Wilson), many of which were evidently addressed to one “Anand”; (b) a receipt showing transfer of funds from the CPI-Maoist to one “Anand T”; and (c) the testimony of certain witnesses regarding meetings between Milind Teltumbde and Anand Teltumbde.

The Division Bench began its analysis by noting the judgment of the Andhra Pradesh High Court Devendra Gupta, where a set of parameters had been laid down for determining what constituted a “prima facie” case under the UAPA: these included evidence of association with banned organisations, prior convictions for terrorist acts, recovery of explosive material, presence at the event, and a swift arrest after the event (readers will note that most of these are highly specific and particularistic) (paragraph 14). With this analytical framework in mind, the Court turned its scrutiny to the Prosecution’s evidence. It noted that (a) none of the letters allegedly addressed to “dear Comrade Anand” or “Comrade Anand” had been recovered from the possession of Anand Teltumbde himself, and (b) that even if one was to assume that the letters were addressed to Anand Teltumbde, there was nothing in them that showed active membership of the CPI-Maoist, or active complicity in terrorist acts (paragraphs 18.1, 18.2, 18.4.3). For example, at one point, the Court observed:

This letter has not been recovered and seized from Appellant. Assuming at the highest that reference in this letter i.e. ‘brother Anand’ is to the Appellant himself, prosecution needs to show the nexus and link of Appellant with the present crime or any specific overt act. There is no material save and except calling upon us to presume that the word ‘brother Anand’ named in the present letter is a reference to Appellant and as such he is directly involved with the activities of CPI(M). It is to be noted here that, this letter refers to names of 17 persons in all, including “brother Anand”. Some names are also with their phone numbers. Not all of these 17 persons have been indicted in the present crime. (para 18.4.3)

This paragraph is important, as it demonstrates the application of both prongs of the jurisprudence of liberty. What was presented to the High Court was a set of letters – not recovered from Anand Teltumbde – which were addressed to “Anand”, and which – in certain vague and general terms – spoke about his involvement with the CPI-Maoists. At the first level, the High Court refused to fill in the gaps in the Prosecution’s case by simply assuming that “Anand” referred to Anand Teltumbde (indeed, later the High Court would point out that there was more than one “Anand” who was part of the CPI-Maoists (para 18.6). At the second level, the High Court noted that even if – at the highest – the letters were referring to Anand Teltumbdge, the UAPA would kick in only if there was complicity with a specific crime or a specified overt act. Thus, it was not good enough for the Prosecution to engage in a roving exercise seeking to condemn individuals on the basis of their ideology, or generalised suspicion. In that sense, the UAPA’s provisions would have to be construed in a limited fashion, so that they would not end up sanctioning prosecution for thought-crimes.

The High Court similarly found that the evidence for the transfer of funds was lacking. Significantly, wherever there was a gap in the Prosecution’s case, instead of filling it under the umbrella of “conspiracy”, the High Court stressed repeatedly that for it to make the presumptions that the Prosecution wanted it to make, “would require further corroboration and evidence.” In other words, bail could not be denied simply on the basis of suspicion and conjecture (paragraph 18.7.1).

Finally, both aspects of the jurisprudence of liberty were evident in the High Court’s consideration of the witness testimonies: the witness statements contained generic allegations of Anand Teltumbde being an ideological fellow traveler of the CPI-Maoists, but – as the Court noted – they did not establish that he had been with a CPI-Maoist member or specifically met Milind Teltumbde (paragraphs 19 and 20). However, even if the NIA’s allegations were to be accepted, the Court noted that at the highest, this showed that Anand Teltumbde was a member of the CPI-Maoist. Under the UAPA, the maximum term of imprisonment for membership was ten years; the more stringent provisions of the UAPA (for example, Section 18) required the involvement of the accused in a terrorist act (crucially, the High Court held that under section 18, one would have to show conspiracy or abetment to a specific terrorist act). Thus, the Court noted in paragraph 22.2:

Section 20 cannot be interpreted to mean that merely been a member of a terrorist gang would entail such a member for the above punishment [i.e. life imprisonment]. What is important is the terrorist act and what is required for the Court to see is the material before the Court to show that such a person has been involved in or has indulged in a terrorist act. Terrorist act is very widely defined under Section 15. In the present case, seizure of the incriminating material as alluded to hereinabove does not in any manner prima facie leads to draw an inference that, Appellant has committed or indulged in a ‘terrorist act’ as contemplated under Section 15 of the UAP Act.

