The ConCast: Episode 3, Part 2 (31 July 2022)

In Part 2 of Episode 3 of The ConCast, Abhinav Sekhri and I continue our discussion: we talk about Mohammed Zubair’s arrest, and what the events between his incarceration and his release tell us about how the Constitution protects – or fails to protect – the individual from the State.

The Podcast is available to listen below, and also on Podbean Google Podcasts, Apple Podcasts, and Spotify.

Selected References:

State vs Anil Sharma.

P. Chidambaram vs Directorate of Enforcement.

State of UP vs Mohammed Zubair (Sitapur bail order).

Coverage of remand hearing before the Mgaistrate’s Court, Patiala House.

Coverage of the bail hearing before the Magistrate’s Court, Patiala House.

Mohammed Zubair vs State of UP (interim bail order).

Mohammed Zubair vs State of UP (extension of interim bail).

Mohammed Zubair vs State of UP (no precipitate action order).

Mohammed Zubair vs State of UP (Supreme Court bail order).

The ConCast: Episode 3, Part I (30 July 2022)

The ConCast is back! In this Episode, Abhinav Sekhri and I look at the Supreme Court’s PMLA judgment, the interface between criminal law and the Constitution, and we talk about why criminal lawyers are a bit like Leonard Cohen.

The Podcast is available to listen below, and also on Podbean, Google Podcasts, Apple Podcasts, and Spotify.

Select References:

Prevention of Money Laundering Act, 2002 [“PMLA”].

Vijay Madanlal Chaudhary vs Union of India [“PMLA Judgment”].

Abhinav Sekhri, “Of Old Wines in New Bottles: The Judgment in Vijay Madanlal Choudhary – Part One.”

Abhinav Sekhri, “Old Wines in New Bottles?: The Judgment in Vijay Madanlal Choudhary – Part Two.”


Peter Alldridge, What Went Wrong With Money Laundering Law? (Palgrave 2015).

Peter Alldridge, Money Laundering Law (Hart 2003).

United Nations Convention Against Corruption.

United Nations Convention Against Transnational Organised Crime.

Financial Action Task Force (FATF) Report on India, 2010.

FATF Follow-Up Report on India, 2013.


Forfeiture Act of 1859.

Narcotic Drugs and Psychotropic Substances Act, 1985 [“NDPS”].


Tofan Singh vs State of Tamil Nadu.

Nikesh Tarachand Shah vs Union of India.

Ramanlal Bhogilal Shah vs D.K. Guha.

State of Bombay vs Kathi Kalu Oghad.

Satender Kumar Antil vs CBI


Abhinav Sekhri, “Confessions, Police Officers, and S. 25 of the Indian Evidence Act, 1872.”

Abhinav Sekhri, “Not so Civil: The Money Laundering Act and Article 20.”

The Executive(’s) Court: Notes on the Legacy of Justice A.M. Khanwilkar

This blog has a long-standing tradition of assessing the judicial legacies of Chief Justices of India, upon their retirement (see here, here, and here). This tradition has hitherto been limited to Chief Justices, because of the sway that they exercise upon the Supreme Court as “master(s) of the roster”, and because during their tenures, they tend to hear significant constitutional cases themselves.

Last year, an exception was made upon the retirement of Justice R.F. Nariman, for reasons explained in this blog post. Today, the retirement of Justice A.M. Khanwilkar requires, I believe, a second exception. One reason for this is that during the course of his career (as we shall see in this post), Justice Khanwilkar has written some of the most consequential judgments concerning State power and the rights of the individual. But secondly – and more importantly – when you study these judgments together, you glimpse a certain judicial philosophy – such as it is – at work. This judicial philosophy – subject to a few important exceptions – is, I believe, largely representative of the Supreme Court today (which also perhaps explains why, across Chief Justices, these kinds of cases have been regularly assigned to Justice Khanwilkar, one of its most forceful proponents).

What is this philosophy? In my earlier analysis of Justice Khanwilkar’s judgment in the FCRA Case (also discussed below) I had compared it to the Peruvian President Óscar R. Benavides famous line, “for my friends, anything; for my enemies, the law.” In a similar vein, the common thread running through Justice Khanwilkar’s constitutional law judgments is: “for the State, anything; for the individual, the law“: it is the philosophy not just of the executive court, but of the executive(‘s) court.

Before we begin, a final point, by way of caveat: it is almost trite to say that I do not agree with the outcomes of the cases that I discuss below. I have criticised some of these judgments when they were delivered, and in the Central Vista Case (that I flag, but do not discuss), I was one of (many) arguing counsel on the losing side. My analysis below, however, is not founded simply upon the fact of disagreement with the outcome, or of dislike of these judgments. Regardless of my predilections, I believe that these judgments reveal something important, both about Justice Khanwilkar’s judicial career, and about the contemporary Supreme Court, which is important to articulate and to discuss. This post should be read in that spirit.

Watali: Taking a Sledgehammer to Personal Liberty

Any discussion of Justice Khanwilkar’s legacy must begin with the 2019 judgment in National Investigation Agency vs Zahoor Ahmad Shah Watali. The case involved the interpretation of Section 43(D)(5) of the Unlawful Activities Prevention Act [“UAPA”], India’s umbrella anti-terrorism statute. Section 43(D)(5) prohibits a Court from granting bail to an accused if “on a perusal of the case diary or the report made under Section 173 of the [Criminal Procedure] Code, [the Court] is of the opinion that there are reasonable grounds for believing that the accusation against such person is prima facie true.” In layperson’s language, Section 43(D)(5) bars the grant of bail if it appears that the police version (through the case diary or the chargesheet) against the accused is, on the face of it, true.

Watali was an appeal by the National Investigation Agency [“NIA”] against an order of the Delhi High Court. In that order, the Delhi HC had granted bail to Watali (the accused), under Section 43(D)(5) of the UAPA. The High Court took into account the (uncontroversial) legal proposition that “as far as the statutes concerning serious offences inviting grave consequences are invoked, the trial Court will scrutinize the material with extra care.” The Court’s job was not to proceed simply on the basis of the statements made by the investigative agency, and nor to act as a “post-office” for the State. On this basis, the High Court subjected the police version – according to which Watali was involved in terror funding – to rigorous scrutiny. It found that many of the witness statements were inadmissible under the law of evidence, that the documents purporting to originate from the accused were neither signed by him and nor on his letterhead, and that other documents were entirely innocuous, and consistent with his position as a prominent Kashmiri businessman. On this basis, the Court found that at that point, the police version was speculative, and there was no ground for denying bail to the accused.

When the case came up in appeal, the Supreme Court – in a judgment authored by Justice Khanwilkar – overturned the High Court’s order, and put Watali back in jail (he stayed in jail – awaiting trial – for three more years, until in February 2022, he was moved to house arrest because of a terminal disease). Crucially, Khanwilkar J’s problem with the High Court was not that it had incorrectly appreciated the facts of the case. Rather, it was that the High Court had applied the wrong legal standard altogether, and that the true role of the Court under S. 43(D)(5) of the UAPA was, effectively, to act like a post office. He noted that while examining the question of bail, “elaborate examination or dissection of the evidence is not required to be done”, and that furthermore, to reject inadmissible statements at the stage of bail was akin to entering into the “merits and demerits of the case.” Instead, the Court was to form a view based on the “broad probabilities” flowing from all the materials supplied by the police.

The judgment in Watali was criticised at the time as being incorrect (see, e.g., Abhinav Sekhri’s blog post), and I do not intend to traverse covered ground once again. It is worthwhile, however, to recall once again just what it did. As is well known, at the time of bail, the defence cannot present its own arguments, put forward its own witnesses, or cross-examine the prosecution’s witnesses. It has no real way to effectively contest the State’s case. All that is for the stage of trial. At the time of bail, all the Court can look at – and all that the defence can point to – is the State’s version of events. Thus, when Section 43(D)(5) prohibits the Court from granting bail if “there are reasonable grounds for believing … that the accusation is prima facie true”, everything turns upon how closely and deeply the Court is authorised to examine the State’s version, on its own terms – for internal consistency, for plausibility, for whether the State is relying on materials that would even be admissible at trial (such as hearsay statements) – to come to its prima facie conclusion. And when, in Watali, Khanwilkar J barred all Courts from “examining” or “dissecting” the evidence, he effectively made the grant of bail in UAPA cases borderline impossible. As Sekhri wrote at the time, he “actively chose a legal position that makes lengthy undertrial detention more likely.”

The asymmetry in power is glaring. UAPA trials in India take years – decades – to complete. If the grant of bail is made borderline impossible, then all the police are required to do is to slap the UAPA onto a chargesheet, and an individual will be condemned to years – or decades – in jail without trial. The chargesheet and the materials need not be persuasive, need not be internally coherent, and in addition to all this, may even rely on plainly inadmissible material (as in Umar Khalid’s case): all that ceased to matter once, in Watali, Khanwilkar J turned all courts into stenographers for the Prosecution, while attaching dumbbells to the feet of the Defence and throwing it into the river to swim or sink. In this sense, Sekhri’s 2019 warning has turned out to be prescient: “it is hard to conceive of outcomes which are anything but fearsome. The decision could make the UAPA an even more attractive tool to law enforcement agencies now that getting bail is harder…” We now know that this is exactly what has happened: the UAPA is the foremost tool of political repression in India, and Watali has become the chant that almost all Courts (barring a few) invoke to justify keeping people in jail for years without trial.

PMLA: Taking another Sledgehammer to Personal Liberty

If the UAPA is the executive’s weapon of choice to keep inconvenient individuals in jail for years without trial, the Prevention of Money Laundering Act [“the PMLA”] is its political weapon. By now, every Indian knows about the “Enforcement Directorate” – or, as it is commonly known by its abbreviation – the “ED”. The eyeball impression that the ED is used to overwhelmingly to jail political opponents without trial, has been confirmed in this detailed analysis; that the purpose is jail without trial is borne out by the fact that while the number of PMLA cases filed by the ED has risen by eight times over the last eight years, the conviction rate under the law is under 1% – a statistic that should send alarm bell ringings for everyone (other than, it seems, the Supreme Court).

Amendments to the PMLA – passed in 2019 – which made the legal regime more draconian, were challenged before the Supreme Court. On 27 July – two days before his retirement – a three-judge bench led by Khanwilkar J delivered judgment, upholding all the provisions under challenge (see here, here and here).

A similarly exhaustive analysis of the judgment is not the subject of this post. However, its underlying philosophy is simple enough: while in every sense the officials of the ED act like the police – as coercive appendages of the State, and in the power that they hold over citizens – the Court liberated them from following the minimal procedural constraints under the Code of Criminal Procedure that do apply to the police. For instance, the Court exempted the ED from sharing the equivalent of the police’s First Information Report – the “ECIR” – with the accused, noting that communicating the “grounds” was enough; the Court held that as an ED summons was not an “arrest” (even though functionally indistinguishable from it), the constitutional right against self-incrimination doesn’t apply to statements made under ED questioning; that because ED officials weren’t “police officers” (even though functionally indistinguishable from them), confessions made to them were admissible in evidence (even though the whole purpose of making confessions to the police inadmissible was the fear of coercion); and that because the ED wasn’t a police force (even though functionally indistinguishable from one), the procedures that it followed (the “ED manual”) wasn’t required to be made public, but could remain an “internal document.” If all of this sounds somewhat reminiscent of the Stasi, it is because it is rather reminiscent of the Stasi (or, in Pratap Bhanu Mehta’s words, “Kafka’s Law“).

