Guest Post: Silence and ‘Pragmatism:’ Skirting bail conditions in the UAPA

[This is a guest post by Nitika Khaitan.]


The denial of bail to Safoora Zargar last week drew fresh attention to harsh conditions in the Unlawful Activities Prevention Act 1967, which make it exceedingly difficult to secure bail. Under S. 43D(5) of the Act, no person accused of certain UAPA offences can be released on bail if the court finds reasonable grounds to believe that the accusations against her are prima facie true. Amidst overly broad definitions of these offences, and a low prima facie threshold, how then have courts granted bail? Sometimes, by narrowly interpreting offences or holding that the evidence against the accused is contradictory. At other times, as is the focus of this post, by simply leaving out any mention of the Act or its mandate altogether.

In 2016, for instance, the Supreme Court granted bail to the Dalit activist Angela Sontakke, accused of being a member of the Communist Party of India (Maoist), which is banned as a terrorist organisation. The Bombay High Court had earlier held that her bail was barred by S. 43D(5), since she appeared to be an active member unlike some of her co-accused (who had thus been granted bail by a different High Court bench). While allowing Sontakke bail, the Supreme Court order records that she is charged with offences under Chapters IV and VI of the UAPA, which attract S. 43D(5). But the order doesn’t mention the section. It doesn’t even briefly refer to the evidence against her, let alone record a different prima facie finding from the High Court’s. It speaks merely of balancing the serious charges against her with the facts that she is a woman, has spent years in custody and has yet to see her trial begin.

This is far from the only order that reads as if S. 43D(5) doesn’t exist. In 2017, while granting bail to three Kabir Kala Manch activists accused in the same case as Sontakke, the SC doesn’t even mention the UAPA. Other orders mention just the section number. In the 2017 bail plea of Malegaon blast accused Lt. Col. Prasad Shrikant Purohit, his counsel contended that S. 43D(5) wouldn’t apply since the blast occurred before the amendment that enacted the section. The SC order holds that this plea must be considered at the time of trial and not now. Without excluding the application of S. 43D(5) though, the SC proceeds to effectively ignore it. (The judgment also, oddly, refers to the state as having “rights” to investigate, instead of calling it what it is, a power.)

The SC’s reasoning for granting him bail reads like an order under ordinary law. The SC refers to prima facie satisfaction in support of the charge as one of the factors to consider (true for regular bail), not as the factor that S. 43D(5) elevates above all else. The only other time the Court uses the phrase is in holding that there is a “prima facie case for release on bail,” decidedly not the finding it is mandated to return. Attempts to hunt for the missing reasoning elsewhere in the order fail. The SC does say that there are “variations” and “material contradictions” in chargesheets filed by different investigating agencies (this was also discussed in the Bombay HC order granting bail to another Malegaon blast accused, Pragya Singh Thakur). But before one can infer that this is what led the SC to believe a prima facie case wasn’t made out, the order promptly states that these contradictions too need “to be tested at the time of trial and this Court cannot pick or choose one version over the other.” Almost as if to overcompensate for its missing finding on the evidence, the order repeats thrice on the same page that “at the stage of granting bail, a detailed examination of the evidence” need not be undertaken. (The general proposition is correct, but as held by the SC in the context of another law imposing similar restrictions on bail, “The duty of the court at this stage [of bail] is not to weigh the evidence meticulously… However, while dealing with a special statute… the court may have to probe into the matter deeper”). The order ends by going beyond the usual caveats and emphasises that the grant of bail here “shall be no consideration for grant of bail to other accused persons in the case.”

In none of the cases above did the Court say it was using its extraordinary powers to grant bail where other courts, in light of S. 43D(5), would have been unable to. How do we read the Court’s silences in these orders? Explicitly engaging with S. 43D(5) of course has its pitfalls. As the Andhra Pradesh HC recognised, at the stage of bail, taking the view that the accusations against an accused are not true could dent the prosecution; while taking the opposite view may be akin to “pre-judging the charges.” But even if we allow for the possibility that the Court wanted to refrain from making any observations that would influence lower courts, to not even mention S. 43D(5) in an appeal from a HC that has rejected bail on these grounds goes too far. That this violates the Supreme Court’s own pronouncements is trite. See, for instance, its 2019 decision making clear that courts must apply their mind to the prima facie truth of the accusations. More importantly, while the SC’s orders above indicate an obvious unease with the years in custody that harsh bail conditions inflict on people, they also reflect an unwillingness to fix this unease with anything besides ad-hoc measures, falling far short of the jurisprudence a constitutional court could choose to build.

In the cases above, long years of incarceration played a key role in the Court’s reasoning in favour of bail—over five years for Sontakke, close to four for the Kabir Kala Manch activists and eight years and eight months for Purohit. (This same concern, and elision of the S. 43D(5) mandate, is evident in some High Court orders as well. A 2019 Bombay HC decision, for instance, partly engages with the lack of grounds to prima facie believe the allegations made for certain offences. But for other alleged UAPA offences, the HC simply states that they are punishable with merely two, seven and ten-year imprisonment terms, and the accused had already served nearly four years in jail.) In implicitly acknowledging the injustice of such pre-trial incarceration, while refraining from any systemic change, these SC orders mirror its earlier decisions in the context of other laws with onerous bail conditions.

In 1994, in the context of the Narcotic Drugs and Psychotropic Substances (NDPS) Act, 1985, the SC noted that “to refuse bail on the one hand and to delay trial of cases on the other is clearly unfair and unreasonable” and “if the period of deprivation pending trial becomes unduly long, the fairness assured by Article 21 would receive a jolt.” But noting also that the constitutionality of similarly restrictive bail provisions in the Terrorist and Disruptive Activities (Prevention) Acts, 1985 and 1987 (TADA) had been upheld earlier that year, the Court asked itself, “What then is the remedy?” The Court passed a “one-time” order. It directed all undertrials charged with certain offences to be released on bail, if they’ve spent a certain number of years in custody. The Court said that given the percentage of acquittals under the Act, “we cannot be oblivious to the fact that many innocent persons may also be languishing in jails.” But nothing in the order of course was “intended to interfere” with the future grant of bail by lower courts, which would continue as restrictively as before and presumably also lead to many innocent persons languishing in jails till another one-time SC order.

