Guest Post: How Long is Too Long? - On the Maximum Period that an Undertrial Prisoner can be Detained

(This is a guest post by Hany Babu and Surendra Gadling, who have been detained in prison as 'undertrials' since 2018 and 2020, respectively. This piece is being published simultaneously with the Indian Constitutional Law and Philosophy Blog)

Dedicated to the fellow undertrial prisoners languishing in the prisons of India with the hope that the system would sooner than later wake up to their plight; and that one day prisons, if not still obsolete, would at least have become places where those proven guilty are held, and not places that incarcerate those who are presumed to be innocent.

The primary author would like to thank his co-defendants, Arun Fereira and Vernon Gonsalves, for the enriching discussions – stolen at times in the corridors of the court, at times on the drives to and from the court, and at times in the ‘After’ hours; Vernon, for his characteristically incisive remark ‘You need to think more!’ after going through – what I now realise was – a much cruder earlier version of this piece even before BNSS had kicked in; and Arun, whose name may have been among the authors had their Lordships not set him free, for owning the idea that the Code should have more provisions for statutory bail. Needless to say, neither of them is to shoulder the shortcomings of this piece.

Introduction

“How long is too long a period of incarceration as an undertrial for a court to conclude the right of speedy trial is defeated?”

This was a question raised by the Delhi High Court in Mohamed Hakim v. State (NCT of Delhi) 2021 SCC OnLine Del 4623, in the context of grant of bail for an undertrial prisoner. We raise a similar question regarding the maximum period of detention permissible by law for an undertrial prisoner. Section 479 of the Bharatiya Nagarik Suraksha Sanhita, 2023 (BNSS), the Section that deals with the question of the maximum period that an undertrial can be detained, largely reproduces the language of its predecessor, Section 436-A of the Code of Criminal Procedure, 1973 (CrPC), the clause that no person should be imprisoned for a period exceeding the maximum period of punishment specified for his offence.

This essay argues that the existing law, which allows a person to be detained for a period equal to the maximum period of imprisonment specified as punishment for his offence only serves to legitimise the award of “surrogate punishment” equivalent to the maximum punishment to that person, without conducting trial and determination of the question of guilt or innocence of that person according to procedure established by law. Such a law has no place in the statute book if presumption of innocence is one of the cornerstones of criminal jurisprudence.

Further, if “bail, not jail” were followed in letter and spirit, the question of prolonged pre-trial detention would not arise. In practice, however, the combination of the question of ‘prima facie guilt’ – another principle that goes against the presumption of innocence – and the tripod test of the possibility of the accused tampering with the evidence, influencing witness and / or evading trial, together conspire against the grant of bail. In such circumstances, it is only those who can afford to hire expensive lawyers and knock the doors of the Constitutional Courts who can manage to secure bail. For the rest jail becomes the rule.

In such a scenario an act of balancing is called for. We argue for a judicial reading of the BNSS that will provide us with the framework of finding such a balance.

We set out by drawing a comparison between Section 479 of BNSS and its predecessor Section 436-A of CrPC. It can be seen that far from being an improvement on Section 436-A of CrPC with respect to reducing the period that an undertrial prisoner can be detained, Section 479 of BNSS is more restrictive in its application. Be that as it may, what makes Section 479 of BNSS problematic are certain anomalies we discuss below. We shall first take a look at Section 436-A of CrPC and then go on to Section 479 of BNSS before putting forth our suggestions regarding the maximum period of detention for an undertrial prisoner.

Section 436-A of CrPC

Section 436-A of CrPC was inserted in 2006, as it was found that there were undertrial prisoners detained in jails for periods beyond the maximum period of imprisonment provided for the alleged offence. It may be noted that in spite of this, undertrial prisoners are still languishing in prisons for periods longer than the maximum punishment specified for their offence, as in the case of Dipak Joshi in West Bengal, who came out of prison after 42 years following the intervention of the Calcutta High Court.

Apart from the stipulation that an undertrial prisoner cannot be held in prison for a period exceeding the maximum period of punishment for his offence, Section 436-A of CrPC also has a provision that a person who has undergone detention for a period extending up to one half of the maximum period of imprisonment specified as punishment for his offence may be released if the punishment of death is not one of the punishments specified for the offence in question. In the prison parlance, this provision is referred to as ‘half ground’, and we shall use this term to refer to this clause. Unlike the maximum period of detention, the provision of half ground was not a statutory obligation and was subject to the Court’s discretion after hearing the Public Prosecutor.

In this article we shall be limiting our discussion to offences under the Indian Penal Code, 1860 (IPC) while discussing CrPC and offences under the Bharatiya Nyaya Sanhita, 2023 (BNS) while discussing BNSS. There are twelve offences in IPC for which death is specified as one of the punishments, and these offences were outside the purview of operation of 436-A for the purpose of half ground.