Significantly, therefore, the Court indicated that there existed a gradation of seriousness even within the UAPA, and that even on the basis of the prima facie case against him, there was no evidence that Anand Teltumbde had committed, or conspired to commit, a terrorist act. The NIA’s allegations of conspiracy, therefore, stood rejected.

Teltumbde, therefore, was granted bail.

Other than the issues around the jurisprudence of liberty that I have pointed out above, there are three further points that I want to make.

First, an important aspect of the High Court’s approach was set out in paragraph 11. While considering Section 43(D)(5) of the UAPA, the Court observed that “this is an extraordinary phenomenon and a deviation from the ordinary Criminal Law. Naturally, therefore the approach to such a case is required to be cautious and careful. By its very nature, the exercise to be undertaken by a Court in relation to this provision is therefore somewhat typical and delicate.” It is this approach that informed the Court’s careful and forensic consideration of the adequacy of the evidence in record, which followed. In this context, it is important to note that there exists another normative tension within UAPA jurisprudence: for one set of courts and judges, the fact that the UAPA exists to deal with serious crimes like terrorism implies that the judiciary should have a reflexive pro-State attitude in such cases, overlook the violation of procedural safeguards (recall the recent Sai Baba order of the Supreme Court) and essentially keep people in jail. For another set of courts and judges, however, the very stringency of the UAPA’s provisions requires courts to be even more vigilant, and consider the facts even more carefully before condemning individuals to years and decades in jail without trial. The Bombay High Court’s judgment belongs to the latter, pro-liberty category.

Secondly, while the High Court’s order is welcome and praiseworthy, it is impossible to ignore – as the Court itself notes – that at the time of the judgment, Anand Teltumbde had spent two and a half years in jail. This is a function of the increasing length of UAPA bail hearings, and indeed the reflexive instinct of trial courts (and in many cases, the High Court and Supreme Court) to refuse bail until a few years of incarceration have passed. Indeed, there is no reason why today’s High Court order could not have been passed by the trial court, at the first time of asking.

And finally, what of the other BK-16 cases? In paragraph 23, the Court attempted to distinguish this case from its denial of bail to Hany Babu and Jyoti Jagtap. While the Court made something of an effort for the former (pointing to various recoveries from Hany Babu), in Jyoti Jagtap’s case, it did not even do that: the Court simply said that “our judgment [in Jyoti Jagtap’s case] speaks for itself”, and that “we do not agree” that the two cases are identical. With the greatest of respect, this is something of a cop-out. The Jyoti Jagtap bench was the same bench as the present one (Gadkari and Jadhav JJ), and as I have written before, the Court’s entire consideration of the evidence in that kind evinced an approach that is the polar opposite from the one in this case: in denying bail to Jyoti Jagtap, the Court did fill in gaps in the Prosecution’s case with presumptions of conspiracy, and did interpret the UAPA’s provisions very broadly (see, for example, the Court’s analysis of criticisms of demonetisation and of the Prime Minister). I respectfully submit that these two judgments cannot stand together: while the Anand Teltumbdge’s judgment deserves appreciation and praise, Jyoti Jagtap’s judgment remains problematic, and it is to be hoped that it will be corrected on appeal.

At the time of writing, the Bombay High Court has stayed its judgment by one week to enable the NIA to appeal to the Supreme Court. We shall, therefore, soon know if Anand Teltumbde will walk free, or whether the Supreme Court will elect to stay this detailed and well-reasoned judgment, and keep him in jail.

Guest Post: Rekindling the Proportionality Test for Protective Discrimination under Article 15(3) [Part I]

[This is a guest post by Kartik Kalra. It is the first of a two-part series on the evaluation and application of a rekindled framework to test protective discrimination under Article 15(3).]

Article 15 consists of a set of prohibited grounds, limiting Parliament’s ability to treat classes differently based on religion, race, caste, sex and place of birth. It creates an exception for “women and children” within the Article, letting the state make sex-based classifications in the form of “special provisions.” The history of Article 15(3) is widely considered one of perpetuating sex-based stereotypes in the name of protection, with some authors locating its role in “justifying provisions which [are] downright derogatory”.