In a similar order two years later in the context of TADA (Shaheen Welfare Association v. Union), the Court even more explicitly recognised that “when the release of undertrials is severely restricted as in the case of TADA” and a speedy trial is “not practical, release on bail… may, in some cases, be necessary to meet the requirements of Article 21.” Acknowledging that “many of the under-trials may be found to have completed the maximum punishment provided by law by being in jail without a trial,” the Court again offered a “pragmatic approach” / “one-time measure.” The Court divided TADA undertrials into different categories based on whether they were roped in for possession offences or overt acts directly attracting TADA sections, or by virtue of vicarious liability and conspiracy provisions; and directed release on bail on different conditions for each category. The Court recognised that it was overriding the ordinary operation of TADA by creating these classes but held that “while adopting a pragmatic and just approach, no one can dispute the fact that all of them cannot be dealt with by the same yardstick.”

What of pragmatism and justice after the one-time measure then? In Shaheen Welfare Assn., the Court recognised that stringent bail conditions “can be justified… on the presumption that the trial of the accused will take place without undue delay.” The Court had then focused on the inadequacy of Designated Courts set up to try TADA cases as the cause for this undue delay. But gross delay is of course pervasive, not exceptional. While upholding TADA’s constitutionality, the Court had also recognised that TADA was often unscrupulously invoked merely to deny bail. It stopped, however, at terming this sheer “misuse and abuse” and merely exhorting prosecutors and courts to do their job better. But what of the extensive material to show that misuse and abuse are woven into the provisions of extraordinary laws? Particularly in the context of the current repository of most of TADA’s provisions, the UAPA, under which the ‘independent’ authority set up to sanction prosecution is appointed by the executive itself.

Despite repeatedly being confronted with the reality that onerous bail conditions equal years of incarceration without guilt, courts have more often than not resorted to elision and ad-hocism. In a series of decisions from November 2019, the Punjab and Haryana High Court called this out. These decisions were delivered in the context of harsh bail provisions for some offences in the Companies Act, 2013, which prohibit release on bail unless the court is satisfied of reasonable grounds to believe that the accused is “not guilty” of the alleged offence amidst other conditions. The High Court orders state that there is an “inconvenient question, which has not been shown… to have been answered by any court so far, including the Hon’ble Supreme Court. The question is – for how long an accused can be kept in custody on the basis of non-fulfillment” of restrictive bail conditions? The HC order decries “unfortunate situations where a court may not even find the moral courage or the legal sanctity to tell to the accused that he shall have to wait in custody till conclusion of the trial, despite and in face of the legislative policy contained in provisions of Section 436A of the Cr.P.C.” S. 436A of the CrPC mandates the release of under-trial prisoners if they have been incarcerated for half the maximum term of imprisonment for their alleged offence. S. 436A itself carves out an exception for offences punishable with death, but non-obstante clauses in special laws like the UAPA exclude the benefit of S. 436A even for offences punishable with imprisonment for three years. Despite such non-obstante clauses, the HC emphasises that years of custody without trial “cannot be used to curtail the liberty of an accused in violation of Article 21” and poses more inconvenient questions—“In such a situation the court would do substantial justice; or would stick to the [bail] conditions…Even if the courts are to stick to such condition; then how much injustice to the accused would be sufficient to off-set or to balance” the conditions? The HC goes on to hold that unless these questions are “categorically answered to say that till the conclusion of the trial such a person cannot be released on bail,” the onerous conditions cannot be held to be mandatory. (Also see the same bench’s 2018 decision with respect to NDPS cases).

In the context of the UAPA as well, certain High Courts reflect a more sustained engagement with these questions. In a 2014 decision, the Andhra Pradesh HC lays out, colourfully, the cautious and delicate approach needed with provisions like S. 43D(5), comparing it to “the care which a cat is expected [to take] while carrying the kitten in its mouth from one place to another.” The order dilates for several paragraphs on motivated prosecutions; emphasises that “an accused cannot be equated to a convict, even before the trial is conducted;” and goes on to lay out guidelines for courts to appropriately form the prima facie opinion required by S. 43D(5), while taking concerns of liberty seriously.

Such guidelines ultimately may not make too much of a difference—the inherently low threshold of S. 43D(5) no doubt ties the hands of lower courts. Till more authoritative pronouncements on these bail conditions, thus, the road ahead looks bleak. But as I’ve argued in the context of a different set of provisions under the UAPA, judicial logics often defer to state ones with anti-terror laws, accepting the need for extraordinary measures to combat ‘extraordinary’ threats, and making any such authoritative pronouncements unlikely. The history of personal liberty, as a judge from another time had said, is largely the history of insistence upon procedure. Not so with S. 43D(5), where liberty has oft been secured by ignoring it.


[The author thanks Jawahar Raja and Chinmay Kanojia for their help as always with locating UAPA orders; and Sanya Kumar and Megha Bahl for their incisive comments.]

Imprisonment by Metaphor: The Safoora Zargar Bail Order

“… when you choose to play with embers you cannot blame the wind to have carried the spark a bit too far and spread the fire.”

 

When a Court needs to rely upon metaphor instead of law to justify keeping an individual in prison, it is perhaps time for the justice system to take a long, hard look at itself. The Order passed today by a District and Sessions Judge at Patiala House, New Delhi, denying bail to Safoora Zargar, an accused in what has colloquially come to be known as the “Delhi riots case”, is a deeply disturbing one. It is disturbing because it takes the Unlawful Activities Prevention Act [“the UAPA”], a law so stringent that it precludes judges from granting bail if even a “prima facie” case is made out, and then stretches its provisions from one side, and the facts from the other, to ensure that the prima facie case is made out. In the process, what it effectively does – as we shall see – is criminalise the exercise of one set of constitutional rights (the freedom of speech and expression), and deny the exercise of another (personal liberty).