Section 479 of BNSS

Section 479 of BNSS corresponds to Section 436-A of CrPC. Unlike in the case of most sections, where BNSS just copies the CrPC provisions, in the case of Section 479 there are a few changes and considerable additions. There are three sub-sections to Section 479 of BNSS. While sub-section (1) reproduces the contents of Section 436-A of CrPC with some slight modifications and additions, sub-sections (2) and (3) are entirely new. We shall look at each of the sub-sections one by one.

Sub-section (1) of Section 479 of BNSS

In sub-section (1) of Section 479 of BNSS, offences punishable with death and imprisonment for life are mentioned to be excluded from the “half ground.” On our count, there are as many as 75 offences under BNS that have death or imprisonment for life as one of the punishments. While Section 436-A of CrPC excludes only 12 offences of the IPC from the ambit of half ground, BNSS excludes 75 offences from its ambit. If only offences punishable with death were excluded, by our count 16 offences of BNS would have been excluded from half ground.

As per Section 6 of BNS (which is almost a replica of Section 57 IPC), imprisonment for life is to be reckoned as equivalent to imprisonment for a period of twenty years for the purpose of calculating fractions of terms of punishment. So, for the purpose of “half ground,” a person who has been detained for an offence for which imprisonment for life is (and death is not) one of the punishments specified under the relevant law, that person would ordinarily have been eligible to be released on “half ground” after ten years in custody. However, that door is closed for such people under BNSS.

If BNSS takes a tougher view of offences punishable with imprisonment for life, as it considers them to be offences of a serious nature and wants to show no leniency towards persons accused of such crimes, there is hardly any room for argument. However, a look at Section 474 of BNSS (which corresponds to Section 433 of CrPC), which deals with the power of the government to commute sentences, would make us doubt this. As per clause (a) of Section 474 of BNSS, imprisonment for life can be commuted to imprisonment for a term of not less than seven years. That is to say, if you are convicted for imprisonment for life, you may stand a chance of being freed after spending seven years in prison provided you are in the good books of the government of the day; however, for the reasons set out above, if you are detained as an undertrial prisoner you are not eligible to be released even after spending ten years in prison.

So, rather than treating serious offences with iron gloves, BNSS seems to leave it to the whims of the executive rather than to the wisdom of the judiciary. What this implies for hapless prisoners who cannot afford to engage lawyers to file bail applications on their behalf is that pleading guilty and begging for the mercy of the executive in the form of commutation to a term of seven years of imprisonment would be a more attractive option than languishing endlessly in prison as an undertrial prisoner.

Apart from the “half ground,” BNSS introduces what we shall call ‘one-third ground’ for first time offenders. A first-time offender can be released if he has undergone detention for a period extending up to one-third period of the maximum punishment specified for his offence under the relevant law. The parenthetical clause explains that a first-time offender is a person “who has never been convicted for any offence in the past.” Let us not overlook the inappropriacy in using the term ‘offender’ for a person yet to be proven guilty. The possibility of getting bail after having spent one-third of the maximum period of punishment raised the hopes of so many prisoners who have been behind the bars for years. Their hopes were further raised with the possibility of the retrospective application of Section 479 of BNSS following the ruling of the Supreme Court in W.P. (Civil) 406 / 2013, In Re Inhuman Conditions in 1382 Prisons. It is not surprising that they missed the fine print that said offences that have imprisonment for life as one of the punishments are excluded from the ambit of Section 479 of BNSS. (Ed Note: A similar conclusion was made on the Proof of Guilt Blog here).

Sub-section (2): More than one offence and multiple cases

Sub-section (2) of Section 479 of BNSS is a non-obstante clause that says a person who has more than one offence or multiple cases pending against him is not eligible to be considered to be released, neither on the basis of one-third ground, nor on the basis of half ground. Section 436-A of CrPC had no such restriction.

“Offence” is defined in Section 2(q) of BNSS (a replica of Section 2(n) of CrPC). It says “any act or omission made punishable by any law for the time being in force…” As per sub-section (1) of Section 243 of BNSS (equivalent to Section 220 of CrPC), a series of acts connected so as to form the same transaction can lead to more offences than one being committed. Sub-section (3) of Section 243 of BNSS states that if an offence falls within two or more separate definitions by which offences are defined under any law, the person may be charged with each such offence.

Illustration (1) pertaining to sub-section (3) of Section 243 of BNSS says if a person wrongfully strikes another person with a cane, he may be separately charged with and convicted of offences under sub-section (2) of Section 115 (i.e. voluntarily causing hurt) and Section 131 (i.e. punishment for assault or criminal force other than on grave provocation) of BNS. This would mean that on a literal reading of Section 479 of BNSS, a person who strikes another with a cane would be ineligible to be considered to be released on the basis of one-third or half ground even if the person is a ‘first time offender’.