In this two-part series, I examine the history of case law on Article 15(3), propose a rekindled proportionality test to save discriminatory laws under that Article, and apply this framework to test the saving of pensionary benefits and their cessation on a widow’s remarriage. At times, striking down stereotype-perpetuating laws solely because of their stereotypical nature may produce socially sub-par outcomes, for some stereotypes are often true and laws are made in cognizance of the same. Section 125 of the CrPC is one such example, which creates a regime of maintenance only for women, children and parents, excluding men from its domain. The pensionary regime, which will be discussed in the following post, is another such example: the state recognizes that following the death of their husbands, women may be dependent, and makes provisions for pensions accordingly. Once a widow remarries, the state considers that they may no longer depend on it, for they would look to their new husband for their sustenance. This logic, while based on a quantifiably true understanding of the economic dependency of women over men, feels uncomfortable. A law of this kind, despite taking cognizance of true stereotypes, also risks actively perpetuating them. The effects of validating such a law would reaffirm the dependency stereotype, causing a disadvantage to women as a class. Simultaneously, women being the beneficiaries of maintenance or pensionary regimes, are also advantaged by such laws.

Navigating this line between recognizing and entrenching stereotypes is akin to walking on eggshells, for there exists no determining principle to save some stereotype-perpetuating laws within Article 15(3) and strike down others. In this piece, I attempt to develop such a determining principle in the form of a four-pronged test that is based on quantifiability, historical disadvantage, lesser-discriminatory means, and the balancing of advantages and disadvantages. In order to do so, I first assess Article 15(3) in light of quantifiably true stereotypes, followed by positing this new test and demonstrating its functionality.

Stereotype Perpetuation and Article 15(3)

The usual account of the evolution of Article 15(3) jurisprudence is as follows: there is one line of cases where the victimization of women occurs in the name of ostensible protection, to which cases such as Yusuf Abdul Aziz v. State of Bombay and Leela v. State of Kerala belong; and that there is a more recent line of cases that includes Anuj Garg v. Hotel Association and Joseph Shine v. Union of India that deny the saving of stereotype-perpetuating laws within Article 15(3). The former line is considered to lie in a paradigm of romantic paternalism that portrays women as the weaker sex that needs the continued protection of men. One author considers romantic paternalism an unwritten bargain where “economic support and protection given by the male [is exchanged] for subordination in all matters, sexual service and unpaid domestic work”. In exchange for the ostensible protection that the state gives women under Article 15(3), it presumes their subordination in the private realm. The latter line of cases does away with this paradigm, taking into consideration the historical disadvantages faced by women that presently necessitate special provisions, and limiting the protective ambit of Article 15(3) only to such provisions.

This separation does not, however, hold true for case law on quantitatively-verifiable stereotypes that also confer a discernible advantage on women. Many laws take into account a state of economic dependency that women have on men: compassionate appointments to government jobs, maintenance u/s 125 of the CrPC, sex-based reservations, and the pensionary regime are some examples. This can also be factually assessed: only a fifth of India’s female population is working, as opposed to nearly four-fifths of men. Laws recognizing this stereotype and making disparate treatment have been subject to multiple constitutional challenges. In State of Andhra Pradesh v. P.B. Vijayakumar, a rule preferring women over men in state employment was under challenge. Recognizing the historical disadvantages faced by women that necessitated a rule that officially prefers women over men, the Court acknowledged that special provisions are primarily intended to address the historic economic exclusion of women:

7. The insertion of clause (3) of Article 15 in relation to women is a recognition of the fact that for centuries, women of this country have been socially and economically handicapped. As a result, they are unable to participate in the socio-economic activities of the nation on a footing of equality. It is in order to eliminate this socio-economic backwardness of women and to empower them in a manner that would bring about effective equality between men and women that Article 15(3) is placed in Article 15. Its object is to strengthen and improve the status of women.