A close reading of the bail order reveals the following:

  1. Taking only the Prosecution’s case (as this was a bail hearing), there is evidence that there existed a “conspiracy” to block a road, which the accused was involved in (the role of the accused in this “conspiracy” – even prima facie – is not spelt out, only some WhatsApp messages and disclosure statements are referred to).
  2. That “one cannot ignore the case of the prosecution that the accused persons have conspired to cause disruption of such an extent and such a magnitude that it would lead to disorderliness and disturbance of law and order at an unprecedented scale.” The Order does not clarify what “unprecedented scale” means. It does not clarify whether the “unprecedented scale” refers to the same “conspiracy” referred to in (1), or whether it refers to something else; if the latter, the Order does not clarify how the participation of the accused was deduced in that separate “conspiracy”; if the former, the Order does not clarify the link between the “conspiracy” to block the road and its “unprecedented scale”, in a country where blocking roads happens every second day.
  3. That although there was no evidence of the accused committing any act or making any speech that instigated violence, nonetheless, as there existed a “conspiracy”, nonetheless “when you choose to play with embers you cannot blame the wind to have carried the spark a bit too far and spread the fire”, and that consequently, the “acts and inflammatory speeches of the co-conspirators are … admissible against the accused.” Now, it is unclear what the “acts” are, as the Order never mentions them; it is also unclear what the “inflammatory speeches” are, as the Order does not mention them either.

The lynchpin of the Order, therefore, is a prima facie finding of a “conspiracy”, in specific terms, to “block a road.” This conspiracy rose to an “unprecedented level” – we are not told how. But the fact that the accused is also – prima facie – one of the conspirators (regardless of specifics, because this remains a prima facie appraisal), meant that ipso facto the “acts and inflammatory speeches” (we are not told which) were attributable to her. It should be immediately clear that such an approach casts the net of criminality so wide, that just about anyone can be brought within its ambit. At the threshold level, it dispenses with the gravity requirement needed to trigger the UAPA, by failing conspicuously to specify how “blocking a road” reaches that threshold; at the more substantive level, upon a prima facie finding of a “conspiracy”, it dispenses with the need to show any causal connection between the accused and the events in question.

This would be problematic for acts (which the accused didn’t commit) as well, but when it comes to “inflammatory speeches” (which the accused didn’t give), it becomes even more problematic. This is because, recognising the problematic character of laws such as the UAPA which make the grant of bail effectively impossible, both the Supreme Court (in Arup Bhuyan, while examining the similarly-worded TADA) and the Bombay High Court (in the Kabir Kala Manch cases) have narrowly interpreted the substantive offence, limiting it to cases involving the incitement of violence. This is, indeed, nothing new: going back to the field of metaphors, as the Supreme Court held in S. Rangarajan, the proximity between speech and consequence needs to be like that of a “spark in a powder keg” for criminality to be imposed.

Now, the image of a “spark in a powder keg” suggests a relationship of immediacy and inevitability. The metaphor chosen by the Sessions Court on the other hand – that of playing with “embers” that the wind then “carries” is the exact opposite of a “spark in a powder keg”. The wind can carry embers as far, and in any direction, that the State or the judge might please; what this effectively does is do away with any causal requirement between speech-act and consequence. Such a doctrine, therefore, buries the fundamental right to free speech: if there is no need for a causal requirement between speech-act and consequence, anything can be criminalised, taking us directly into the territory of thought-crimes.

A reading of the Order, therefore, makes it clear that insofar as both the law and the facts stood in favour of bail, the Court got around the first barrier by replacing legal doctrine with a metaphor of its own invention, and vaulted the second barrier by replacing an accounting of the facts with a set of adjectives (“unprecedented scale” and “inflammatory speeches”) that spared it the necessity of an explanation. In this way, the law was stretched from one side, and the facts from the other, and they met in the middle to make out a prima facie UAPA case.

This prima facie case was then used to justify keeping a pregnant woman in an overcrowded prison in the middle of a nationwide pandemic. What that says about the state of the justice system is best left to the readers’ judgment.

NDPS and the Rise of Punitive Constitutionalism

[Editorial Note: Justice is an indivisible concept. We cannot, therefore, discuss contemporary Supreme Court judgments without also acknowledging the Court’s failure – at an institutional level – to do justice in the case involving sexual harassment allegations against a former Chief Justice. This editorial caveat will remain in place for all future posts on this blog dealing with the Supreme Court, until there is a material change in circumstances.]


Among the many colonial holdovers that continue to persist in our post-colonial legal framework, there is one that stands out as particularly egregious. In a series of “special criminal laws”, there is a provision that – in effect – prohibits a judge from granting bail to an accused unless that judge has “reasonable grounds to believe” that the accused is not guilty. This provision is found, for example, in the Unlawful Activities Prevention Act, in the Narcotic Drugs and Psychotropic Substances Act, and – until it was struck down by the Supreme Court – in the Prevention of Money Laundering Act.

This provision is repugnant to all notions of personal liberty, and it should be easy to see why: it effectively makes the grant of bail next to impossible. An adverserial legal system is premised on the assumption that in a case of competing narratives (i.e., every legal case), the “truth” can be known only when both sides have had a chance to contest each other’s version. The law of evidence – with its focus on cross-examination – is geared towards achieving this: through a detailed set of rules, it affords to both prosecution and defence an opportunity to test each others’ cases, bring out contradictions and conflicts, test the credibility of witnesses, and so on.

An application for bail is not, in principle, about truth or falsehood. In very simple terms, the basic logic of bail is that because the criminal justice system is premised upon the idea of “innocent until proven guilty”, it is presumptively wrong to keep a person in jail until the outcome of their trial – i.e., until a definitive finding has been returned on their guilt or innocence. This is why – traditionally – the considerations that are supposed to weigh with a Court in deciding to grant bail are the potential harms that might accrue from letting an accused out pending trial (he might, for example, flee the jurisdiction, or – if he is a powerful person – intimidate witnesses), measured against the core value of personal liberty (admittedly, this has been watered down substantially over the years, with the “nature of the offence” often playing a dispositive role in courts’ decisions whether or not to grant bail).