A combined reading of sub-sections (1) and (2) of Section 479 BNSS gives us a peculiar situation. The condition in sub-section (1) that a person should be a first-time offender is applicable only for the one-third ground and not for the half ground. Or, in other words, if you have a previous conviction, you cannot be considered to be released on the basis of one-third ground, but you can be released on the basis of half ground. Well and good. Now consider this: as per sub-section (2) of Section 479 BNSS, a person who has multiple cases pending against him (or even more than one offence against him in a single case) cannot be considered either for one-third ground or for half ground. That is to say, BNSS takes a more considerate view of a person who has a previous conviction than of a person who is yet to be proven guilty but has more than one case pending against him in that the former is eligible to be released on half ground, but not the latter. Does this not eviscerate the presumption of innocence?

Sub-Section 3: A silver lining

Sub-section (3) of Section 479 of BNSS is indeed a welcome addition as it casts a duty on the Jail Superintendent to make an application to the Court to release a person who has completed one half or one-third of the period mentioned in sub-section (1), as the case may be. Under the previous scheme, it was left to the accused to move court. Often the persons were hardly aware of such a provision or would not have the legal support to make such an application.

Applying the Law

Having gone through Section 479 of BNSS, which deals with the maximum period that an undertrial prisoner can be detained, we shall turn to two related issues: (1) How to reduce the number of undertrial prisoners in the prisons? and (2) How to ensure that an undertrial prisoner does not get incarcerated for too long a period? Albeit related, these two questions are distinct, and different steps need to be taken to tackle them effectively.

The ABC of reducing the number of undertrial prisoners

If the number of undertrial persons in the prisons is to be reduced, the three necessary steps are:

A: Avoid arrest unless necessary;

B: Bail to be granted unless circumstances warrant otherwise;

C: Conduct speedy trial.

In spite of the repeated pronouncements of the Supreme Court about the need for the police to be judicious in the use of the power to arrest, BNSS makes only a feeble attempt to strengthen the safeguards offered by CrPC. Sections 41, 41A, 41B, 41C and 41D of CrPC, which deal with arrest, are reproduced and renumbered as Sections 35 to 38 of BNSS. The only additional protection BNSS offers – and what perhaps justifies renaming of the Act as “Nagarik Suraksha” – is the addition of sub-section (7) of Section 35, which stipulates that prior permission of an officer not below the rank of Deputy Superintendent of Police is needed to arrest a person who is infirm or above sixty years of age, for an offence punishable with less than three years of imprisonment.

There are at least more than one hundred offences in BNS (just as was the case with IPC) that have a punishment of imprisonment for a maximum period of less than three years. However, barring roughly half a dozen of the rest are all bailable and non-cognizable offences. Therefore, the introduction of sub-section (7) of Section 35 of BNSS may hardly make a practical difference. Moreover, experience shows that in order to circumvent provisions like those of Section 35(7), police may easily add a Section that provides for punishment of three years or more and may also get away with it.

Going into the issue of grant of bail is beyond the scope of this piece. Some of the recent judgements of the Supreme Court that have upheld the right to life of the undertrial prisoner and granted bail upholding Article 21 as an effective means against the prolonged incarceration and award of surrogate punishment seem to offer a beacon of hope. However, as noted by Abhinav Sekhri in his blog post [‘A curious Constitutionalism’ dated July 19, 2024], it is still as though prolonged incarceration of an undertrial prisoner and the issue of grant of bail are issues only the highest Court has to concern itself with. For those who cannot afford to reach that level, there still isn’t much scope for hope.

The culture of seeking adjournments and postponements are usually cited as the main causes of delay in judicial proceedings. Section 309 of CrPC that dealt with powers of the Court to postpone or adjourn proceedings presents a curious case. The Section said that proceedings in every trial or enquiry should be conducted on a “day-to-day” basis and postponement of the commencement of trial or adjournment beyond the following day is to be resorted to only if necessary, and the reasons for the same have to be recorded. However, postponement of the commencement of the trial for months and years and adjournments of proceedings have become the norm rather than exception. There is also a stipulation that trial in offences related to rape should be completed within a period of two months from the date of filing of the chargesheet, which would be a miracle in our trial courts.