The effect of the judgment is clear: it is the sanctioning of a State-controlled, coercive militia, exempt from the basic principles of due process and the rule of law. To this heady cocktail, the Court added further, dangerous mixes: it upheld a bail requirement even harsher than section 43(D)(5) of the UAPA, and which the Supreme Court had itself struck down four years before (Khanwilkar J overruled precedent, simply to ensure that bail would become almost impossible under the PMLA), and upheld the “reverse burden” clause – i.e., that under the PMLA, the burden was on the individual to prove their innocence, and not on the State to prove guilt.

And finally, to expand the scope of the PMLA, Khanwilkar J went further: Section 3 of the Act stipulates that “whosoever directly or indirectly attempts to indulge or knowingly assists or knowingly is a party or is actually involved in any process or activity connected with the proceeds of crime including its concealment, possession, acquisition or use and projecting or claiming it as untainted property shall be guilty of offence of money-laundering.” In other words, for the PMLA to be attracted, two conditions had to be satisfied: involvement (whether intentional or unintentional) in connection with proceeds of crime, and the (definitely) intentional “projecting” or “claiming” it as untainted property. Khanwilkar J held, however, that actually, the word “and” meant “or” (just like “day” means “night”), and that therefore, simply being in possession of “tainted” property was enough for guilt under the PMLA.

When you now combine this with the reverse burden clause (that under the PMLA, the individual is guilty until proven innocent), and Khanwilkar J.’s finding that any criminal offence could be brought under the PMLA (thus effectively making the CrPC wholly redundant), the effects of this judicial rewriting exercise are terrifying. They also exacerbate and worsen the already wide definition of tainted property under the PMLA, which effectively covers just about everything (and makes just about everything subject to attachment orders (see here), financially crippling someone under PMLA scrutiny; note that Khanwilkar J also held that property can be attached right from the beginning of PMLA proceedings).

But there are three things really of importance here. The first is that the re-worded section makes no grammatical sense (try reading it aloud and see for yourself). The second is that this interpretation turns basic criminal law principles on its head: because criminal legal statutes are coercive, and impose jail time on people, there is a time-honoured, well-worn principle in criminal law that they are to be read strictly and narrowly. In Khanwilkar J’s judicial philosophy of “for the State, everything; for individuals, the law”, however, every canon of interpretation is upside down, and nobody is safe from arbitrary State action; and finally, of course, to accomplish this task, he had to rewrite the section, taking the word that existed and replacing it with its opposite. I have previously referred to this as “Humpty Dumpty jurisprudence“, where the Court – like Humpty Dumpty in Alice Through The Looking Glass – decides that words mean what it decides them to mean, just because it can:

“When I use a word,” Humpty Dumpty said, in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”

“The question is,” said Alice, “whether you can make words mean so many different things.”

“The question is,” said Humpty Dumpty, “which is to be master—that’s all.”

In his analysis of the judgment, Abhinav Sekhri points out that there was material on record to show that while drafting Section 3, the legislature had made a genuine error, and used the word “and” while it meant to use the word “or”. However, when it comes to criminal law, it is most certainly not the Court’s job to save the legislature from the consequences of its own incompetence (especially when the same leniency is hardly accorded to the individual!): the whole point of the doctrine of reading criminal statutes literally, narrowly, and strictly is that, given the differences in power between the State and the individual, the reach of the criminal law is not to be expanded any further than what the words can bear. It is that principle that is Khanwilkar J entirely forsook in rewriting Section 3.

Let us take a step back, and sum up. When we look at the judgment in a broader context, it is important to keep in mind Sekhri’s observation that not all of this is entirely new. In many respects, the PMLA judgment is a continuation of the Indian Supreme Court’s long-standing tradition of expanding the State’s coercive powers and erasing the procedural safeguards that the law extends to individuals. In the PMLA context, however, the statute’s provisions magnify that substantially: the statute “weaves together all the restrictive, rights-effacing clauses from this illustrious past in one fine blanket, and it then goes further.” And the PMLA judgment, in turn, is perhaps unique in that it brings all of those rights-effacing judicial predilections together, in one case – what Sekhri calls a “greatest hits” video, and to which we can add: the band is the Supreme Court and the “hits” are direct hits to our constitutional rights: in sum, Khanwilkar J rewrote a criminal statute to substantially widen its ambit; authorised the State to bring any offence within that ambit; upheld the reverse burden of proof within that widened ambit; deprived individuals of their procedural and constitutional rights within that widened ambit; made the grant of bail almost impossible within that widened ambit; and exempted the State authorities from any effective constraints, once they began to operate within that widened ambit. When you put all of these together, what emerges is the classic definition of a lawless law, blessed by the executive’s Court.

Noel Harper: Taking a Hatchet to the Freedom of Association

In April 2022, Khanwilkar J wrote a judgment upholding various amendments to the Foreign Contributions (Regulation) Act of 2022. Elsewhere, I have analysed this judgment at some length, and pointed out how the Court accorded its imprimatur to a set of provisions that had turned India’s NGO regulation law into a Russian-style legislation that effectively made the work of most NGOs either impossible, or prohibitively difficult. A few salient points stand out from this judgment.

First, at the time of hearing Noel Harper, there were challenges to the FCRA pending in High Courts. Noel Harper itself was a limited challenge to one set of restrictions. Now, ordinarily, the Supreme Court is quick to talk about how the High Courts should not be bypassed; however, it seems that all that rhetoric ceases to matter when legislation that the political executive really cares about is at stake. Here, the Khanwilkar J-led bench could not wait to bypass those same High Courts, and hear and decide all questions about the constitutional validity of the FCRA, thus effectively depriving the High Courts from hearing the cases before them.

Secondly, the judgment in Noel Harper applied differential standards to the State and to the petitioners, where the State’s factual claims (contrary to the prevailing legal standard of proportionality) were taken as true without any scrutiny, whereas the petitioners’ claims – and bona fides – were taken with the highest level of mistrust. In my post analysing the judgment, I wrote that:

The Court begins by framing the issue in a way that is most favourable for the State, and least favourable for the citizen. Having framed the question thus, it then goes on to accept the State’s factual claims at face value, but does not extend the same courtesy to the citizen. Having done that, it then applies those parts of existing legal doctrine that favour the State, and ignores – or misrepresents – those parts that protect the rights of citizens. Having framed the question in favour of the State, accepted the State’s version of reality, and applied the doctrine in favour of the State – voila! – the conclusion is that the challenged State action emerges validated from the tender caresses of judicial review.

Indeed, this is a thread that runs throughout Khanwilkar J.’s judgments, and for a more elaborate articulation in this case, interested readers may consult the above blog post as a whole.

However, the most glaring aspect of Khanwilkar J.’s judgment (other than its impact on the freedom of association) – is that he explicitly and unashamedly framed its arguments in ideological terms, and this ideology was evidently the ideology of the political executive. Lines from the judgment include: The question to be asked is: “in normal times”, why developing or developed countries would need foreign contribution to cater to their own needs and aspirations?“; “Indisputably, the aspirations of any country cannot be fulfilled on the hope (basis) of foreign donation, but by firm and resolute approach of its own citizens“; “There is no dearth of donors within our country.”

These are familiar lines. These are lines that we hear from the mouths of authoritarian leaders across the world, when they justify clamping down on civil society, and in particular, on NGOs. None of these words have anything to do with the law, legal reasoning, the Constitution, and the practice of constitutional adjudication. Yet here they are, serving as the articulated major premise of a Constitutional Court judgment that is supposedly about whether restrictions upon the freedom of association – achieved via choking off funds to NGOs – are reasonable or not. But as we have seen, that is not really what this judgment is about: what this judgment is really about is giving formal judicial imprimatur to some of the more extreme and prejudicial rhetoric of the political executive, giving a dressing down to citizens who have the temerity to want to raise funds for NGO work, and telling them to be “resolute and firm” if they want to have rights. This is the language not just of the executive court, but of the executive(‘s) court.

Teesta Setalvad and Himanshu Kumar: Taking a Dagger to Article 32

The language of the executive’s court is present most starkly in Justice Khanwilkar’s notorious opinion in the Zakia Jafri case. Once again, it is not my task here to examine the correctness of the judgment in refusing to set aside the SIT Report that had found that there was no controversy at high governmental levels during the horrendous 2002 Gujarat Riots (interested readers may refer to Nizam Pasha’s analysis of the judgment, here; see also the discussion in Episode 2 of the ConCast, with Abhinav Sekhri, on the criminal legal standards applied – or not applied – by the Court). For the purpose of argument, let us say that the Court found – as was its prerogative to find – that the petitioners had failed to provide adequate evidence to dislodge the SIT’s findings of no political conspiracy, and that therefore, the writ petition had to be dismissed.

But that is not the only thing that Justice Khanwilkar did. First, he spent some time in the judgment lavishing fulsome praise on the executive authorities (“indefatigable work”) – something particularly embarrassing, coming from a constitutional court, in a case involving large-scale riots. Most seriously, however, he then went on to note that this case was the result of a “coalesced effort by disgruntled officials”, that those who had brought the present proceedings “had the audacity to question the integrity of every functionary … to keep the pot boiling”, and “all those involved in such abuse of process, need to be in the dock and proceeded with in accordance with law.”

There are a few things we need to note about these lines. The first is that in a functioning legal system, lines such as these would invite an immediate action for defamation, with heavy damages to follow. None of that, however, applies here: following the example set by Khanwilkar J., it seems that Supreme Court Justices, in the course of their official duties, are free to engage in character assassination, insinuations, and personal attacks, without being called upon to provide a shred of evidence for the same. Forget evidence, the Supreme Court did not even accord the petitioners the courtesy of a hearing on this point before damning them through its judgment. Needless to say, at the next available opportunity – judicial or extra-judicial – the same Supreme Court is likely to issue moral lectures on the principles of natural justice.

But what followed is even more alarming. The day after these “observations”, Teesta Setalvad – petitioner no. 2 in this case – was arrested by the Gujarat Police. The paragraph of the Supreme Court judgment that I have extracted above was the literal basis of this arrest: it was cited in the FIR. In other words, the Supreme Court – through Khanwilkar J – by making statements such as “all those involved in such abuse of process need to be in the dock” laid the groundwork for an arrest that State authorities followed up on within hours. And this arrest – it is important to note – was on the basis of a judgment in a case filed under Article 32 of the Constitution, which guarantees the right to move the Supreme Court for the enforcement of rights; in other words, the petitioner in a case filed against alleged State impunity, before the Supreme Court, was arrested by the State, based on the judgment of the Supreme Court.

At the time of writing, Teesta Setalvad remains in jail.

Perhaps you may say that this is a one-off, an aberration. Except that, a few days later, the same thing happened all over again, and once again it was Justice Khanwilkar who was the senior judge on the bench (although the actual judgment was written by a future Chief Justice of India, Justice J.B. Pardiwala). Himanshu Kumar vs State of Chhatisgarh involved a 2009 petition regarding extra-judicial encounter killings in the state of Chhatisgarh. As in Zakia Jafri’s case, this was an Article 32 petition against State impunity, seeking police accountability for a massacre of adivasis. As in Zakia Jafri’s case, the Supreme Court dismissed the petition, and then took it upon itself to do more. First, it imposed a fine of Rs 5 lakhs on the petitioner, Himanshu Kumar. And then, as in Zakia Jafri’s case, it laid the groundwork for legal action against the petitioner. It noted that:

We leave it to the State of Chhattisgarh/CBI (Central Bureau of Investigation) to take appropriate steps in accordance with law as discussed above in reference to the assertions made in the interim application. We clarify that it shall not be limited only to the offence under Section 211 of the IPC. A case of criminal conspiracy or any other offence under the IPC may also surface.