In Ramesh Chander v. Veena Kaushal, the constitutionality of Section 125 of the CrPC was challenged on the ground that it provides for maintenance only for women, excluding men. This was saved on the ground that it recognizes an existing reality and functions as a measure of social justice (¶9). Chaturbhuj v. Sitabai reaffirmed this purpose of the Section 125, pointing to the general state of economic dependency of all classes eligible for maintenance (¶5); Ramesh Rege v. Gauri Rege held that a law providing for maintenance for unmarried daughters is based on their economic dependence and confers an advantage on them (¶8); Cyril Britto v. Union of India held that a law prohibiting the arrest of women judgement-debtors would be saved due to their economic dependance, for the law serves as a “recognition of a reality” (¶8);  Shrikrishna Eknath Godbole v. Union of India reaffirmed the role of a law preventing the arrest of women judgement-debtors to concern their economic dependence (¶9,10); Dattatreya Motiram More v. State of Bombay upheld a law reserving seats for women by referring to the advantage being conferred on a historically excluded group:

7. The proper way to construe Article 15(3), in our opinion, is that whereas under Article 15(1) discrimination in favour of men only on the ground of sex is not permissible, by reason of Article 15(3) discrimination in favour of women is permissible, and when the Stale docs discriminate in favour of women, it does not offend against Article 15(1). Therefore, as a result of the [j]oint operation of Article 15(1) and Article 15(3) the State may discriminate in favour of women against men, but it may not – discriminate in favour of men against women.

In the above set of cases, we can observe the invocation of distinct determining principles to save laws under Article 15(3): Ramesh Chander relies on quantitative verifiability, P.B. Vijayakumar focusses on remedying historical disadvantage, and Dattatraya More relies on the conferral of an advantage on women in the form of reservations of seats. Another principle has been invoked in Anuj Garg:

50. “The test to review such a protective discrimination [under Article 15(3)] statute would entail a two-pronged scrutiny:

(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.

Further, consider the following remark in Vasantha v. Union of India, a case assessing the constitutionality of a rule prohibiting the employment of women in factories at night:

72. In the present case, the provision is not a protecting provision so that it could be tested with reference to Article 15(3) or Article 15(4) of the Constitution but it is a restriction and therefore, the validity of such restricting provision has to be tested in the anvil of Article 15(1) of the Constitution.

The Court, therefore, drew a line between protective and restrictive provisions, indicating that a law curtailing the rights of women cannot be saved under Article 15(3). This was also invoked by Chandrachud J. in his concurrence in Joseph Shine:

189. Article 15(3) encapsulates the notion of “protective discrimination”. The constitutional guarantee in Article 15(3) cannot be employed in a manner that entrenches paternalistic notions of “protection”. This latter view of protection only serves to place women in a cage. Article 15(3) does not exist in isolation…Neither Article 15(1), nor Article 15(3) allow discrimination against women. Discrimination which is grounded in paternalistic and patriarchal notions cannot claim the protection of Article 15(3).

It can, therefore, be observed that Courts have offered fragmented, and at times, intuition-based reasonings to save some laws under Article 15(3) but reject others. In the following section, I coalesce these distinct reasonings into a four-pronged test to determine whether a law can be saved by Article 15(3).

A Test to Evaluate Protective Discrimination

At this stage, it must be noted that all cases except Anuj Garg and Joseph Shine addressed laws based on stereotypes whose quantitative veracity was verifiable, which ultimately concerned the socioeconomic disadvantage and economic exclusion of women. Preventing their arrest due to judgement-debt, calling for special maintenance provisions, and the reservation of seats are all ultimately exercises in the recognition of stereotypes. Laws made in pursuance of the dependency stereotype, therefore, have been unwaveringly upheld.

Stage 1: Quantitative Verifiability

On this basis, I propose the first prong of the test, which would require the law to be made in pursuance of a stereotype that is quantitatively verifiable. If the law is based on a stereotype that isn’t true at all, no question of its saving under Article 15(3) arises. If the stereotype of lack of safety in establishments selling alcohol is false, then all matters end there, and the law cannot be saved within Article 15(3). Consider the use of this prong in Joseph Shine: since the law was based on the stereotype of a woman’s lack of sexual autonomy, which is undoubtedly false, the law cannot be saved as a “special provision” under Article 15(3).

Stage 2: Remedying Historical Disadvantage

The use of the second prong is essential to paint a true image of the reasons necessitating the use of Article 15(3). Its conceptual underpinnings are reflected adequately in the abovementioned extract from P.B. Vijayakumar, which pertains to the structural economic exclusion of women that causes the present status of economic inequality, necessitating remedial measures. Article 15(3) has the same conceptual underpinnings as that of reservations under Articles 15(4) and 16(4), well-summarized in Subba Rao J.’s dissent in T. Devadasan v. Union of India, approved in all cases thereafter:

26. Centuries of calculated oppression and habitual submission reduced a considerable section of our community to a life of serfdom. It would be well nigh impossible to raise their standards if the doctrine of equal opportunity was strictly enforced in their case. They would not have any chance if they were made to enter the open field of competition without adventitious aids till such time when they could stand on their own legs.