Consequently, prohibiting a judge from granting bail unless she is reasonably convinced that the accused did not commit the crime, takes a sledgehammer to the criminal justice process. It requires a judge to take a call on guilt or innocence at the beginning of, or in the middle of, the trial process, without all the information that she needs to do that. In practice, it invariably benefits the prosecution: first, in its very terms – it requires the judge to reach a finding that there is a reasonable likelihood that the accused did not commit the crime, as opposed to a finding that he did. Proving a negative is always substantially more difficult than its opposite. And secondly, it is slanted towards the prosecution because at the time of bail applications, the judge – in the normal course of things – effectively has before her only the Prosecution’s version (the FIR and the chargesheet). The defence may controvert it, but without the opportunity to attack the Prosecution’s case in the course of trial, at the best of times, it will simply be a clash of two rival versions. In such circumstances, a judge can hardly come to a finding against the Prosecution, unless the Prosecution’s own case is so riddled with inconsistencies, that it collapses under its own weight (it has been known to happen).

While such provisions might potentially be defensible in a legal system where trials are completed quickly, in the Indian justice system, where trials take years, they simply cannot be defended. They enable incarceration for years on end (recent UAPA cases have involved people being jailed for over a decade before being found innocent), without any judicial finding of guilt.

What do the Courts do in such a situation? One solution, of course, is to hold that such provisions are unconstitutional (which Nariman J. did, in a judgment on the PMLA). If Courts cannot – or are not willing – to do that, then the other option is to interpret the law narrowly, and in favour of the accused. The logic of this is simple: the more draconian the law, the more it impacts personal liberty, the more cautious a Court must be in interpreting it. A good example of this is the Bombay High Court’s bail judgments in the Kabir Kala Manch cases: Justice Thipsay interpreted the substantive provisions of the UAPA narrowly, in order to hold that on its own materials the Prosecution had failed to make out even a prima facie case for guilt. Consequently, he granted bail even within the constraints of the section.

On a different note, in a series of detailed and brilliant judgments, Justice Rajbir Sehrawat of the Punjab & Haryana High Court – after noting many of the basic arguments highlighted above – held that in certain statutes condition in question would not be “mandatory” for a Court to follow, while in an NDPS case, held the arresting officer to strict compliance of the procedural requirements under the statute – and granted bail as they were not followed.*

The judgments of Thipsay and Sehrawat JJ from the High Courts of Bombay and Punjab & Haryana show that even under draconian statutes, Courts have interpretive wiggle room to ensure that the individual rights are not entirely effaced by the logic of “reasons of State.” It is in that context that the judgment of the Supreme Court in State of Kerala v Rajesh, delivered yesterday, is disconcerting and disappointing. Handed down by a bench of Ajay Rastogi and Indu Malhotra JJ., this judgment swings the pendulum the other way, holding that in NDPS cases, there cannot be a “liberal” approach to bail (it is unclear what is meant by the word “liberal” here). Moreover, the judgment then holds, in paragraph 21:

The expression “reasonable grounds” means something more than prima facie grounds.  It contemplates substantial probable causes for believing that the accused is not guilty of the alleged offence.  The   reasonable   belief   contemplated   in   the   provision requires existence of such facts and circumstances as are sufficient in themselves to justify satisfaction that the accused is not guilty of the alleged offence.

 

In other words, the Court holds that before the Defence has even had a chance to fully controvert the evidence against the accused, there must be a finding of “substantial probably cause” that the accused is innocence. In other words, this finding must – to all effects and purposes – be made on primarily on the basis of materials adduced by the Prosecution. But how is this ever going to happen? If the Prosecution, for example, has listed three witnesses who saw me commit a crime, until the time I can cross-examine them to bring out the inconsistencies in their testimonies – something that happens at an advanced stage in the trial – how can I ever establish “a substantial probable cause” to believe that I am innocent? If the judgment is followed in letter, therefore, it effectively means that granting bail is almost impossible wherever there exists such a section – or at least, requires substantial ingenuity from High Court judges hearing such cases.

I call this form of reasoning “punitive Constitutionalism”, and it has been on the rise at the Supreme Court of late. Punitive constitutionalism operates on the assumption that the Constitution is not a charter of freedom or emancipation, but a document meant to discipline a recalcitrant or troublesome society. “Punitive constitutionalism” makes rights contingent on what the Court considers “good behaviour” (witness the Chief Justice refusing to hear cases until “violent protests” stop), invents new ways to whittle down rights against the State (witness what was done with habeas corpus in Kashmir), views every accused as a potential criminal rather than innocent until proven guilty, and sees procedural safeguards as impediments to an efficient criminal justice system rather than as essential safeguards of liberty. It is, ultimately, a school of thought that is closely allied with the idea of “executive Courts” – i.e., a Court that sees itself as part of a joint project with the Executive (however defined), rather than the institution that stands between the citizen and the Executive.

It remains to be seen whether this drift will continue in the near future.


(Thanks to Gautam Khazanchi for bringing these judgments to my attention.)

Guest Post: Acquitted but not yet Free – the Constitutionality of Section 437-A Cr.P.C.

(This is a guest post by Abhinav Sekhri, cross-posted from The Proof of Guilt blog with permission.)

The Criminal Procedure Code 1973 [Cr.P.C.] was subjected to significant amendments in 2009. The law on arrest was drastically altered following Supreme Court admonition, and victims were given a real foothold in the criminal process for the first time. Amidst all this, a provision was added to the section on Bail in the Cr.P.C.: Section 437-A. What does it say?

Before conclusion of the trial and before disposal of the appeal, the Court trying the offence or the Appellate Court, as the case may be, shall require the accused to execute bail bonds with sureties, to appear before the higher Court as and when such Court issues notice in respect of any appeal or petition filed against the judgment of the respective Court and such bail bonds shall be in force for six months [Section 437-A(1)]

Through this post, I will try and convince the reader that Section 437-A Cr.P.C. is unconstitutional. The post first cursorily explains the concept of bail and engages with the problematic consequences flowing from the text of Section 437-A. It then discusses the origins of the provision, before moving on to argue that it is contrary to Articles 14, 19, and 21. The last section considers that there are two options, reading down Section 437-A or striking it down completely, and I support the latter course.