Notice, once again, the loose language used by a Constitutional Court in a case that involved the undisputed massacre of adivasis: that a “case of criminal conspiracy or any other offence” under the IPC “may also surface.” Without evidence. Without a hearing. Once again, this is exactly the kind of stuff that gets you cleaned out for defamation in functioning legal systems; maybe it even would in India, unless you’re the Supreme Court. If you’re the Supreme Court – and especially if Justice Khanwilkar is on the bench – it’s open season, especially on citizens who take Ambedkar seriously when he said that Article 32 was the “heart and soul of the Constitution.”

It is also important to note that during the pronouncement, the Court only referred to the State of Chhatisgarh. The reference to the Central Bureau of Investigation [“CBI”] was added subsequently to the judgment, on the oral request of the Solicitor-General, after the pronouncement. Once again, you can see the attitude of the Constitutional Court in cases like this: just add a reference to a central investigative agency in the judgment, on the request of the union government’s lawyer, as if it was the correction of a typographical error. What else can we call this, other than the executive(‘s) court?

These two judgments – driven by Justice Khanwilkar – mark a profoundly dangerous shift in the history of the Supreme Court. It is one thing for the Court to dismiss Article 32 petitions against State impunity. However, it is quite another – and truly unprecedented – for the Supreme Court to turn upon the petitioners themselves, and pass prejudicial remarks against them that then become the basis of FIRs and jail time. In every way, this is an inversion of the rule of law, of the Constitution, and of the Supreme Court itself: from the protector and guarantor of fundamental rights, to persecutor-in-chief. Idi Amin famously said: “I can guarantee freedom of speech, but I cannot guarantee freedom after speech.” Likewise, through these judgments, Justice Khanwilkar has said: “I can guarantee freedom to come to Court; but I cannot guarantee freedom once you’ve come to Court.”

Sabarimala: The Unreasoned Volte-Face

The final case that I want to (briefly) analyse is not strictly in the same line of cases as the others, but does bear a family resemblance, in terms of significant judicial action not backed up by any reasons whatsoever.

In November 2018, a five-judge bench of the Supreme Court held that the Sabarimala Temple’s ban upon the entry of women between the ages of ten to fifty was unconstitutional. The verdict was 4 – 1. Chief Justice Dipak Misra and Justices Khanwilkar, Chandrachud, and Nariman held against the exclusion. Justice Indu Malhotra dissented. All judges except for Justice Khanwilkar wrote separate opinion; Khanwilkar J joined the opinion of the Chief Justice.

I do not, in this post, intend to re-litigate the correctness of the Sabarimala judgment. The point, however, is this: an application for review was filed. Recall that for the Supreme Court to review its own judgment, it is not enough to just show that the judgment under review was mistaken on law, but to show that there was an inescapable error, on the very face of the record (that phrase, prima facie, again!) – and that this has to be demonstrated before the same bench that passed the original judgment.

The Sabarimala review was heard in open court. At the time, Chief Justice Dipak Misra had retired, and had been replaced by Chief Justice Gogoi. The rest of the bench was the same.

By a 3-2 verdict, the Supreme Court decided to “refer” certain “questions” about the correctness of the Sabarimala judgment for interpretation to a larger bench (this, effectively, stayed the implementation of the judgment). Two of the judges who voted to refer were CJI Gogoi (new to the case) and Malhotra J (a dissenter in the original judgment). Two of the judges who dissented were Chandrachud and Nariman JJ (both in the majority in the original judgment). The tie-breaking vote was that of Khanwilkar J, who had been in the majority one year before, but now seemingly believed not only that the judgment that he had signed on to was arguably wrong, but so wrong – so prima facie wrong – that the threshold for review was activated.

Can a judge change their mind about the correctness of a judgment they have signed onto? Yes, of course. We are all changeable creatures. Can a judge change their mind about the correctness of a judgment they have signed on to so much that they not only believe they were wrong, but blatantly, egregiously wrong – within a year? Perhaps. Perhaps Justice Khanwilkar had a Damascene moment about the rights of women to enter temples. But if that is the case, is there not a minimum – a bare minimum – requirement for a judge to explain themselves? To provide reasons for a 180-degree turn? What is notable is that in neither of the two cases – Sabarimala or Sabarimala “Review” – did Justice Khanwilkar do us the courtesy of a reasoned opinion. We do not know the reasons why he agreed with his brother, the Chief Justice, in 2018; and we do not know the reasons why he came to believe that his brother, the Chief Justice, was egregiously wrong in 2019. Walt Whitman could well ask the rhetorical question, “do I contradict myself?”, and expect his readers to nod knowingly when he answered, “very well then, I contradict myself”, but that is not open to a Supreme Court Justice who, with a stroke of the pen, can extend or withdraw rights from millions of people.

Conclusion: The Executive(‘s) Court

These examples could be multiplied. One could talk about Khanwilkar J.’s majority opinion in Romila Thapar vs Union of India – another UAPA case – where the Supreme Court turned a blind eye to obvious police misconduct in the prosecution of a case (see Abhinav Sekhri’s analysis here), and at the time of writing, the accused are still in jail without trial (can you see a trend here?); one could talk about the Central Vista Judgment, where Khanwilkar J’s majority opinion laid down a standard of public participation, and then refused to apply it to the facts at hand; one could talk about all these, but there is little benefit in belabouring the point.

And the point is this: the cases that we have discussed involve some of the most basic and crucial civil rights in our Constitution. Watali and PMLA involved the right to personal liberty; FCRA involved the right to freedom of speech and freedom of association; Zakia Jafri and Himanshu Kumar involved the right to enforce fundamental rights, and the right to seek judicial remedies against State impunity. Enforcement of these rights is at the heart of the rule of law, at the heart of what it means to be a constitutional democracy governed by the rule of law rather than by State arbitrariness. Each of these rights is a crucial bulwark between the individual and the State, and it is the task of the Court to preserve and maintain that bulwark.

However, when we look at the judgments in these cases (four out of five were authored by Khanwilkar J, and he was a party to the fifth), a disturbing picture emerges. It is not simply that the State always wins, and the individual always loses; regrettably, that is a familiar story in the history of our constitutional jurisprudence, with only a few exceptions scattered on the sands of time. Rather, it is the manner in which the State wins. When it comes to the State’s claims, the State’s interests, the State’s (presented) facts, the State’s vision of the world, the Court treats all this with a feather-light touch, takes everything as true, and occasionally takes the time out to praise the State and its authorities for the great job that they are doing. On the other hand, when it comes to the individual, the Court turns into the proverbial “lion under the throne”, baring its fangs and unsheathing its claws. Under this judicial philosophy, rights are nuisances, individuals are dispensable, and to approach the Court for justice is like playing a game of Russian Roulette: it’s you who might end up in jail after the dust has cleared. And, as Justice Khanwilkar’s conduct in Sabarimala shows, none of this needs justification: it is not the exercise of reason that drives this judicial philosophy, but the exercise of raw power. The Court does, because it can. And that’s about it.

This phenomenon of judicial rule by decree, of orders without reason – the language of the executive, in other words – is why, in a previous post, I referred to the Court led by the previous Chief Justice as an “executive Court”: “an institution that speaks the language of the executive, and has become indistinguishable from the executive.” Judgments in cases like Watali, for example, are classic examples of the workings of an executive court. But at the same time, the observations in the FCRA Case, and in Zakia Jafri and Himanshu Kumar’s cases, are more than just that: it is not simply that the Court is speaking the language of the executive, but has become an institution where executive ideology can be laundered, and shown to the world as sparkling, judicially-declared truth. This is what happens when, in FCRA, Khanwilkar J speaks about citizens needing to “be firm and resolute” so that they wouldn’t need foreign remittances; and this is what happens in Zakia Jafri, where Khanwilkar J’s character assassination of Teesta Setalvad and the suggestion that she be “put in the dock” is immediately followed up by an FIR (which quotes his very words), arrest, and jail.

Khanwilker J is now gone. His individual legacy can be measured in the months, the years, and the decades that people have spent and will spend in jail, without trial (indeed, the State’s lawyers have already begun arguing that under the PMLA, a Court can only ever grant bail on health grounds, and never otherwise). It can be measured in ruined lives and broken futures. But it is the coming time that will reveal whether the normalising of the Supreme Court as the executive(‘s) court would, at the end of the day, be his most significant contribution to Indian constitutional jurisprudence.

Guest Post: MeitY-enforced horizontal application of Constitutional Rights

[This is a guest post by Utkarsh Srivastava.]


On 6 June 2022, the Indian Ministry of Electronics and Information Technology (MeitY) released a draft of the proposed amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (Intermediary Guidelines). While the major concern in the amendments is the proposed appointment of Government Appellate Committees (GAC) which could become the ultimate censorship body for content on online platforms, another proposed obligation (which has largely flown under the radar) could have a massive impact on content moderation as well. The proposed addition of Rule 3(1)(n) to the Intermediary Guidelines will require the intermediary to ‘respect the rights accorded to the citizens under the Constitution of India’ (Rights Obligation). As I discuss below, the consequences of this obligation could cause extreme confusion regarding content moderation with regards to posts by Indian citizens on social media platforms.

Intermediary Protection

Intermediary protection allows entities which handle data on behalf of another person or provide any service with respect to that data to avoid liability with regards to that record. Indian law requires intermediaries to comply with certain conditions to avail intermediary protection. One of these conditions is compliance with the Intermediary Guidelines. If the proposed amendments are adopted into law, then it would require intermediaries to comply with the Rights Obligation and accordingly ‘respect the Constitutional rights of Indian citizens’. The language in Rule 3(1)(n) puts a mandatory obligation on the intermediaries as it uses the word “shall” (MeitY has also stated that it may be inclined to harden the language of this rule). This would entail a horizontal application of Fundamental Rights, marking the first time that the State has made a private entity liable to respect Constitutional rights through its rulemaking power.

Horizontal application of Constitutional rights

This blog has earlier examined how while the default rule is to enforce Constitutional rights against the State (Vertical Application), Constitutional courts have found ways to apply Constitutional rights in matters involving private actors (Horizontal Application). Of the four approaches discussed in the earlier analysis, two require the entity against whom the rights are being enforced to have a nexus with the State and are therefore inapplicable in the present scenario. The third approach is one of indirect horizontality wherein the respondent is a private actor who in its capacity as a private actor uses a law to justify its acts. The law is then challenged on the basis of indirect horizontality. This approach too is inapplicable in the present scenario. Finally, the fourth approach is where the private act of a private party is challenged on the touchstone of the Constitution. This approach typically applies only where the Constitution specifically outlaws horizontal rights violations, i.e. (a) restriction from access to public places on the basis of religion, race, caste, sex or place of birth; (b) untouchability; and (c) human trafficking and bonded labour.

The Rights Obligation is a fifth approach of Horizontal Application where the government has made private entities liable to comply with Constitutional protections. Bear in mind that this action would be dubious even if done by the legislature. That it is done by the executive makes it even more Constitutionally suspect. The language in Rule 3(1)(n) is broad and does not carve out any exceptions.

Effect of the Rights Obligation on the Freedom of Speech

Currently, social media platforms regulate speech on basis of their community standards/terms of service (Platform Restrictions). Therefore, while certain speech may not be illegal per se, it may be removed on the grounds of infringing the Platform Restrictions. However, under the Rights Obligation, users will be able to enforce their Fundamental Right to Freedom of Speech (FoS Right) against the platforms. The FoS Right is a broad one which accords citizens the right to freedom of speech and expression, and is limited only by Article 19(2). Since the Article 19 was clearly intended to govern State action (as recognised by the Supreme Court in PD Shamdasani v. Central Bank of India, AIR 1952 SC 59), Article 19(2) only applies to laws and platforms cannot use it to protect their content moderation actions. Theoretically, the Rights Obligation could lead to a scenario where the user commands absolute right to freedom of speech on platforms with no restrictions (including those that the platform might seek to enforce under Platform Restrictions).