The use of Article 15(3), therefore, must be cognizant of the reasons that necessitate it. Like reservations under Articles 15/16(4), special provisions are created in a context where the historic political and economic exclusion of women has continuing implications. Arguments of petitioners in Cyril Britto or Shrikrishna Godbole, which point to the unfairness in creating criminal immunity for women in offences that men are imprisoned for, miss the larger picture of historical disadvantage that necessitates such laws. A stereotype-perpetuating law, therefore, must have its roots in remedying historical disadvantages.

Stage 3: Availability of Lesser Discriminatory Alternatives

The third prong is an exercise in finding lesser-discriminatory means that can reach the same end that the law seeks to achieve. Essentially, this prong is a test of over-inclusiveness: is there a possibility of the creation of a narrower class to whom the impugned stereotype-perpetuating law can apply, which can avoid the vice of essentialism? The test of over-inclusiveness has been used in equality jurisprudence in, inter alia, Indian Hotel and Restaurant Association v. Union of India, which held the prohibition of any dance performance in some establishments to be bad due to the simultaneous prohibition of both sexually inappropriate and ordinary dances, making it over-inclusive (¶20); Navtej Singh Johar v. Union of India, which held the inclusion of both consensual and non-consensual sexual intercourse u/s 377 to be over-inclusive (¶221); and Inspector Ravina v. Union of India, which held the inclusion of all forms of unwillingness, that which was the result of the exercise of bodily autonomy and that which concerned one’s true unwillingness as a ground to disqualify one from promotion to be bad for over-inclusiveness (¶12).

The state, therefore, would have the obligation to prove that there is no narrower class to whom the stereotype-perpetuating law can apply, which would prevent over-inclusiveness and the portrayal of an entire sex-based class a certain way.

Stage 4: Balancing Between Benefits and Burdens

It is evident from case law that the conferral of a benefit on women is necessary for the law to be saved by Article 15(3), and a restriction cannot masquerade as a protection. Malhotra J.’s concurrence in Joseph Shine also relied on the following reasoning:

274. The true purpose of affirmative action is to uplift women and empower them in socio-economic spheres. A legislation which takes away the rights of women to prosecute cannot be termed as “beneficial legislation”.

There is, at the same time, harm that could result from the saving of laws under Article 15(3), which pertains to the perpetuation of stereotypes. Given the balancing between the conferral of benefits on women and the harm resulting from the perpetuation of stereotypes, this prong may be considered akin to the last prong of the traditional proportionality test. The state, therefore, would have the burden to prove that the benefits being derived from via the protective measures are greater than the harm caused due to the perpetuation of stereotypes.

This stage, like its traditional proportionality counterpart, is also value-based and requires a subjective assessment of the value accorded to particular principles. Thus, even though the exclusion of men from the domain of Section 125 of the CrPC has the effect of perpetuating a stereotype of women as dependent, a judge may come to a conclusion that the benefits being conferred on women via this law outweigh the disadvantages caused by the perpetuation of stereotypes. The treatment of a law of this kind under the third prong, however, is a different issue.

Navigating the Thin Line Between the Recognition and Entrenchment of Stereotypes

The simultaneous evaluation of these four prongs is necessary in order to navigate the thin line between the recognition of a stereotype and its active entrenchment. While Anuj Garg posits a two-pronged test, its context pertains to stereotypes that are neither quantitatively true (lack of safety for women in establishments selling alcohol), nor have the effect of conferring an advantage on women, for the law caused their economic exclusion. For laws that concern a quantitatively verifiable stereotype, an immediate declaration of unconstitutionality may cause a devastating levelling down: the state can argue that there is no scope for maintenance provisions for women if economic independence is the constitutionally mandated starting point. This was the disagreement between O’Regan and Kriegler JJ. In the South African Constitutional Court’s judgment in President of South Africa v. Hugo, a case concerning a remission of prison sentences of mothers with children younger than twelve years. In a challenge to the exclusion of men constituting the perpetuation of sex-based discrimination due to the assignment of child-care duties to women, O’Regan J. made the following remark:

112. To determine whether the discrimination is unfair it is necessary to recognise that although the long-term goal of our constitutional order is equal treatment, insisting upon equal treatment in circumstances of established inequality may well result in the entrenchment of that inequality.