Understanding Bail and the Text of 437-A
One often comes across “bail” in context of criminal trials and investigations. What does this mean? In such scenarios (and others), where a person is in the crosshairs of the legal system, the law wants to ensure that legal proceedings are not frustrated by persons fleeing the jurisdiction. An obvious way to address this is to arrest everyone. But that is hardly proportionate to the needs of law enforcement and is far too heavy a strain on State resources.

Bail is the answer to this problem. The person is notionally still in the custody of the court and not at liberty, but is not actually in fetters. Note, that as the law would always need a guarantee of personal appearance, all defendants once in the crosshairs of the system are either on bail or in custody. How does it ensure appearance when required? By imposing certain conditions while releasing the person, chief among which is a requirement to appear in court or before the police. Non-compliance with the conditions is met by the threat of arrest, often along with a threat of imposing financial consequences such as forfeiture of property to the State. The financial threat often extends to other persons called “sureties”, who are thus incentivised to ensure the defendant does not flee.

Now, consider the text of Section 437-A Cr.P.C. It is very broad: the court shall require bail bonds, with sureties, before conclusion of trial and disposal of appeal. This throws up a bunch of questions. First, does it mean that the court will not proceed with the trial or appeal before getting such bail bonds? Second, if the court does proceed with the trial / appeal and finds the defendant innocent, would she then remain in custody if she cannot find sureties or comply with the other conditions imposed for bail?

Section 437-A Cr.P.C. allows for both of these eventualities. And it is for this reason that the High Courts of BombayAllahabad, and Himachal Pradesh have clarified that courts within their jurisdiction must not apply the provision in a way that causes either of these results to follow. There will be some states that I have missed, but I am certain that there are many others where no such clarification exists today. Nor has any guidance been issued by the Supreme Court, and so, it is very possible that both of these problematic outcomes are being seen across the country. From here on, this post will focus on the second of the two outcomes: the continued detention of persons acquitted of all charges for their failure to post adequate bail bonds.

The Genesis and Object of Section 437-A
I mentioned that the guidelines issued by certain High Courts curbed certain uses of Section 437-A Cr.P.C. but have not yet explained how they wanted the provision to be applied. The Courts suggested that the provision is a means to ensure that an acquitted person is available to contest any eventual appeal by the State, and so the bail bonds should only be required at the end of a trial before judgment. They also suggest that bail might be given without sureties if an acquitted person cannot find sureties.

The history of Section 437-A supports this reading. Before it was added to the Cr.P.C. in 2009, the only other provision dealing with a need to detain persons pending an appeal against acquittal was Section 390 Cr.P.C. This empowers the appellate court to detain persons pending an appeal against acquittal, if it is convinced that of the threat of them evading the legal process. But in this scheme there still exists a period between the acquittal and appeal when a scheming defendant could still flee and frustrate the State’s appeal. Taking note of this (and abortive attempts by the Gujarat High Court to fill the gap) the Law Commission in Report No. 154 of 1996 recommended insertion of a Section 437-A Cr.P.C. Why? It said that the Cr.P.C. was “silent on the point of securing attendance” during an appeal, and there had been instances where appeals against acquittals were delayed or dismissed due to this failure in securing attendance.

Two problems are immediately apparent here. First, the Law Commission said that the Cr.P.C. was “silent” on securing attendance for appeals but did not even look at Section 390 Cr.P.C. which did cater to this need, albeit differently. Second, the bogey of appeals against acquittals being dismissed was raised without any empirical data about how many such dismissals happened and why. For instance, if the prosecution filed an appeal years after acquittal (as it often does) and then failed to find the original defendant, then it is rather unreasonable to claim that a person fled or frustrated the appeal and piggyback on the dismissal of the appeal to create a perceived need for Section 437-A

Importantly, the Law Commission acknowledged that this measure might be challenged under Article 21 of the Constitution. Its basis for concluding that the proposal was constitutional was simple: the Cr.P.C. allowed appeals against acquittals, and so seeking bail bonds till the limitation period for filing an appeal subsisted was not a “restraint” on personal freedom. Further, proposed Section 437-A was eminently reasonable where it involved no “restriction of liberty or his freedom of movement”.

When Section 437-A Cr.P.C. was ultimately passed by Parliament, there were two big changes from the suggested draft in Report No. 154. The final version of Section 437-A said that a court shall require bonds while the draft version did not make it a mandatory requirement. At the same time, the final version only needed the bail bonds for six months, down from the one year period that the Law Commission had suggested.

Unconstitutional Fetters on Personal Liberty
Practically, Section 437-A Cr.P.C. does not make much of a difference to defendants already out on bail during trial. In such situations, it is easy to extend the period of that bail bond for six months after acquittal by changing the form of the bail bond. But Section 437-A works very differently for those defendants who are in actual custody, unable to post bail by finding sureties or complying with any financial conditions that a court might impose. It is only for these persons that the two scenarios highlighted earlier – delayed trial and delayed effect of acquittal – are possibly realised.

On the face of it, Section 437-A Cr.P.C. though seemingly neutral, is very selective in its impact and discriminates against one class of persons. The classification that it effects is purely built on levers of wealth, influence, and privilege, rather than pursuit of the object behind Section 437-A (perhaps they indirectly affect that object, at best). The effect of this discrimination is to deprive such persons of their right to life under Article 21 of the Constitution, by not only possibly denying a trial itself, but more importantly, by not allowing them to enjoy the liberty that is the natural concomitant of an acquittal. Thus, Section 437-A in its present form offends the equality guarantee of Article 14.

Actual confinement of a citizen after acquittal obviously curtails the freedom of movement that she is guaranteed under Article 19(1)(d). But both the legal and actual fetters on personal liberty curtail the fundamental right that Article 21 protects. Which means we must consider whether Section 437-A Cr.P.C. is protected by the tests governing restrictions of these fundamental rights.