This blog has already discussed how the GAC proposed under the amendments may force the platforms to host speech contrary to Platform Restrictions and how that may violate the platform’s own free speech rights. This current post argues that the government may have added rules to provide backing to such compelled speech by way of Rule 3(1)(n).

Harmonious reading

Intermediary Guidelines themselves permit content moderation under Rules 3(1)(b) (which lists out prohibited information) and 3(1)(d) (which requires removal of prohibited information). Further, they provide for a mandatory grievance redressal mechanism that the intermediary must implement. Therefore, there is some grounds for the platforms to moderate content.

However, since Rights Obligation must also be recognized, the Intermediary Guidelines must be read harmoniously and the only way to do so is such that content moderation can only be on the basis of the categories of prohibited information listed in Rule 3(1)(b) and the platforms may not enforce any other restrictions including Platform Restrictions. This is different from the earlier approach of intermediary regulation as earlier there was a baseline set up by listing prohibited information and the platforms were allowed to go over and above the prohibited information restrictions and add their own restrictions.

Government views

The Rights Obligation must be seen in light of the government’s public views on the matter. The Minister of State for Electronics and Information Technology Rajeev Chandrashekhar has been the driving force behind recent Indian laws governing the Internet and chaired the discussion on proposed amendments on 23 June 2022. In the context of former US President Donald Trump being banned from various social media platforms, Chandrashekhar had tweeted: “Deplatforming is a big deal – Its a violation of fundamental rights of users n must hv force of law behind it for any platform to exercise n must never ever be be done arbitrarily.

The government therefore seems to be inclined to protect the rights of the user on the platforms even where the platform finds them in violation of its Platform Restrictions. Indeed, the MeitY has stated before the Delhi High Court that “Only in cases where the majority of the contents/posts/tweets in a user account are unlawful, the platform may take the extreme step of taking down the whole information or suspending the whole account.” At the very least, this implies that the government is comfortable in overruling the Platform Restrictions and severely limit the control platforms have over the content hosted by them if they wish to retain their intermediary protection.

This would also be in line with certain efforts in the USA where State governments have sought to put limitations on how platforms regulate content. Given the effect social media has had on the real world, it appears that governments are uncomfortable with the power wielded by platforms and are trying to enact laws which overrule Platform Restrictions. However, this would take away from the important function served by the platforms where they prevent/reduce undesirable/harmful speech from flooding their users’ timelines.

More questions than answers

The Rights Obligation through a one-line addition to the Intermediary Guidelines raises some very complicated questions:

  1. Are platforms now required to abandon their own Platform Restrictions and only moderate content posted by Indian citizens from the lens of information prohibited by Rule 3(1)(b)?
  2. Would moderation techniques which do not involve outright removal of content, but instead reduce the reach of the content (for instance by making posts from certain accounts ineligible for sharing/recommendation on the platform or displaying replies from the account in a lower position in conversations) be hit by the Rights Obligation?
  3. In addition to approaching the GAC in case of moderation of content, do citizens now also have the right to approach the Constitutional courts for violation of their Constitutional rights by platforms?
  4. What other Constitutional rights are protected by the Rights Obligation? For instance, Article 301 provides for the freedom of trade, commerce and intercourse. Does this mean that platforms must take advertisements from all businesses regardless of whether it might be part of their Platform Restrictions to prohibit certain kinds of businesses from advertising on the platform?

Another potential wrinkle to this situation may be through the Delhi High Court’s decision in Sanjay R Hegde v. The Ministry of Electronics And Information Technology, W.P.(C) 13275/2019, CM APPLs. 53972/2019, 5543/2020 wherein the petitioner has challenged the “illegal suspension” of his Twitter account. While the Intermediary Guidelines themselves did not exist at the time of the suspension of Mr Hegde’s account, the court may be roped into the controversy as well, especially if the proposed amendments are passed into law prior to the delivery of the judgment in the matter.

The proposed amendment to the Intermediary Guidelines are still undergoing discussion and consultation. The government must consider the Pandora’s box it may open with the proposed amendment including through the Rights Obligation before passing the amendments into law. Otherwise, platforms may find it extremely difficult to moderate content as they will be caught in a Catch-22 situation where (a) users will abandon the platform for showing undesirable content; or (b) the courts may decide that platforms are not eligible for intermediary protection and are therefore liable for the content posted.

As this blog has already articulated:

The Union government is trying to solve a real problem. Online platforms have often acted arbitrarily in taking down content (see here and here) and providing users some redress against unreasoned takedowns may be well-intentioned.

However, the Rights Obligation is an extremely broad solution to this problem which will at best cause extreme confusion and at worst result in the platforms being full of undesired content.

Guest Post: Two Courts, Two Conclusions: Abortion Law in India

[This is a guest post by Gauri Pillai.]


On 15 July 2022, the Delhi High Court, in Ms X v The Principal Secretary of Health and Family Welfare Department Government of NCT of Delhi (‘Ms X’), denied an unmarried woman permission to terminate her unwanted pregnancy at 23 weeks under Section 3 of the Medical Termination of Pregnancy Act, 1971. A week later, through an ad-interim order, the Supreme Court allowed the woman’s appeal against the Delhi High Court’s order, permitting termination as long as a Medical Board confirms that abortion is medically safe for the pregnant woman. In this post, I discuss both these orders, which differ not just in their outcomes but also their reading of the law, and the factors they take into account in decision-making. But first, I briefly set out the legislative framework governing abortion in India.

Legal Background

The legal regulation of abortion in India began with criminalisation. The Indian Penal Code 1860 (‘IPC’) declared criminal voluntarily causing a woman with child to miscarry, unless done in good faith for saving the life of the woman. It brought within the scope of the law both the person causing the miscarriage, and the woman herself. Predictably, criminalisation did not eliminate the need for abortions and only compelled women to seek abortions in unsafe settings from unskilled practitioners, causing death. To quell the rising mortality rate, the State introduced the Medical Termination of Pregnancy Act (‘MTPA’) in 1971, conditionally legalising certain categories of abortions, and exempting them from criminal sanction.

The MTPA allows termination up to an outer limit of 20 weeks. Termination is permitted only when continuation of pregnancy would cause grave injury to the physical or mental health of the pregnant woman, or where there is substantial risk that the foetus, if born, would suffer from such physical or mental abnormalities so as to be seriously handicapped. The 1971 Act was amended in April 2021. As per the amended law, the limit of 20 weeks has been extended to 24 weeks—as long as the above conditions are met—for certain categories of women, as specified in the Medical Termination of Pregnancy Rules 2003, amended in October 2021. Rule 3B lists the following categories: (a) survivors of sexual assault or rape or incest; (b) minors; (c) change of marital status during the ongoing pregnancy (widowhood and divorce); (d) women with physical disabilities (e) mentally ill women including mental retardation; (f) the foetal malformation that has substantial risk of being incompatible with life or if the child is born it may suffer from such physical or mental abnormalities to be seriously handicapped; and (g) women with pregnancy in humanitarian settings or disaster or emergency situations as may be declared by the Government. An assessment as to whether the conditions specified are met is carried out by one medical professional for termination before 20 weeks and two medical professionals for termination between 20 and 24 weeks.

Inclusion of Unmarried Women

An immediate question before both the Delhi High Court and the Supreme Court in Ms X was whether an unmarried woman with a unwanted pregnancy at 23 weeks fell within the ambit of Rule 3B. The Delhi High Court adopted a literal interpretation, holding that since the rule did explicitly not specify unmarried women (unlike, say, widows or divorcees), it did not extend to the petitioner. The Supreme Court, in contrast, deviated from this ‘unduly restrictive’ interpretation, and adopted a purposive reading of the MTPA, making four crucial observations.

First, the Court held that the listed categories of widows and divorcees were simply illustrative of the broader category of change in marital status; they did not exhaust it. This implied that other instances of change in relationship status would also fall within the ambit of the rule. In this case, the petitioner, who was in a consensual relationship, sought termination because she was deserted by her partner: in other words, she decided to terminate her pregnancy due to a change in relationship status, accounted for by Rule 3B. Second, the Court noted that the 2021 amendment to the MTPA modified Section 3 to extend it to unmarried women. Section 3 states that a grave injury to mental health justifying termination can be presumed if pregnancy is on account of failure of contraception. While the 1971 Act saw such failure of contraception as occurring only between a ‘ married woman and her husband’, the 2021 Amendments extend it to ‘any woman and her partner’. This, the Court held, clearly indicated legislative intent to bring pregnancies outside marriage within the reach of the MTPA. Third, the Court observed that the MTPA recognises the ‘reproductive choice of a woman and her bodily integrity and autonomy. Both these rights embody the notion that a choice must inhere in a woman on whether or not to bear a child’. In light of this underlying purpose, denying the petitioner the choice of termination simply because she is unmarried, the Court remarked, would be at odds with the spirit of the MTPA. Finally, the Court held that live-in relationships have previously been recognised by the Supreme Court, which has refused to impose subjective notions of social mortality through the law to ‘unduly interfere with the domain of personal autonomy’. Drawing the four arguments together, the Supreme Court allowed the unmarried petitioner to terminate her unwanted pregnancy.

This is certainly a crucial development in Indian abortion law. The exclusion of unmarried women from Rule 3B speaks to the law’s questionable elevation of marriage as the only form of relationship that deserves legal recognition, on the basis of which certain groups of women are denied an essential form of healthcare (access to abortion). The exclusion also cements the patriarchal assumption that women ought to express their sexuality only within the institution of marriage. Women who experience a change in marital status (on account of death of their partner or divorce) are thus seen as deserving of the law’s sympathy, while women who engage in sex outside of a marital relationship are punished through withholding their right to access an abortion. In reading Rule 3B and Section 3 to include unmarried women, the Supreme Court rightly refuses to perpetuate these assumptions.

Ignoring the Body

Beyond the Delhi High Court’s restrictive reading of Rule 3B, the High Court also shockingly minimised—to the extent of ignoring—the bodily impact of pregnancy. During the hearing the Court told the petitioner that ‘hardly 12 weeks’ were left to carry the pregnancy to term, after which she could give up the child for adoption. When the petitioner remained firm in her decision to terminate the pregnancy, the Court responded by offering to pay for childbirth in a good hospital. All the petitioner then had to do, the Court’s observations implied, was to carry the pregnancy to term for the remaining 12 weeks. Is that not a reasonable ask, the Court appeared to entreat, to ensure that the foetus survives? In refusing termination, the Court in essence concluded that it was, indeed, a reasonable ask.

However, in arriving at this conclusion, the Court completely disregarded the impact of an unwanted pregnancy on the body of the pregnant women. Pregnancy involves ‘all the body systems, displacing body parts, depleting the body of its necessary elements and changing its chemical balance’. As has been noted within the literature, it increases blood volume by 50%, stroke volume of the heart by 35% and renal function by 50-60%. It decreases lung volume by 20% and the respiratory rate by 15% (2-3 breaths per minute). The pituitary gland enlarges 135% and the production of growth hormones increases dramatically. A whole new organ, the placenta, is generated by the body. The excess progesterone produced by the placenta can cause fluid retention, increase in blood pressure, weight gain and inability to sleep. The size and position of the heart and the uterus change, the latter displacing and compressing other organs in the gastro-intestinal tract. The resulting pressure may affect the circulation of blood, sometimes causing irreversible varicose veins, haemorrhoids and disabling thrombophlebitis. The interference with the gastro-intestinal tract can also cause constipation. The displacement of the urinary tract can result in urinary tract infections. The weight of the uterus can create sacroiliac strain, backache and pressure on the cervical spine, potentially resulting in numbness, tingling, and proprioceptive acuity reduction in the hands. Metabolically, the pregnant woman is in an accelerated stage of starvation due to the nutritional demands of a growing foetus. Increase in oestrogen levels can cause nausea and vomiting in the first four months, resulting in dehydration, extreme fatigue, headaches, confusion, fainting, low blood pressure, rapid heart rate, and anxiety or depression. Childbirth frequently damages the pelvic organs which might then require corrective surgery. Bladder control may be permanently lost.