O’Regan J. considered that the denial of the present stereotypical reality of women engaging in child-rearing, while desirable in the long run, would produce an undesirable outcome of their continued incarceration at present. While Kriegler J. ultimately concurred with the majority on the validity of the remission, he considered its underlying reasoning in women’s status as caregivers of children to be constitutionally unsound:

80. In my view the notion relied upon by the President, namely that women are to be regarded as the primary care givers of young children, is a root cause of women’s inequality in our society. It is both a result and a cause of prejudice; a societal attitude which relegates women to a subservient, occupationally inferior yet unceasingly onerous role. It is a relic and a feature of the patriarchy which the Constitution so vehemently condemns.

The navigation of the issue of quantifiably verifiable stereotypes must, therefore, occur with great caution. While the production of socially sub-par outcomes due to an unrealistic commitment to constitutional ideals is undesirable, so is the indefinite perpetuation of sex-based stereotypes. In order to navigate this, the four-pronged test would perform the following tasks: the first prong of quantitative verifiability will ensure that only those rules concerning true stereotypes can be saved by Article 15(3), with the perpetuation of false stereotypes being entirely excluded from that Article’s domain; the second prong will ensure that there is a nexus between the allegedly protective provision and its historical underpinnings, ensuring that a true image of the need of such a law is painted; the third prong ensures that wherever a lesser-discriminatory alternative in the form of a narrower class is available, it is employed instead of the impugned law; and the last prong of the weighing of advantages and disadvantages creates an additional possibility of saving stereotype-perpetuating laws where the Court assesses the benefits being conferred on women to outweigh the disadvantages via the perpetuation of stereotypes.

On this basis, the rekindled test to save stereotype-perpetuating laws within Article 15(3) comprises of the following prongs:

  1. the law must be based on stereotypes that are quantitatively true and verifiable;
  2. the stereotype must be perpetuated in pursuance of remedying historical disadvantages faced by women;
  3. there must be no availability of a lesser-discriminatory alternative than the impugned law; and
  4. the advantage conferred on the historically disadvantaged group (i.e., women) must be greater than the disadvantage generated via the perpetuation of stereotypes.

As I have demonstrated, the present regime on saving laws within Article 15(3) functions in an ad-hoc manner, offering very little in definite determining principles to save some laws but reject others. Even Joseph Shine, which is widely lauded for affirming the true role of Article 15(3), doesn’t rely on a strict principle to save some laws but reject others, relying only on the notion that laws “entrench[ing] paternalistic notions of protection” cannot be saved by the Article. For all practical purposes, this doesn’t leave us in a position better than Cyril Britto and Ramesh Chander, both of which posit that laws made in the “recognition of a reality” can be saved under Article 15(3). The navigation of the line between recognizing and entrenching stereotypes, I submit, can best be done via this four-pronged test.

Guest Post: Schrodinger’s Substantive Equality – Conceptual Confusions and Convenient Choices in Justice Maheshwari’s Plurality Opinion in the EWS Case

[This is a guest post by Ayan Gupta.]

The Supreme Court (SC) in its split verdict in Janhit Abhiyan v Union of India upheld the 103rd Amendment to the Indian Constitution. The amendment permits states to introduce reservation up to 10%, on top of already existing reservations, for “Economically Weaker Section [EWS]” of the citizenry. The qualifications for being classified as EWS are notified by the Govt separately and currently include persons with a family income of less than 8 Lakhs per annum.

At almost 400 pages, the judgement comprises of four opinions. The plurality opinion of Maheshwari J., along with concurrences from Trivedi and Pardiwala JJ., find consistent with the basic structure of the Constitution (a) that reservations can be grounded in solely economic basis and (b) that such reservation can be introduced to the exclusion of Scheduled Castes, Scheduled Tribes, and Other Backward Classes. The dissent by Bhat J. (joined by Lalit CJ.) disagrees explicitly on the second point regarding the exclusion of SC, ST, and OBC beneficiaries.

Keiran Correia has already provided an insightful analysis on the dissenting opinion of Bhat J, while Kartik Kalra has likewise examined the majority judgments. In this post, I focus specifically on Maheshwari J.’s plurality opinion. I argue that the decision highlights a severe lack of conceptual clarity concerning the nature and role of reservations within the Indian Constitution. 