Section 437-A Cr.P.C. contains no sense of proportionality. It does not require the State to satisfy a court that an acquitted persons might possibly flee to avoid the appeal. Nor does it require the State to show that a person, if immediately released, will pose a threat to public safety. Instead, the provision demands every acquitted person to remain in custody of the court despite till the State can make up its mind about pursuing an appeal. Thus, the rights under Articles 19(1)(d) and 21 are rendered subservient to administrative convenience, pure and simple. And this after a person is declared “not guilty”, after a full-length trial or appeal.

Reading Down vs. Striking Down
No wonder those High Courts which have recommended that personal bonds be taken are effectively reading down the text of the provision (supported in this paper too, which discusses other issues with the provision). The constitutional problems in giving Section 437-A Cr.P.C. its fullest expression are obvious, and even the Law Commission in 2017 also suggested a relook is now necessary. But here, I argue that reading down cannot save the provision and it must be struck down altogether.

First, a question of means. Saving Section 437-A Cr.P.C. does not involve merely filling in gaps or creatively interpreting the text. It involves actively re-writing it, and that is something courts cannot do. The requirement that a court “shall” take bonds with sureties will be re-written as something it “may” do. Further, the scope of discretion is altered not to allow a court to forego the demand for bail bonds altogether, but to insert the words “personal bond” in Section 437-A to change the kind of bail bonds that are required.

But far more important is the question of principles. Section 437-A Cr.P.C. must be struck down for it automatically breaks the link between a judgment of acquittal and its legal effects, in the absence of any appeal preferred by the State. This is perhaps the most problematic part of the provision At one level, it creates a conflict within the Cr.P.C. Today, Section 354(1)(d) Cr.P.C. still states that a judgment of acquittal requires that a court direct the person be set at liberty. Without amending what it means to be acquitted directly, the legislature has indirectly rendered all acquittals subject to a condition of complying with Section 437-A. Can the legislature indirectly alter the very meaning of an acquittal at all stages within our criminal justice system in this indirect fashion? No, it cannot. Because this link between a judgment of acquittal and being set at liberty is protected through Article 21 itself. It cannot be severed, and certainly not to cater to administrative convenience.

Conclusions
Perhaps I am “fetishising” what an acquittal means – after all, it is not final till confirmed in appeal. But that finality is in respect of an acquittal being legally unassailable. It does not make the effects of an acquittal automatically contingent upon the possibility of appeal proceedings. Rather, not treating an acquittal as final allows an appellate court to delay giving it effect. This delay can only occur after giving a full hearing to both sides. Any other position would deprive the verdict of a lower court of all sanctity lest it be confirmed in appeal. Moreover, it would mean that persons are condemned from the date of arrest till their case is resolved by the highestappellate court, and continue to suffer all the collateral consequences of criminal convictions for this unconscionably long period of time as well.

This is why the remedy provided by Section 390 Cr.P.C. makes sense. The state can seek detention of the acquitted person pending appeal if it can show that it is necessary, but the default is still that a person remains at liberty. Moreover, an appeal having been filed gave the court proper jurisdiction to hold someone in custody. What if, after some empirical study, it is found that something like Section 437-A is necessary to prevent persons from fleeing and frustrating appeals? Then, a hearing similar to that under Section 390 is the answer, not a position that changes the default position. If the trial court / appellate court is satisfied of a need to detain after having heard both sides it could pass appropriate orders, with the denial of liberty narrowly tailored to account for how long the State might take to file an appeal rather than simply hold persons in custody for fixed periods.

Under no situation is Section 437-A Cr.P.C. the answer. Parliament cannot pass statutes that deem an entire population to be a suspect class for administrative convenience, even after a court of proper jurisdiction has pronounced them innocent. If this is so, then the guarantee under Article 21 might soon be no better than a fig leaf.

Guest Post: Bail Provisions of Section 45 PMLA Struck Down – Some Hits and Misses

(This is a guest post by Abhinav Sekhri, which first appeared on the Proof of Guilt blog)

Two days ago, a Two Judges’ Bench of the Indian Supreme Court decided a batch of writ petitions led by Writ Petition (Crl) No. 67 of 2017 titled Nikesh Tarachand Shah v. Union of India & Anr. [Nikesh Shah] in which it struck down the parts of Section 45 of the Prevention of Money Laundering Act 2002 [PMLA] which concerned the grant of bail. The Court held that these parts violated Articles 14 and 21 of the Constitution – guaranteeing a right to equality, and protection against deprivation of the right to life and personal liberty by a procedure not established by law. The effect of this judgment is that bail petitions earlier subject to a stringent standard under Section 45 PMLA will now be tested on the less taxing standards of Sections 439 and 437 of the Criminal Procedure Code 1973 [Cr.P.C.]. This post has four parts – (i) explaining how money laundering and the PMLA work (which I’d urge you to skim through even if you’re a lawyer, because at times the judgment reflects some lack of knowledge on the Court’s part), (ii) charting out how the Court did what it did, (iii) showing where the Court goes wrong, and finally (iv) what this judgment might mean for the many other statutes with similar clauses that have not been examined by the Court yet.

What is the PMLA, What are the Schedules, and What does Section 45 do?

The PMLA is India’s answer to its global commitment to tackle money laundering, which (at the cost of oversimplification) means representing assets obtained through illegal acts as untainted. In line with global standards, the PMLA covers all kinds of conduct connected with this process of representing black as white (doing, aiding, abetting, attempting etc), as long as one knowingly did so [Sections 3 and 4]. The PMLA not only makes this is an offence but also triggers connected civil actions of attaching and confiscating the tainted assets themselves [Sections 5-8].

Notice how the entire idea of money laundering is linked to some underlying illegal act which results in generating some proceeds – cash or kind. While some countries don’t require that illegal act to be a crime, India does, and the PMLA calls it a ‘Scheduled Offence’ [Section 2(y)] i.e. offences that are part of the Schedules to the PMLA. There are three Schedules – A, B, and C – and Schedule A contains the bulk of offences and Schedule C is basically the same thing applied in a transnational context. Schedule B contains only one offence – Section 132 of the Customs Act 1962 which criminalises making false declarations before customs officers. Importantly, when the underlying offence is one from Schedule B, the PMLA will only apply if the allegations involve a value of at least one crore rupees. There is no such minimum monetary limit for cases with Schedule A offences. It wasn’t always like this, and the history behind these Schedules became quite important in Nikesh Shah which requires me to discuss it here.