Notably, these are the biological indications of a medically ‘normal’ pregnancy. Complications during pregnancy could result in a worsening of these conditions or an exacerbation of pre-existing health conditions such as asthma, high-blood pressure or thyroid disease. When the pregnancy is desired, the pregnant woman willingly assumes this demanding physical responsibility. However, when the pregnancy is unwanted, these significant biological changes are externally imposed, and their cost undervalued (or ignored), motivated by the ‘latent assumption’ that ‘motherhood is women’s ‘normal’ condition’, rendering state actors ‘oblivious to the life-consuming consequences of forcing women to perform its work…[A] legislature may not decide that it is reasonable to save unborn life by compelling pregnancy ‘but for’ the archaic or stereotypic assumptions about women it holds’.

The strength of the stereotype is heightened when we acknowledge that in no other context does the law demand that one individual offer bodily assistance to another. As some authors note, the law does not mandate that one jump into a river to save a child from drowning. Closer to the context of pregnancy, the law does not require that a parent run into a burning house to rescue a child, or compel a parent to donate a kidney to a child who needs a kidney even for one day, forget 12 weeks (see here, here and here). Of course, a parent might willingly go into a burning house or donate a kidney to save her child, just like a pregnant woman, who desires the pregnancy, willingly provides her body to nurture the foetus. However, at issue here is an unwanted pregnancy, not a wanted one. Despite refusing to impose a similar responsibility in other analogous contexts, the law on abortion expects pregnant women to protect the foetus in this manner, indicating the role of underlying assumptions about women as mothers.

The Delhi High Court, in its order or during the hearing, made no reference to this demanding, one-of-a-kind physical responsibility imposed on women. The bodily cost borne by women in carrying to term an unwanted pregnancy was thus not just undervalued by the Court, but outrightly ignored. The Supreme Court’s order offered some respite, by emphasising women’s right to bodily integrity. However, even the Supreme Court did not acknowledge the sheer extent of the bodily demand made on women. Forcing a woman to carry to term an unwanted pregnancy is not just a question about her body being used against her will, but also about how her body is being used, and the magnitude of the ask being made of her. Without explicitly recognising the immense bodily responsibility that pregnancy is, especially when it is unwanted, the cost to women from compelling pregnancy will always be diluted (or worse, disregarded).

Growing Prominence of Foetal Interests

The Delhi High Court’s observations during the hearing also speak to another worrying trend: the growing prominence of foetal interests in India’s regulation of abortion. The Court repeatedly stated that allowing termination at 23 weeks would ‘virtually amount to killing the child’. In so observing, the Court seems to view the 23-week old foetus as a child (presumably with a right to life). This is contrary to the earlier Supreme Court decision in Suchitra Srivastava, which views the foetus only as a ‘prospective child’, and the Bombay High Court decision in High Court on its Own Motion where the Court held that ‘an unborn foetus is not an entity with human rights… A child when born and takes first breath, is a human entity’. Even parliamentary debates on the MTPA make clear that the foetus, under Indian law, is not seen as an unborn child. Though two members of the Parliament in 1971 and one member in 2020 opposed the MTPA on the basis that abortion is ‘virtually murder’ and a ‘crime against humanity’, their objections were rejected, pointing out that ‘there is no violation of [the right to life] in any manner’. The Delhi High Court decision, then, is inconsonant with precedent and legislative intent in India.

The Supreme Court’s order, in contrast, speaks a different language. In assessing whether the termination ought to be permitted or not, the Supreme Court does not refer to the foetus. While that is certainly a position the law can take—that the foetus ought to be irrelevant in determining the permissibility of abortion—that conclusion has to be reasoned, especially in light of the earlier Supreme Court decision in Suchitra Srivastava which takes the foetus’ status as a prospective child into account as a ‘reasonable limitation’ on the right to abortion. It is therefore important for the Supreme Court in Ms X to set out a principled basis for its stance, failing which it risks contributing to doctrinal confusion on the role of the State interest in preserving the potential life of the foetus in limiting the abortion right; it remains to be seen whether this will happen when the case is finally considered on merits. On this point, instead of silence, a more helpful response would be for courts to ask, and to hold the State accountable for answering with evidence, several key questions: Is the potential life of the foetus a legitimate aim which the State can pursue? Even if it is, are restrictive abortion laws suitable in achieving the aim? Are such laws necessary to protect foetal potentiality? Going forward, the answers to these questions will be central in determining the shape of the law on abortion in India.

Notes from a Foreign Field: The US Supreme Court’s Abortion Judgment in Dobbs v Jackson

[This is a guest post by Aakanksha Saxena.]


The United States Supreme Court delivered its decision in Dobbs v. Jackson on 24th June 2022, holding that there was no fundamental right to abortion guaranteed by the US Constitution, and that decisions about regulating abortion were to be left to the legislatures of the US states. There are five opinions: the majority judgment, three separate concurring opinions, and one dissenting opinion (signed by three justices).

All the opinions deal with various questions – was the Roe Court correct when it recognised the unenumerated right to abortion? Does stare decisis demand following Roe, Casey, and therefore demand upholding the right to abortion? Do the basic principles of liberty and equality require that women be granted bodily autonomy?

This post shall attempt to unravel these threads, and ultimately posit that Chief Justice Roberts’ separate opinion in Dobbs is correct on each of these questions, and ought to have obtained the approval of the Court.

The background to Dobbs

The background to Dobbs is decades of threat, both explicit and implicit, to the decision in Roe v. Wade. Roe held that as part of the right to privacy, a woman has a fundamental right to decide whether to carry a pregnancy to term, subject only to regulations which may be placed by the State in furtherance of its legitimate interests in protecting the health of a woman and foetal life. Roe therefore judicially recognised a fundamental human right, which is otherwise unenumerated in the United States Constitution. This right was held to be enshrined in the 5th Amendment Due Process Clause which provides that no one shall be “deprived of life, liberty or property without due process of law.

The Roe Court also laid down the bright line viability test as the barometer to determine whether a restriction on the right to abortion was valid. This test revolved around whether a foetus was viable outside the womb in order to justify the State interest in protecting it.

A call to overturn Roe was made but turned down by a 5-4 majority in Planned Parenthood v. Casey. The Casey Court however, rejected the viability test, and instead held that the right to abortion stemmed from liberty under the 14th Amendment Due Process Clause, and restrictions on this right were to be tested on an ‘undue burden’ standard. The 14th Amendment came into effect after the Civil War, and is the bulwark of protections of rights against interference by the states.

In any event, Casey cemented Roe as precedent about the recognition of the right to abortion. More relevant to this post is the fact that Casey is also a precedent about precedent and formulated a test to be met by future courts to overrule what is otherwise binding precedent by virtue of the doctrine of stare decisis. Critics of Casey state that it in fact diluted Roe by negating the viability bright line – and they are wrong. Casey, by narrowing Roe but upholding the unenumerated right recognised by Roe, secured Roe’s position and furthered its core holding qua abortion rights into present day – that is, until Dobbs.

A small but relevant detour – the US Supreme Court grants certiorari to limited cases each year, each of which requires to be heard on a constitutional question, and each of which is admitted on this question alone. The question on which certiorari was granted in Dobbs was consideration of the constitutional validity of the Gestational Age Act enacted by the State of Mississippi (“the Act”). Enforcement of the Act had been enjoined by the District Court and affirmed by the Fifth Circuit Court of Appeals. It was specifically urged at the time of granting certiorari that Petitioners did not require Roe to be overruled; a judgement in its favour simply required reconsideration of the bright-line viability test. To be clear – the State of Mississippi sought a clarification whether abortion prohibitions before ‘viability’ are always unconstitutional (see, majority opinion @ pg. 8). Only after certiorari was granted, meaning basically that the Supreme Court would hear the case, did the State launch into a full-frontal assault on Roe and its essential holding.

The majority opinion

The majority opinion (authored by Alito, J., joined by Thomas, Gorsuch, Kavanaugh, and Barrett JJ.) opens with a moral, not a legal debate, laying the foundation for the Court to overturn Roe. What follows is an overbroad discussion of whether the US Constitution confers a right to abortion. At the risk of repetition – this had not been the question before the Court. The Court had already recognised and protected the right to abortion; all it was originally called upon to do, was to examine the validity of a restriction on this right by the State of Mississippi.

There are several reasons marshalled for not adhering to stare decisis without actually engaging with the contours for the same: that the Roe Court was “remarkably loose in its treatment of the constitutional text” since while holding that the right to abortion was part of the right to privacy, it didn’t specifically identify which part of the Constitution provided for either of these rights. The majority further argues that in any event, this was repudiated by the Casey Court when it held that the right to abortion stemmed from the 14th Amendment Due Process Clause.

The Court applies Timbs v. Indiana, to argue that when a right is to be recognised under the 14th Amendment Due Process Clause, it must be deeply rooted in the history and tradition of the United States and be essential to the nation’s ‘scheme of ordered liberty’. The opinion however descends into what must have been meant by the 14th Amendment as ‘liberty’ – at the time it was ratified. This culminates with the majority echoing the words of Thornburgh v. American College that “Roe itself convincingly refutes the notion that the abortion liberty is deeply rooted in the history or tradition of our people.”

The 14th Amendment was ratified and adopted on 9th July 1868. It defined “citizens” and “voters” as male. In addition to being unable to vote, discrimination against women was very much part of the social and national fabric in the USA – in equal wages and hiring, in federally supported education programs, being excluded from the draft, and so on. In fact, it was only in 1923 that the draft ‘Equal Rights Amendment’ was proposed to the Constitution – its deadline was extended multiple times, and it has not yet been ratified. Incremental progress over the years has brought women into the 21st century in the USA. Despite this history of recognition of women’s rights from 1868 till Dobbs, the majority though it fit to go back to 1868 and the position in 1868, to ascertain whether the right to abortion was required to be protected. In fact, as recognised by the Court itself, most states had criminalised abortion at all stages. This is, on a plain reading, a leap unfathomable to the basic belief that the Constitution was envisaged as a document meant to evolve with time and the march of civilisation. This is, as aptly put in the starkest, most truthful line in the Dobbs dissent reflective of the harsh reality that “people did not ratify the Fourteenth Amendment. Men did.”.

The concurring opinions

It is natural that the dangers of the majority opinion are highlighted and attacked in the dissent. This is why the post will now turn to the criticism of the majority opinion by 2 of the 3 concurring opinions, i.e., those authored by Chief Justice Roberts and Justice Kavanaugh. Kavanaugh’s opinion claims to remain neutral on the question of abortion; that the Court’s decision does not outlaw abortion in the Unites States. This concurring opinion also opines that Casey, though relevant to the stare decisis analysis, cannot dictate the question of overruling Roe. This is a convenient sidestepping of Casey which is a precedent about precedent itself, for which the only reason forthcoming is that post-Casey as well, several states have enacted abortion legislations which are contrary to Roe. The opinion balances the precedential value of Roe and Casey and holds that it does not outweigh the political will of these anti-Roe states and their laws. Was this a factor outlined by Casey in its four-pronged test? No.