Clarifying Indra Sawhney: The Task at Hand

The reservation jurisprudence of the SC has always been puzzling, with the Court often moving back and forth between the choices it makes. Nonetheless, since State of Kerala v NM Thomas, it is generally accepted, at least doctrinally, if not in practice, that article(s) 15(4) and 16(4) are not exceptions to the Constitution’s equality code, but emphatic restatements of it; in other words, the Equality Code embodies a vision of substantive equality, of which Article 16(1) (for example) is the general statement, and Article 16(4) a specific application.

That is not to say that this notion has not had its fair share of resistance. In Indra Sawhney, a 9-Judge bench of the SC, despite affirming NM Thomas, upheld MR Balaji’s 50% ceiling on reservations. This position reflects a “normative tension” in Sawhney, where on the one hand, the Court commits to understanding reservations as a facet of equality, and on the other, holds that the equality of opportunity function rationale of reservation must be balanced against the right to equality of everyone else.

Affirmative Action as a Facet of (In)Equality

Indeed, this tension is the exact conflict that Janhit Abhiyan required the Court to resolve. In having to decide whether SC, ST and OBCs can be excluded from the scope of EWS reservations, the Court would have had to deal with the internal tension that Sawhney presented before it. It would have had to reason whether exclusion of backward classes within a substantive equality framework could be balanced against the equality claims of “everyone else.”

Unfortunately, as we see in the plurality opinion, the Court undertakes no such exercise. Instead, it begins with the presumption that reservations necessarily do function as an exception to the equality. Thus, after articulating the issues it is to decide, in para 31.1 the Court says:

31.1. All these points are essentially structured on three important components namely, (i) the general rule of equality enshrined in Article 14 of the Constitution; (ii) the reservations enabled in Articles 15 and 16 as exception to the general rule of equality; and (iii) the doctrine of basic structure that defines and limits the power of the Parliament to amend the Constitution.

However, despite this presumption, the Court does not reject substantive equality-oriented interpretation of the Equality Code. Though in para 44, the plurality articulates the equality function of the constitution in the classic form of “equals must be treated equally while unequals need to be treated differently,” it follows up in the very next paragraph by noting that in Indian Constitutional Jurisprudence has “…held the guarantee of equality to be substantive and not a mere formalistic requirement” [para 46].

The plurality’s articulation of the Constitution’s equality, in these two paragraphs, highlights the first signs of the conceptual confusion that plagues the reasoning that follows. In the very next section, titled “Affirmative Action by ‘Reservation’: Exception to the General Rule of Equality,” this confusion becomes glaringly visible.

The Court begins by reasoning that the state is “tasked with affirmative action” as India’s “multifaceted social structure, ensuring substantive and real equality, perforce, calls for consistent efforts to remove inequalities” [para 48]. It goes on to characterize reservations as the “basic gateway to tread the path of all-around development” [para 52.1]. Yet, it concludes the section by stating that

56. However, it need be noticed that reservation, one of the permissible affirmative actions enabled by the Constitution of India, is nevertheless an exception to the general rule of equality and hence, cannot be regarded as such an essential feature of the Constitution that cannot be modulated…

Interestingly, the Court does cite NM Thomas to say that Article 14 embodies the ”the generic principle of equality (as genus) and Articles 15 and 16, [enact] the facets of general equality (as species)” [para 52]. Yet, it does not engage with, or even acknowledge, the impact of this statement (and the Thomas decision) on its presumption that reservations represent an exception to equality since Thomas had explicitly held, and Indra Sawhney had explicitly reaffirmed, that reservations do not function as exceptions to equality!

The rest of the opinion (including this particular section) offers no reasons for why reservations have been characterized as an exception to the “general rule of equality.” It is important to note that the Court does not engage at all with the decisions in NM Thomas and Indra Sawhney in so far as they question the nature of reservations within the Constitution’s equality code.

It begs to be asked, then, if the Court is clear in its determination that reservations function as an exception to the general rule of equality, why does it articulate the equality code’s function as a substantive equality feature? Why does it characterise reservations as exceptional and yet justify their presence on a substantive equality principle?

The answer to this question perhaps lies in the plurality’s need to justify reservations based solely on economic determinant(s), a classification that did not exist prior to the amendment, and at the same time find reasons for the exclusion of SC, ST and OBCs from the scope of the new amendment.