When the PMLA came into force in 2005, Schedule A only had two paragraphs carrying offences punishable under the Indian Penal Code 1860 [IPC] for waging war against India and nine offences from the Narcotics, Drugs and Psychotropic Substances Act 1985 [NDPS Act]. Schedule B contained the bulk of offences, along with a lower minimum threshold of thirty lakhs for the value of allegations. Then around 2010 India wanted to join the Financial Action Task Force [FATF] as a member. The FATF is a global body created by the G-8 for money laundering and membership is a big deal [India is the only South Asian member state till today]. When the FATF conducted its evaluation of Indian money laundering laws, it heavily criticised the monetary limit for the cases in Schedule B [paragraph 167 of the linked report]. The logic was that the limit would allow money laundering to escape under the radar as people would just deal in smaller tranches over a slightly elongated period of time. So the FATF recommended the limit be abolished [paragraph 175]. The government sought to do this by simply moving all Schedule B offences to Schedule A, which was done through the 2013 Amendment, leaving Schedule B empty for the time being.

In all this moving around offences, nobody thought fit to look at what impact it would have on the rest of the PMLA – specifically, on Section 45 which spoke about bail. Since the money laundering offence was tied to the Scheduled Offence, Section 45(1) looked at that underlying offence and this decided how difficult it would be to get bail. If it was a Schedule A offence with a sentence of more than three years, the law placed two additional conditions for getting bail: (i) the public prosecutor had to be given a chance to oppose bail, and if the prosecutor chose to oppose bail, then (ii) the court had to satisfy itself that the defendant was “not guilty of such offence” and was not likely to commit any offence on bail, and the burden fell on the defendant to satisfy the court. For all other Schedule A offences, and all Schedule B offences, the regular bail clauses from the Cr.P.C. continued to apply. You can see how the 2013 amendments to the Schedules completely changed the look of Section 45 – the exceptional process became the norm. This new normal was under challenge before the Supreme Court in Nikesh Shah.

SC on Section 45 – Violates Articles 14 and 21
Petitioners argued that the constitutional protections of Articles 14 and 21 were violated by Section 45 PMLA, and the Court agreed to both contentions. Rather than address arguments first and then move to the Court’s appreciation, I discuss both together for brevity.

Article 14
The Petitioners argued that linking the stringent bail clauses to offences in Schedule A that carried at least a three year maximum sentence was creating several irrational and arbitrary classifications which the Court encapsulated through examples [Paragraphs 24-27, and 35]. The Court found no basis to differentiate the harsh treatment meted out under Section 45 from the following hypothetical cases which according to the Court did not attract Section 45:

  • When there is only the PMLA charge as the trial for the Scheduled offence was complete;
  • When the PMLA allegation is based on a Schedule B offence;
  • When the PMLA allegation is based on a Schedule A offence carrying a maximum sentence below three years;
  • When a person is tried for a Part A offence with at least a three year term (versus a joint trial where the same person is tried together with the person with PMLA charges);
  • When the person is released on Anticipatory Bail under Section 438 Cr.P.C. for allegations of the Scheduled Offence, before the PMLA charge was brought in.

The Court was of the view that the seriousness of money laundering cases depended on the amount of money involved [Paragraphs 29-30]. Since Schedule A had no monetary limits, the Court concluded that the likelihood of being granted bail was being significantly affected under Section 45 by factors that had nothing to do with allegations of money laundering [Paragraphs 26-27]. When the Attorney General attempted to defend the scheme by painting the classification as a punishment-based one, the Court easily rebuffed his argument. First, the Court suggested there was no such scheme, but noted that even then, the idea should have something to do with the object of the PMLA. The Court showed how Schedule A had many offences that didn’t seem related to money laundering [taking particular objection in Paragraph 34 to offences under the National Biodiversity Act being there], leaving out others that might have more rational connections to money laundering such as counterfeiting currency [Paragraphs 29-30]. The Court also adversely commented on how Schedule A had lumped different NDPS offences together, at the cost of ignoring how the parent Act treated those offences differently [Paragraph 32-33].

The Court noted also that Section 45 of the PMLA was different from other laws that carried similar requirements such as Section 20(8) of the Terrorism and Disruptive Activities (Prevention) Act 1987 [TADA]. The ‘such offence’ in TADA required a court to be satisfied that the defendant was not guilty of the TADA offence in question before granting bail. But in the PMLA, ‘such’ offence referred to the Scheduled Offence instead of the PMLA offence. So, the restrictions imposed by Section 45 PMLA were held to have no connection to the objects of the PMLA itself and thus the rational classification, if any, violated Article 14 [Paragraph 28].

Article 21

The Petitioners argued that requiring defendants to satisfy the court that they were not guilty of ‘such’ offence violated Article 21 by reversing the presumption of innocence and required the defendant to disclose her defence at the outset of the case. In the judgment the Court doesn’t really address Article 21 independently – instead the Court suggest that because the provision violates Article 14 it cannot be ‘procedure established by law’ and therefore violated Article 21. Towards the end of the decision the Court begins discussing the argument though. It labels Section 45 a “drastic provision which turns on its head the presumption of innocence which is fundamental to a person accused of any offence.” [Paragraph 38]. In the same paragraph it goes on to observe that “before application of a section which makes drastic inroads into the fundamental right of personal liberty guaranteed by Article 21 of the Constitution of India, we must be doubly sure that such provision furthers a compelling state interest for tackling serious crime. Absent any such compelling state interest, the indiscriminate application of section 45 will certainly violate Article 21 of the Constitution. Provisions akin to section 45 have only been upheld on the ground that there is a compelling state interest in tackling crimes of an extremely heinous nature.”

Hits and Misses
There are two questions that were at stake here: (i) did any part of Section 45 offend the Constitution, and if so, (ii) did the Court have no other option but to repeal the provision. Reading the decision, it seems like the Court felt there was so much wrong in the PMLA scheme it decided to throw the kitchen sink at one point rather than explain the issues. The Court answered both affirmatively but never explained to us whether any argument dispositive, or does every case need this sort of broad argumentation to succeed.