Justice Roberts’ opinion as well as his stare decisis analysis and application are clear from the quote he uses from Washington State Grange v. Washington State Republican Party (quoting Ashwander v. TVA), and which seems apposite to reproduce – “… the difficulty of a question “admonishes us to observe the wise limitations on our function and to confine ourselves to deciding only what is necessary to the disposition of the immediate case”.

This, it is submitted, is the essence of stare decisis. A doctrine which demands an adherence to precedent to enable consistency and robustness, and requires deviation to be tested and limited to a narrow path. Applying stare decisis especially as laid down by Casey, Roberts J. implores the majority not to conflate the 2 distinct rules of constitutional law laid down by Roe – one, the recognition to the constitutional right to abortion, and second the bright line viability rule formulated basis the State’s interest once a foetus become viable outside the womb. These, he correctly records, are 2 independent rules, and the abandoning of the latter in Casey and subsequent case law did no manner of harm to the validity and binding nature of the former. This is critical – he correctly opines that an independent holding which identifies the mode in which a constitutional right is to be treated ought not to be regarded as part and parcel of the right itself. The viability test raises questions of arbitrariness given that there arise issues in determining and pin-pointing at what stage viability occurs. By way of an example, the Court may recognise the constitutional right to euthanasia / physician assisted suicide, and then also lay down a test on how this right is to be effectively realised, given the State interest in protecting life. A test may be formulated basis how terminal a patient’s illness has to be, in order to avail this right. These are two discrete holdings, and the latter can easily be detached and struck down by a subsequent court, while still preserving the former.

The record of overruling precedent cited by the majority

A large plank of the majority decision was that the SCOTUS itself had plentiful times overruled precedent, and veered off the course of stare decisis. The majority chose to cite 3 instances, all of which are distinguishable on one fundamental basis, i.e. the doctrine of non-retrogression. The 3 judgements cited by the majority all deviated from precedent so as to recognise a new right. However, the Dobbs Court deviates from precedent to achieve the opposite – striking down a previously protected right. This is directly contrary to the doctrine of non-retrogression, which recognises that protection may be granted to rights which are otherwise unenumerated, but once such rights are protected, the protection cannot be taken away or revoked.

Brown v. Board of Education overruled Plessy v. Ferguson and repudiated the doctrine of “separate but equal” i.e., racial segregation in public education. This judgement Roberts, CJ says cannot be compared with Brown which was a unanimous, pithy judgement of the Court. This is relevant since a long-winded and splintered judgement like Dobbs discloses the conflict within the Court and renders the judgement itself vulnerable. The conflict is in fact made clearer from Thomas J.’s opinion which proclaims that other rights under the 14th Amendment require reconsideration – a proclamation that other members of the majority take pains to distance themselves from. Brown on the other hand was a resounding rejection of segregation after reviewing the history of public education since the ratification of the 14th Amendment. The Court in one voice declared that “separate but equal” legislation was unconstitutional and had no place in public education, and the right to a good, equal education was fundamental to democratic society. Several cases in the run-up to Brown had already challenged segregated schools furnishing a judicial basis to depart from precedent, as contrasted from Kavanaugh J.’s argument that several states had enacted laws violative of Roe, i.e. furnishing a legislative basis to defeat precedent.

Adkins v. Children’s Hospital of DC, which held that laws requiring minimum wages to be paid to women was violative of the 5th Amendment’s Due Process Clause, was overruled by West Coast Hotel Co. v. Parrish. This was brought about in a sea-change in economic and social conditions that necessitated a departure from precedent. Up until then, the Lochner era meant that in the garb of freedom of contract, SCOTUS invalidated innumerable worker and consumer protection laws.

Lastly, West Virginia Bd. Of Ed. v. Barnette overruled Minersville School Dist. v. Gobitis, holding that public school students could not be compelled to salute the national flag if it violated religious beliefs, and that consequently schools did not have the right to expel students if they refused to salute the national flag on these grounds. Barnette invoked both freedom of religion and an individual’s freedom of speech—and that freedom of speech included the right not to be forced to speak against one’s will.

The common thread therefore is that all 3 previous incidents of overruling cited by the majority to support their overruling of Roe, were instances where the Court by overruling precedent, recognised a right. I think it is bears relevance that no rights that were previously protected were being eroded or erased.

Slippery slopes recognised by the Court  

There are a few slippery slopes identified in Dobbs. This is yet another reason that Chief Justice Roberts’ approach ought to have been adopted, keeping the judgement’s scope confined and narrow.

The majority holds the view that a broad right to bodily autonomy would somehow justify or encompass fundamental rights to “illicit drug use, prostitution, and the like”. It is telling that the majority chooses to compare the right of a woman to control her own body to activities that have been considered largely illegal / unlawful and have been criminalised across jurisdictions over the years. Further, Dobbs only serves to provide the foundation in the larger movement to expand state surveillance and eventually perhaps erode the right to privacy – a campaign that sex workers have been fighting for decades.

The most chilling is the outright call for reconsideration of rights such as the right to obtain contraceptives, right to privacy in sexual acts, right to same-sex marriage, in the Thomas concurrence. This approach will descend into regression of unimaginable proportions – which is why the other majority judges are keen in distancing themselves from the view.

The other slippery slope is identified by the dissent, with respect to what kinds of legislation would now be possible post-overruling Roe: laws compelling women to carry to full term pregnancies that are the outcome of rape, or compelling women to carry to full term foetuses with known terminal illness, or even laws criminalising the act of travelling across state lines to obtain abortions. There is no answer to this in any of the opinions penned by the majority judges.

Conclusion

Returning the power to decide abortion-related policy to “the people” while at first blush seems to be correct, the structure of federalism in the US means that very often, state legislatures do not perfectly or even adequately represent the will of the people. Further, the deviation from stare decisis to revoke protection to a right which had been recognised and protected for decades is, in no small measure, a cause for alarm. Not only has the Supreme Court retrogressed and undone years of progress as regards the constitutional right to abortion, but the deviation from precedent has also now reopened the gamut of protections which were similarly afforded by substantive due process.

Let there be no mistake – overruling Roe was a major plank on which Trump rode to presidency, which set in motion the chain of events that led to Dobbs. SCOTUS was packed with judges who were appointed simply to overrule Roe, despite some them in fact professing loyalty and deference to Roe and Casey in their congressional hearings. But the politics surrounding the SCOTUS aside, a reading of the opinions displays a sheer wanton disregard for basic principles of constitutional law that now stands available to support similar overruling of settled constitutional precedent. It will do well to heed to the warnings by way of slippery slopes contained in the opinions – none of those rights remain safe.

Guest Post: The Most Recent Clarifications to the Supreme Court’s Bail Guidelines

[This is a guest post by Abhinav Sekhri, and is cross-posted from the Proof of Guilt blog.]


On July 11, 2022, the most recent clarifications were issued by the Supreme Court in respect of the bail guidelines which it had first issued in October, 2021 [MA No. 1849 of 2021 in SLP (Crl) 5191 of 2021, titled ‘Satender Kumar Antil v. CBI‘ (Order dated 11.07.2022)]. This blog had covered the guidelines in October, and then had taken up the first set of clarifications issued by the Court in December, and readers can turn to those posts to get a sense of the background to the most recent order in this series. To be clear, the guidelines in issue here were limited to the issue of bail in scenarios where persons were not arrested during an investigation. 

This short post will only take up the contributions made by the July 11 order, which are, broadly, of two kinds — a further set of clarifications to the existing guidelines, and fresh directions altogether.

The Fresh Clarifications to the Bail Guidelines

Recall that the guidelines worked with a logic of creating four categories of offences for deciding bails in  cases where persons were not arrested during investigation — Category A dealt with offences punishable with imprisonment up to seven years, Category B with offences punishable with more than seven years or death, Category C dealt with offences under special acts with restrictive bail clauses, and Category D was for economic offences not covered by special acts. 

The most lenient approach was asked of courts in respect of Category A, and in respect of Category B cases, the guidelines demanded a ‘case by case’ approach. Not much appears to have changed here at least going by Paragraph 63. But, is it really so for Category B cases? Paragraph 63 does reiterate that “these cases will have to be dealt with on a case-to-case basis” which is the same as the earlier orders, but then it goes on to add that this determination is “keeping in view the general principle of law and the provisions, as discussed by us“. The discussion referred to here takes place through Paragraphs 19 to 62 and it asks courts to follow an approach where coercive processes are strictly kept as a last resort in the non-arrest cases that the guidelines covers. Potentially then, the July 11 order gives a new lease of life to personal liberty for even Category B cases.

In respect of Categories C and D, the clarifications are much more direct, and very substantial. The earlier orders made it uncertain as to whether the fact that a person was not arrested during investigations under a special act would be entirely immaterial when such a person is ultimately appearing before court for bail after completion of investigation, and bail would be governed strictly by the restrictive bail clause. Now, it appears that the Court has made a clean break from this view: 

“65. We may clarify on one aspect which is on the interpretation of Section 170 of the Code. Our discussion made for the other offences would apply to these cases also. To clarify this position, we may hold that if an accused is already under incarceration, then the same would continue, and therefore, it is needless to say that the provision of the special act would get applied thereafter. It is only in a case where the accused is either not arrested consciously by the prosecution or arrested and enlarged on bail, there is no need for further arrest at the instance of the court. …” [Emphasis mine]

Thus, even in cases of special acts, the fact that the prosecution has ‘consciously’ not arrested an accused is significant as it suggests ‘no need for further arrest at the instance of the court’ upon the start of judicial proceedings. This means that, practically agencies would be barred from invoking the harsh bail clauses at least in such cases, and bail would practically be a matter of asking.

The residuary set of economic offences under Category D was the subject of some criticism on this Blog and elsewhere as it had a breathtakingly wide amplitude. It would appear that the Court has acknowledged its error, noting that “it is not advisable on the part of the court to categorise all the offences into one group and deny bail on that basis“. Instead, the Court has turned back the clock, and restored an approach where courts would look at the seriousness of allegations and the gravity of the offence as relevant factors [Paragraph 65].  While this is certainly welcome, one would assume that the same express clarifications rendered in respect of Category C cases — that a conscious decision to not arrest signals no need for further arrest — would also equally apply to Category D cases no matter the seriousness of allegations.              

Breaking New Ground

The first half of the July 11 order is where the Court has broken new ground, as a result of which the guidelines have gone much beyond the initial issue of cases where investigations conclude without arrest. 

Predominantly, this new ground is in respect of the discretion vested in police officers to exercise powers of arrest. Paragraph 23 of the order states that courts will have to be satisfied on compliance with Section 41 of the Criminal Procedure Code which outlines the circumstances in which an arrest can be made, and further that “non-compliance would entitle the accused to a grant of bail” (emphasis mine). Besides Section 41, the Court also turned its focus to the directions given by an earlier judgment [Arnesh Kumar v. State of Bihar, (2014) 8 SCC 273], that ordinarily arrests ought not be made for alleged offences punishable up to seven years imprisonment at the very first instance and instead notices should be sent under Section 41-A of the Cr.P.C. [Paragraphs 24-28]. It has reiterated the importance of these directions,  called upon state governments to facilitate issuance of standing orders for police to secure compliance [Paragraph 29], and also called upon courts to “come down heavily on the officers effecting arrest without due compliance of Section 41 and Section 41A” [Paragraph 30].