In having to justify reservations to the exclusion of caste and group identities, the plurality takes resort to the space provided by substantive equality formulations of the Constitution. For instance, in para 63, it notes that

63. … this Court distinctly pointed out that the equality clause in the Constitution does not speak of mere formal equality but embodies the concept of real and substantive equality, which strikes at inequalities arising on account of vast social and economic differentials.

Indeed, the Court follows up by noting that in conversations concerning substantive equality, economic justice has acquired “equal focus” alongside “the principles of social justice” [para 65]. In articulating this substantive equality argument, it goes so far as to say

65. …[a]ny civilized jurisdiction differentiates between haves and have-nots, in several walks of life and more particularly, for the purpose of differential treatment by way of affirmative action.

Thus, on the one hand, the Court refuses to acknowledge that substantive equality articulations do not understand affirmative action policies as exceptions to equality. On the other hand, it uses the language of substantive equality to reason that the Constitution not only permits, but demands [para 79], that economic poverty be recognized as a group/ class of its own.

It is important to note, in this context, that the Court’s decision to allow the breach of the 50% ceiling for EWS reservations must necessarily result from a substantive equality conception of the Constitution. The introduction of the limit was based on the need to “balance” equality of opportunity and the right to equality of “everyone else.” It functions on the assumption that reservations are an exception to the “general principle” of equality, and therefore, cannot exceed 50% as exceptions cannot swallow the whole.

Thus, though the stated reasoning for permitting a breach of the limit is that it concerns Article(s) 15(4) and 16(4) and is irrelevant for EWS reservations, which form a separate category of its own [para(s) 93 and 94], any breach of the 50% limit would necessarily require substantive equality as an underlying philosophical justification. Otherwise, as is the case presently, the Court is faced with a situation where the balance is tilted in favor of one group against another, and the general rule of equality would be defeated by more than 50% reservations. This would be a result inconsistent with the vision of formal equality. Once again, this is a tension that remains unaddressed in the plurality decision.

Conclusion: A Constitutionalism of Convenience

These two contradictory choices represent, to me, a constitutionalism of convenience. In creating this conceptual confusion, the plurality made space for envisioning reservations for a group otherwise not entitled to it without having to include already existing beneficiaries of affirmative action. In essence, on both questions, first, concerning validity of economic reservations and, second, concerning exclusion of SC, ST, OBCs from such reservation, the Court used these competing visions of equality to justify the conclusion it chose.

In answering the first question, the Court held that since reservations function as an exception to equality (the formal vision), it cannot be “…regarded as such an essential feature of the Constitution that cannot be modulated” [para 56]. In answering the second question, using substantive equality’s commitment to acknowledging and countering structural injustice (substantive vision), it held that that ”..but for this exclusion [of SC, ST, and OBCs], the purported affirmative action for a particular class or group would be congenitally deformative” and therefore, the exclusion would not violate the basic structure of the Constitution [para 79].

In making these choices, the Court refuses to acknowledge the contradictions inherent in its approach. The approach is intellectually suspect, and adds to the conceptual confusions in a jurisprudence already plagued by lack of clarity across its various aspects.

Indeed, even the Court’s use of substantive equality as a bridge for creating exclusions without considering the interaction between caste and poverty only serves to highlight its evasion in refusing to engage with the complete consequences of the concepts and choices it employs. Finally, by failing to engage with the normative tensions in Indra Sawhney, the Court refuses to perform the job it was tasked with of clarifying the nature and place of reservations within the equality code.

On a parting note, I also wish to highlight an interesting observation by the Court that deserves more exploration. In para 97, the Court notes that the extent of reservations made under the EWS quota may be “determined with reference to the relevant analysis of the material data justifying a particular percentage.” In doing so, it brings forth the ghost of Nagaraj. In Nagaraj, the SC held that the State must submit quantitative data on backwardness of the beneficiaries and impact on administrative efficiency for granting reservations under Article 16(4A).

In doing so, it made granting such reservations difficult especially as it provided little to no guidance as to what such data must indicate. Though the EWS reservations do not have backwardness and efficiency related qualifications, it will be interesting to see the challenges that follow based on lack of or a need for data justifying the reservations granted (and the many methodological confusions that categorizing “EWS” would bring). Would this lead the Court away from the approach in data-focused approach in Nagraj or would it only solidify Nagaraj’s presence remains to be seen.