Classification and Article 14 first. After reading the legislative history behind the 2013 amendments and the FATF argument, do you think that the Court is right in concluding that higher the monetary allegations, more serious the PMLA case? I’m not so sure. Nor do I think there is much to be gained by placing emphasis (like the Court does) on how Schedule B today has a higher limit than the initial thirty lakhs to suggest that this is in fact the case. It is far more plausible that the one crore limit was placed keeping in mind the underlying offence (false declarations to customs officials in an enquiry) and the concerns of the export industry, which is already subject to Schedule A through Section 135 of the Customs Act 1962 (evading customs duty). Rather than attempt at answering what might be the basis for such a classification for the PMLA (and indirectly giving hints to the government on what might pass muster), the Court would have done well by restricting itself to answering whether the present classification between (i) PMLA allegations based on a type of Schedule A offences versus (ii) all other PMLA cases was intelligible and connected to the objects of the PMLA. As there was enough to show that the original intent (if any) behind Section 45 had not kept apace with the subsequent amendments to the Schedules in 2013, the Court could strike down this classification. But did that require striking down the whole clause?

This brings us to the other part of what did that classification achieve. If it sought to serve as a filter for PMLA cases when it came to administering a strict bail clause, we are left with no filter. Does that mean no PMLA case is serious enough to warrant an application of the clause, or will the clause apply to every PMLA case? Deciding this would need the Court to decide whether clauses such as Section 45 that required a court to find defendants ‘not guilty’ at the bail stage were constitutional. Rather than directly address this, the Court turned to how the text of Section 45 was flawed, as it referred back to the Scheduled Offence on deciding bail petitions. Since the scheduled classification had been struck down, there was nothing to refer to, and so the clause had to go. While there is little to fault this approach, I remain unconvinced that the Court had no option but to strike down the clause because of the text. The Court has performed far greater feats of legislative reconstruction than being asked to read ‘such offence’ in Section 45 PMLA as referring to the PMLA allegations rather than only the Scheduled Offence. After all, it stands to reason that a bail provision in the PMLA would want a PMLA special court to consider the PMLA allegations. In fact, many High Court decisions show this is how they were doing it. Heck, this is how the Court itself was doing it in Rohit Tandon at the start of November [Paragraphs 21-23 of the link]. I think this course was adopted as it helped secure two objectives. Not only did this take care of the PMLA clause which this bench of the Court clearly did not like much, it also helped to protect other statutes with similar clauses which the Court held met a ‘compelling state interest’ test.

This brings me to one last bit about Article 21 and the Eighth Amendment of the U.S. Constitution. The Court cites a previous decision in Rajesh Kumar v. State (NCT) of Delhi [(2011) 13 SCC 706] for the proposition that Article 21 of the Indian Constitution has incorporated the Eighth Amendment and its protection against excessive bail [Paragraphs 13, 19 of Nikesh Shah]. The Court also cites two American decisions [Paragraph 37] on bail for good measure. This is, unfortunately, wrong. Rajesh Kumar cited previous precedent in Sunil Batra to suggest that even though India did not have the Eighth Amendment or the ‘Due Process’ clause, the consequences were the same to prevent cruel and unusual punishment. Not only did both those decisions not mention the excessive bail clause, the references to the cruel and unusual punishment clause itself are highly contentious as an earlier bench of the Supreme Court had held it couldn’t be pressed in India, and that decision continues to be cited.

Conclusion

The slapdash manner in which the PMLA Schedules were amended in 2013 to appease the FATF had already caused some High Courts to address this issue of Section 45. The closest it came to striking down the clause was the Punjab & Haryana High Court’s decision in Gorav Kathuria v. Union of India & Anr. where it held the bail provisions would not apply retrospectively to offences previously in Schedule B [Paragraphs 43-45 of Nikesh Shah]. When the Court declined to hear an appeal against the High Court order in Kathuria I thought that it had indirectly affirmed the validity of Section 45. The judgment in Nikesh Shah comes as a surprise, and marks the first occasion when the Court has looked at any part of the PMLA through a constitutional lens. There are other parts that are equally problematic – the asset forfeiture scheme and the compulsion on witnesses to make truthful declarations, for instance – that litigants may take to the Court being encouraged by this judgment.

As for the future impact of Nikesh Shah on other statutes that carry the same ‘drastic provision’, the stage is set for some litigation on that front as well. The Supreme Court has only approved of the TADA and the MCOCA provisions in the past, leaving the many others open to scrutiny on this new test of whether the provision furthers a ‘compelling state interest’. The Court never answered that for the PMLA context while deciding the petitions in Nikesh Shah. Do you think it might conclude that the PMLA does not meet the test? What about the other statutes? I’ve re-pasted my list of statutes containing the clauses below after accounting for the ones that are not relevant anymore. Comments, as always, are welcome.

  1. Section 437(1), Cr.P.C. (in cases of death and life imprisonment).
  2. Section 12AA (inserted in 1981), of the Essential Commodities Act, 1955.
  3. States of Punjab and Tripura inserted this provision as Section 439-A to the Cr.P.C. so applicable within their territory, in 1983 and 1993 respectively. This restricted bail to persons accused of certain offences, inter alia Section 121, 124-A IPC.
  4. Section 37 (amended in 1989) of the Narcotic Drugs and Psychotropic Substances Act 1985 [NDPS].
  5. Section 7A (inserted in 1994) of the Anti-Hijacking Act, 1982.
  6. Section 6A (inserted in 1994) of the Suppression of Unlawful Acts against Safety of Civil Aviation Act 1982.
  7. Section 8 of the Suppression of Unlawful Acts against Safety of Maritime Navigation and Fixed Platforms on Continental Shelf Act 2002.
  8. Section 51A (inserted in 2002) of the Wildlife Protection Act, 1972.
  9. Section 43D (inserted in 2008) of the Unlawful Activities Prevention Act, 1967 [UAPA] (nearly identical).
  10. Section 36AC (inserted in 2008) of the Drugs and Cosmetics Act, 1940.