With respect to bail jurisdiction itself, there are a few additional contributions made expanding the scope of the guidelines. First, the order notes that delay, where not attributed to the accused, should be a factor in favour of granting bail, and towards this the Court has suggested quick timelines for disposing bail applications [Paragraph 73]. Second, that a magistrate exercising jurisdiction under Section 437, Cr.P.C. is competent to consider bail in respect of offences that are punishable with life imprisonment or death, so long as the offence is one that is triable by a magistrate [Paragraph 55] — suggesting thus that in other cases, magistrates may not be so entitled. Third, simply because Section 439, Cr.P.C. does not explicitly state that young age, sickness, or that the applicant is a woman are factors in favour of granting bail, does not mean that these are not applicable for Section 439 — they are applicable in all cases [Paragraph 58]. Fourth, bail conditions ought not to be mechanically imposed and reasonableness of the bond and surety is something which the court must keep in mind [Paragraph 62].      

Conclusion — Some Old, Some New, Lots Left to Hope

This specific bench of the Supreme Court was first presented with instances of police effecting arrests at the conclusion of an investigation presumably invoking Section 170 of the Cr.P.C. in July 2021, and since then it has made significant efforts to try and curb what it viewed as an approach which unjustly curtailed personal liberty. The guidelines approach was new and one which many, including this Blog, do not agree with. The Court has tried to smooth over some rough edges by melding this new approach with what was the law for some time, and time will tell if this amalgam bears rich fruit. Aware of the socio-legal realities of the Indian criminal process in which our jails are predominantly occupied by undertrial prisoners, and bail ends up being driven more by considerations of guilt or innocence rather than securing appearance of an offender, the Court has expanded its efforts to also try and push for changing this status quo. All in all, the Court can only be commended for making the effort. 

Of course, we have been here before. Many times in fact. In a setup where decisions of arrest and bail are based on exercise of discretion without much statutory guidance on how actors should go about the task, the Supreme Court and various High Courts have tried to fill the gap by issuing guidance on these matters.  Going by the fact that this issue of better exercise of discretion by police and courts is one that is revisited ever so often, it is reasonable to think this guideline-passing exercise only manages to shift the needle ever so imperceptibly on each occasion. Courts obviously know this, and it is for this reason that in the July 11 order the Supreme Court has, once again, called for some legislative guidance on the matter of bail [Paragraphs 67-73]. Statutory guidance through legislation on the lines of the UK Bail Act (referred to by the Court here) is imperative to assure a measure of consistency across individual cases, which is a hallmark of fairness. 

Seven years ago, such a suggestion came from the legislature itself and it led to the issue going before the Law Commission of India; however, midway through the consultative process, the Commission was told that the government no longer wanted to introduce a bail legislation. This exchange resulted in the deeply problematic 268th Report of the Law Commission in 2017. One can only hope that this latest plea for a new legislation from the Supreme Court does not result in mindlessly bringing to life that carceral zombie which the Law Commission had sought to give birth to. Nothing could be more drastically distant from the values that the Supreme Court’s bail guidelines exercise has demonstrated thus far. 

Can the Government’s Appellate Committee for online speech compel platforms to host content?

In June 2022 the Ministry of Electronics & Information Technology unveiled proposed amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (“Intermediary Guidelines”). The amendments propose the creation of one or more Grievance Appellate Committees (“GAC”) to hear appeals against platform decisions to remove or keep-up disputed content. Users can approach the GAC in two situations: (i) if the user complains against content and the platform fails to remove it; and (ii) if the platform voluntarily takes down a user’s content or suspends a user’s account, and the user wishes to reinstate their content or account. The GAC can direct the platform to either remove or reinstate content respectively.

The amendments raise several questions, beginning with whether the creation of the GAC is within the rule-making powers bestowed on the Government under the IT Act. However, this post analyses the power of the GAC to reinstate content that a platform has voluntarily taken down pursuant to its own Terms of Service (“ToS”) (or ‘Community Guidelines or Standards’). Imagine a situation where a platform removes a politician’s post for violating the platform’s ToS on misinformation or hate speech, but the GAC directs the platform to reinstate such speech. In such a situation, the GAC decision would compel the platform to host speech that violates its own ToS. This post argues that under current Article 19(1)(a) doctrine, compelling platforms to host speech contrary to their ToS would violate the platform’s own free speech rights unless the reinstatement furthers the informed decision-making necessary for self-governance. Thus, the GAC’s role should be limited to analysing whether the platform’s decision to remove the content was consistent with the platform’s own ToS.

At the outset, it is important to recognise a few key points. First, a platform’s ToS forms a contractual relationship between user and platform. In other words, a platform’s ToS sets out the kinds of content the platform will allow a user to post on its site – these are often branded as Community Guidelines or Standards. Second, ToS’ are often much broader than actual laws, and platforms may take down legal speech because they believe that such speech would lower the quality of a user’s experience (often referred to as ‘Lawful but Awful’ content). Thus, when enforcing their ToS’, platforms may remove that constitutes misinformation, spam, or nudity even if it doesn’t violate any laws. Third, a user who was aggrieved with a platform’s decision to remove their content could go to court and make a contractual claim that that the user’s content did not actually violate the platform’s ToS. But ToS’ typically grant platform’s broad discretion in the kinds of content they can remove (and users have agreed to this contract); thus it is practically impossible to sue for reinstatement of content. This may also explain the government’s motivation in introducing the GAC, to provide users with a method of getting content reinstated even after it is taken down by a platform.

Compelled speech and free speech

Free speech doctrine has long protected against government interference in the editorial discretion of organisations (e.g., newspapers) to decide what content to publish and what content not to publish. This protects the autonomy interest of the organisation by allowing them to control their message – ensuring the government cannot compel an organisation to publish content inconsistent with their beliefs. It also protects democratic self-governance by ensuring the government does not distort public discourse by compelling various organisations to carry biased pro-government content. However, as with all such freedoms, the right against compelled speech is not absolute.

In Union v Motion Picture Association, the Supreme Court analysed the validity of various statutes which compelled private cinema operators to screen scientific and educational movies. Prima facie this interfered with the editorial discretion of cinema operators to decide what movies to screen. The Court upheld the statutes, but laid down the following test for when it was permissible for the Government to compel organisations to carry speech:

Whether compelled speech will or will not amount to a violation of the freedom of speech and expression, will depend on the nature of a “must carry” provision. If a “must carry” provision furthers informed decision-making which is the essence of the right to free speech and expression, it will not amount to any violation of the fundamental right of speech and expression. If, however, such a provision compels a person to carry out propaganda or project a partisan or distorted point of view, contrary to his wish, it may amount to a restraint on his freedom of speech and expression.

The test laid down by the Court is clear. Where compelled speech furthers the democratic self-governance interest behind Article 19(1)(a), compelled speech may be a permitted interference on editorial discretion. Crucially, the government could not compel cinemas to screen propaganda or biased or false information. The Court also examined the reasonableness of the restriction, specifically noting that the statutes clearly prescribed the types of movies to be screened (scientific and educational) and the ratio of compelled movies to freely chosen movies (1 to 5). Thus, the restriction was narrowly tailored towards achieving the government’s legitimate aim of furthering informed decision-making and did not grant the government unrestricted power to prescribe content.

(As an aside, this is also broadly the approach adopted by the United States Supreme Court and interested readers may refer to Miami Herald Publishing Company v Tornillo and Turner Broadcasting v FCC. The former invalidated a requirement that newspapers publish replies to editorials that the newspaper disagreed with, and the latter upheld a content-neutral “must carry” provision on cable operators on the grounds that ensuring more viewers had access to minimum number of broadcast stations furthered an informed public.)

Interfering with social media’s editorial discretion

Social media platforms exercise editorial discretion when they remove content pursuant to their ToS. Different social media companies have their own ToS’ and they are free to curate speech on their platform based on what they believe will best serve their users. For example, when former President Trump posted misinformation about election ballots, some social media platforms chose to add fact-checking labels and even take down content, while others carried Trump’s posts as is. In other words, platforms decide what content appears on their sites and what content does not, and these decisions reflect the platform’s view on what speech has value to its users – like a newspaper exercising editorial discretion on what content to print.   

When platforms have been accused of favouring certain categories of content, they have been quick to take the defence that they are ‘neutral’ and do not interfere with content. Even in court, they often take the defence that they are neutral intermediaries who cannot be asked to interfere with content or determine its legality. While politically expedient, this claim is also influenced by how Indian law regulates platforms. Section 79(2) of the Information Technology Act, 2000 states that platforms must not interfere with content if they wish to retain legal immunity for hosting unlawful content (colloquially referred to as safe harbour). The principle of safe harbour recognises that while platforms may exercise editorial discretion, the nature of this discretion is different given the high volumes of user uploaded content they host – and platforms need to be protected from the liability of any unlawful content that one of their millions of users may upload. Without safe harbour, platforms could be sued for any unlawful content they may be hosting and thus platforms aim to comply with Section 79(2).

This may suggest that platforms are indeed neutral and do not interfere with content (or exercise editorial discretion). However, Rule 3(1)(d) of the Intermediary Guidelines 2021 expressly notes that where platforms remove content voluntarily, they are not in violation of Section 79(2). In other words, Indian law expressly recognises that platforms remove content voluntarily based on their ToS – and thus exercise what amounts to editorial discretion in determining what content stays up and what content is taken down.

The GAC’s interference with editorial discretion

In the event a platform voluntarily removed content pursuant to their ToS, under the proposed amendments, a user could approach the GAC to have it reinstated. An order by the GAC to reinstate such content contrary to the platform’s ToS would be compelled speech and a direct interference in the platform’s freedom to decide what content to host and what content not to host. Under the principle set out in Union v Motion Picture Association, such an interference would be impermissible unless the content being reinstated furthered a democratic self-governance interest. The Court’s test would not permit compelling a platform to host propaganda or biased or false information.

It is also important to analyse the scope of the power granted to the government. The proposed amendments do not specify the types of content the GAC may direct to be reinstated (unlike the ‘scientific and educational’ films) nor do they specify the volume of content the GAC can direct to be reinstated. Thus, the current framing of the GAC confers unrestricted powers of interference with a platform’s freedom of speech and may not form a ‘reasonable’ restriction on free speech. This concern is exacerbated by the fact that: (i) the Union Government is responsible for appointing members to the GAC but the Union Government’s instrumentalities, or affiliates may also be parties before the GAC; and (ii) the independence and impartiality of the GAC Chairperson or Members are not guaranteed through traditional safeguards such as a transparent selection procedure, minimum qualifications, and security of tenure and salary.

Conclusion

A couple of important caveats must be made about the above argument. First, the proposed amendments are yet to be formally adopted and may undergo changes. Second, most major social media platforms are foreign actors, and the extent to which Article 19(1)(a) rights may be invoked by them remains contested. For example, the Union Government recently argued that Facebook and WhatsApp could not challenge Indian laws that allegedly violated their users privacy because they were foreign companies. Third, platforms may strategically choose to avoid expressly invoking their free speech rights to edit and curate content as it shines a light on how they structure online speech.

Lastly, the Union Government is trying to solve a real problem. Online platforms have often acted arbitrarily in taking down content (see here and here) and providing users some redress against unreasoned takedowns may be well-intentioned. However, any such mechanism must also respect the free speech rights of platforms to not carry speech that they determine violate their ToS. The newly adopted Digital Services Act in Europe for example, allows users to appeal to a dispute settlement body on the ground that the platform violated its own ToS when taking down the user’s content – i.e., the platform acted arbitrarily. If the GAC’s role was limited to ensuring that platforms are enforcing their own ToS’ fairly, by examining whether content removals are in accordance with the relevant platform’s ToS (rather than determining whether the content is legal or illegal), the GAC may provide users with recourse against platform’s that otherwise exercise extraordinary powers over online speech. Nonetheless, the GAC would still have to possess sufficient safeguards to ensure independence, impartiality, and fair procedures. In its current iteration, it risks being a tool for the State to compel platforms to host speech they may have otherwise taken down.  

Home Demolitions and George Orwell’s Supreme Court

There is a line in George Orwell’s 1984, which goes “the Party told you to reject the evidence of your eyes and ears. It was their final, most essential command.

Ongoing proceedings before the Supreme Court pertaining to the spate of home demolitions, which have been carried out across the country by municipal authorities, present a striking example of how judges can continuously reject the evidence of their eyes and ears. These proceedings follow a similar pattern: it is pointed out to the Court (as we have discussed previously on this blog) that the home demolitions – which have now been going on in sporadic fashion for many months – are punitive, and designed to extract retribution for participation in protests. State counsel argue that the municipal authorities are acting in accordance with local laws. The Supreme Court bench – it tends to change – makes a rhetorical statement about how demolitions must follow legal process, makes another rhetorical statement about how it can’t pass “omnibus” orders against the demolitions, and then adjourns the case, as it did today (while the demolitions continue).

In continuously refusing to take cognisance of the fact that the home demolitions are punitive and illegal, and follow the same pattern across the country (as argued previously on this blog), these Supreme Court judges reject the evidence of their eyes and ears. They manage to ignore the fact that, coincidentally, the home demolitions in question come immediately upon the heels of a protest that turns violent, time after time, and are specifically targeted against people who are named by the police in FIRs about rioting, time after time; that in Uttar Pradesh, Javed Anand’s home was demolished one day after the UP Police claimed that he was the “mastermind” behind the June 10 riots, and that in Khargone (MP), Khambhat (Gujarat), Nagaon (Assam), Jahangirpuri (Delhi) and in other places, the exact same pattern is followed (indeed, in Jahangirpuri, demolitions swiftly followed a letter from the BJP leader to the Mayor, asking for bulldozer action against “illegal properties of the rioters”).

Not only that, these Supreme Court judges reject the evidence of their eyes and ears where the punitive character is laid bare by agents of the State. A non-representative sample includes, for example, statements by the Home Minister of Madhya Pradesh, in the wake of the Khargone riots, that the “demolition drive against rioters would continue”; a statement by the Divisional Commissioner, Indore, that “the main idea behind the move is to instil fear of financial losses among the accused”; the Khargone District Collector telling journalists that the demolitions were done to “send a message to rioters”; multiple other statements by Khargone officials, collected here; a statement by the District Administration in Khambhat, Gujarat, that “the encroached properties belonging to the accused are being demolished”; the SDGP in Nagaon, Assam, telling journalists that “some of the suspects [involved in a riot] had encroached upon land … the eviction drive was carried out after a case was registered against them”; and this tweet by the media advisor to the government of UP, with an image of a destroyed home, and the caption “उपद्रवी याद रखें, हर शुक्रवार के बाद एक शनिवार ज़रूर आता है…”.

Examples could be multiplied, but what is abundantly clear is the two-faced character of the State. To maintain the veneer of legality, in its formal orders, the State claims that the demolitions are following due process, and the action has nothing to do with retribution. These are the arguments that the State’s counsel then make in Court. Quite apart from the fact that these arguments fail on their own terms, the basic point is this: the only way that you can accept the State’s arguments – pace Orwell – is if you choose to reject the evidence of your eyes and ears, not once, not twice, but every single time that State agents engage in targeted demolitions after protests, publicly brag about “teaching the rioters a lesson”, and then send their lawyers to argue in Court that the demolitions have nothing to do with the protests.

Finally, it is important to note that the Court is not precluded from taking the evidence of its eyes and ears into account, and crafting appropriate relief. After all, this is a Court that has, over the last four decades, prided itself on turning postcards into PILs, basing interim orders on (credible) newspaper reports, and wielding Article 142 as a sword of complete justice. Nor is it helpless when it comes to framing remedies: in a previous blog post, we discussed the doctrine of an “unconstitutional state of affairs”, which can allow the Court to take cognisance of a systemic pattern of home demolitions across the country; but even without that, existing doctrines such as that of continuing mandamus can serve to address the situation.

However, having taken on this power to do substantive justice, the Court’s refusal to use it in a case where the violation of the rule of law is clear and unambiguous, is a choice from which it cannot escape responsibility. The Court’s bland, oral observations about “omnibus orders” and “following the law”, and its continued kicking of the can down the road while the demolitions continue, allows this two-faced State action to continue with impunity. And its continued refusal to even acknowledge the evidence of its eyes and ears – the evidence of all our eyes and ears – makes us wonder whether the Supreme Court is on the way to becoming George Orwell’s Court.

Guest Post: The Afterparty: Who Is The “Real” Shiv Sena, And Who Decides?

[This is a guest post by Karan Kamath.]


The dispute over which Shiv Sena is the “real” Shiv Sena entitled to the registered party’s name and its bow-and-arrow symbol will be determined by the Election Commission of India. The Commission has the power to do adjudicate factional disputes within registered parties under paragraph 15 of the Election Symbols (Reservation and Allotment) Order 1968. There are several such factional disputes that the Commission has resolved until now. Nonetheless, there are no substantive standards provided in the Order to make that evaluation. Instead, the Order merely provides the procedural threshold that the Commission must take the decision:

…after taking into account all the available facts and circumstances of the case and hearing such representatives of the sections or groups and other persons as desire to be heard.

The overall practical effect of this decision-making process is that it is violative of right to freedom of association. This is because the Commission, on most occasions, prefers to hear legislators rather than primary party members. In any case, it does not even consider the party’s internal rules and regulations, despite mandating that parties have rules on membership and factional dispute resolution. Moreover, besides the political, the Commission’s decision also has wide impact on the economic and proprietary interests of the association as well.

The Political Party as an Association

There is no statutory restriction on what form a political party can take. Section 29A of the Representation of the People Act, 1951 states that “any association of body of individual citizens of India” referring to itself as a political party and intending to register as one, can apply to the Election Commission. Theoretically, a private limited company with individual citizen shareholders could register itself as a political party. (This would not exactly be a novel innovation, for example, the anti-lockdown Reform UK Party is registered as a company, and the defunct centrist Change UK initially was a non-trading company). Even if one considers companies to be incapable of registering as political parties, the words “association or body” of citizens are wide in ambit. This is further evinced by the other provision of Section 29A, which make references to, among other things: “the memorandum or rules and regulations of the association or body, by whatever name called”; “the numerical strength of its members, and if there are categories of its members, the numerical strength in each category”; and “whether it has any local units; if so, at what levels”. This creates a greenfield for the shape or form the party may take. The party can have a memorandum or rules; it can have different categories of memberships; it may or may not have local units.

However, the Commission Guidelines on registration of political parties mandate a “Party Constitution”, which should provide for: democratically elected office bearers (only a third can be nominated); “rules of dispute resolution and discipline”; membership that is open to “all adult citizens of India”; and “merger, split, and dissolution procedure”. This significantly curtails the associational freedom envisioned in Section 29A. But ironically, despite mandating such provisions in the ‘party constitution’, the Commission is never keen on seeking adherence to those rules.

Legislators as the Determinative Factor

If the Commission mandates such a detailed ‘constitution’, it should ideally follow the same in resolving factional disputes. But the practice suggests otherwise. In Sadiq Ali v Election Commission of India, for example, two factions of the Indian National Congress were vying for the grand old party’s erstwhile ‘two bullocks with yoke on’ symbol. The Commission relied on the allegiances of the party’s MPs and MLAs because there were “obvious difficulties” in assessing the primary members’ opinion. Instead, the Commission relied on the opinion of the All India Congress Committee (consisting of delegates to the party conference). The Supreme Court later agreed that the legislators and AICC delegates would correctly reflect the primary members’ views. One of the factions had argued that the party’s constitution be followed to adjudicate on the dispute. But the Commission thought the constitution was “hardly of any assistance” because one faction had removed the other faction’s members from various parts of the party organisational structure. Simultaneously, the Commission concluded that the removals had been “doubtful and open to question”. The Supreme Court confirmed this view of the party constitution.

This approach has two fundamental issues: Firstly, it completely ignores the primary members’ views. To some extent, that may be justified, as not all parties are mass-parties and neither are they obliged to be so. (In theory, parties are free to choose their associational structure, but the Commission Guidelines insist on “democratic spirit”). However, even if the party chooses a governance structure, the Commission’s approach does not consider that either. This is the second issue: if one faction removing the other was “open to question” under the party constitution, the Commission should have ideally examined that question. This does not have the obvious difficulties such as those involved in canvassing opinion of primary members, and also ensures that party constitutions are adhered to. But the Commission did not do that either.

The statute clearly provides freedom of association to political parties. The Commission then constrains that freedom by mandating merger and split procedure, and internal dispute resolution. But when it comes to assessing competing claims to the party legacy, the Commission instead relies on legislators and delegates. On the one hand, delegates may or may not be empowered to provide decisive opinions in case of splits under the party rules. On the other hand, legislators are, historically speaking, mercurial in matters of factionalism. For example, in Edappadi K Palaniswami v TTV Dhinakaran, while the matter was pending before the Commission, the number of MLAs in each faction oscillated. This is not to suggest that primary party members or whoever it is that the party constitution empowers to resolve disputes, should be unchanging in their views. But there ought to be a certainty as to the procedure that is followed. When a person becomes a member of an association, they subscribe to its rules. If the rules say, for example, that only members with ten years of unbroken membership can decide factional disputes, then a new member has subscribed to that rule and to the consequences that follow. But irrespective of what the rules say, if the Commission is to decide matters on the basis of support in legislatures, that is both arbitrary and violative of freedom of association.

It is possible that a party’s rules do not provide for splits or dispute resolution. For example, the Shiv Sena’s constitution does not contain a provision on mergers or splits. Even then, the Commission’s first inquiry must be relate to the party rules to assess who are the officeholders of the party, according to the rules. In some cases, the Commission does take that approach. In Re Dispute in Kerala Congress (M), the Commission did consider support for the rival factions on an organisational level. But, as the dissenting Commissioner pointed out, the lists of organisational level members had inconsistencies and neither faction was in possession of original record. Therefore, until these matters could be resolved, there could not be a judicious assessment of which faction was the ‘real’ party. Alternatively, the majority opinion overlooked these issues and noted that a majority of legislators, MLAs and MPs as well as members of organisational structure supported one faction, which the Commission deemed as being in control of the party. Like Sadiq Ali,instead of resolving the dispute as a dispute over control of an association, the Commission comfortably used support by legislators as a determinative factor. Once again, the Commission’s approach neglected the freedom of association.

The Afterparty

The effect of the Commission’s decision is not limited to immediate formation or dissolution of governments or to claims to party symbols. Political parties, formed as associations, can hold a good amount of funds and property. The Commission’s decision has a real impact on associational property rights. Half a century after Sadiq Ali, the Supreme Court in Janata Dal Party v Indian National Congress was called on to decide whether a lease granted by JDP was valid. The property in question had belonged to INC. After its split, the possession had passed to the losing faction, which eventually merged with JDP. The Court relied on Sadiq Ali to rule that:

ECI, after applying the test of majority at the organizational level and the legislative wings, took the view that Congress (J) group of Congress came to be recognized as the Congress for all purposes. The order of ECI and this Court clearly indicate that the Congress then led by Indira Gandhi had established rights on the properties in question. (Emphasis supplied)

The Commission’s decision on an association’s split is consequential not only for the political nature of the association, but also to its economic and proprietary interests. The current model of dispute resolution followed by the Commission is neither fair nor defends freedom of association. If the Commission is going to have the effective powers to determine questions as to a voluntary association’s property and funds, in addition to its political nature, then, it must have a better approach than the one it currently follows.