Apart from the time limit of 60 / 90 days for the completion of investigation, CrPC had no time limit specified for the framing of charges, commencement and completion of trial, or the delivery of the judgment. BNSS, on the other hand, sets a time limit of “sixty days from the date of first hearing on charge” for framing of charges in Sections 251 and 263 (corresponds to Sections 228 and 242 of CrPC, respectively) and a period of forty five days for pronouncement of judgment after conclusion of the trial in Section 392 (which corresponds to Section 353 of CrPC). The efficacy of these steps, however, will have to be tested in practice as no remedies are prescribed in the eventuality of these limits not being observed. There is hardly any reason to assume that the time limits introduced in BNSS will not remain a dead letter, like the clause about “day-to-day” trial and the sixty day limit for the completion of trial in offences related to rape, which were part of Section 309 of CrPC; or that Section 346 of BNSS will ensure framing of charges and commencement of trial without undue delay. It has to be underlined that no time limit has been specified as to when “hearing on charge” should commence after the filing of the chargesheet.

In reality, as observed by the Bombay High Court while granting bail to the accused in a POCSO case in which charges were not framed even after five years, once the chargesheet has been filed the case is often kept on the back burner and the investigation agency and the court forgets all about the accused after remanding him to judicial custody. The remand gets mechanically extended from time to time in most cases, without the accused person being produced before the court either physically or through video conferencing. This process goes on for months or even years. In order to curb this, effective checks should be built into the system similar to the statutory bail granted to the accused in case investigation has not been completed within the stipulated time. We turn to this issue in the next section.

How Long is too Long

As we argued right at the outset, a law that states no undertrial prisoner shall be detained for a period exceeding the maximum period of imprisonment specified as punishment for his offence is an anathema to the principle of presumption of innocence, as it only legitimises the detention of the undertrial prisoner to a period equal to that of the maximum period of imprisonment specified as punishment for that offence. It thereby provides for inflicting surrogate punishment on those who are considered to be innocent until proven guilty by procedure established by law. The question, then, is how long a period of incarceration is too long for an undertrial prisoner?

In an ideal world, no person should be punished until proven guilty. And without mincing words let us admit that the label ‘judicial custody’ does not make detaining an undertrial person within the confines of a prisonhouse qualitatively very different from punishment. Nevertheless, setting idealism aside, we may still arrive at a reasonable period for which an undertrial prisoner can be detained by confining ourselves to the four corners of BNSS. In order to arrive at such a figure, we need to turn our attention to another Section of BNSS – Section 474 (which corresponds to Section 433 of CrPC) that deals with the power of the government to commute sentences.

Unlike most other Sections, Section 474 of BNSS is a considerable improvement upon its counterpart, namely Section 433 of CrPC. Sentences can be commuted as given below as per Section 433 of CrPC:

• death sentence to imprisonment for life;
• imprisonment for life to imprisonment for a term not exceeding fourteen years;
• rigorous imprisonment to simple imprisonment; and
• simple imprisonment to fine.

Section 474 of BNSS offers the following possibilities for the commutation of sentence:

• death sentence to imprisonment for life;
• imprisonment for life to imprisonment for a term not less than seven years;
• imprisonment for seven years or more to imprisonment for a term not less than three years;
• imprisonment for a term less than seven years to fine; and
• rigorous imprisonment to simple imprisonment.

As is evident from a glance at the above, BNSS takes a much more liberal view with respect to conviction, in cases where the power to show leniency and commute sentences is vested in the hands of the executive. If we consider the arithmetic behind the calculation involved in arriving at the number of years to which terms of sentences of imprisonment are commuted, we can see that it hovers around one third of the sentence. For instance, imprisonment for life is reckoned as equivalent to imprisonment for twenty years for the purpose of calculating fractions of terms of punishments (as per Section 6 of BNS / Section 57 of IPC). So, one third of imprisonment for life is six years and eight months. Similarly, one third of seven years is two years and four months. If after conviction, the term of sentence can be commuted to its one third at the sweet will of the executive, we propose that the same formula can be adopted in fixing the maximum period for which an undertrial prisoner can be detained.

We propose that seven years be fixed as the upper limit beyond which a person in custody as an undertrial for an offence punishable with imprisonment for life or death cannot be detained. Similarly, for offences punishable with imprisonment for seven years or more, the maximum period that a person can be detained should be three years (in line with clause (c) of Section 474 of BNSS). Since there is no term of imprisonment specified in the case of commutation of conviction for a term of less than seven years in Section 474 of BNSS, we could follow the principle of one third of the maximum punishment in such cases. For instance, the maximum period that a person can be detained for an offence punishable with three years of imprisonment can be fixed as one year.

As the Supreme Court has reiterated in some of the recent judgements, the right to life of the accused cannot be obliterated, no matter how serious the alleged offence. An accused has the right to speedy trials and if the State and the Court cannot provide it to him, they have no right to detain him for a prolonged period and award surrogate punishment, but have to release him after a reasonable time. The bogey of the possibility of tampering with the evidence, influencing witnesses, and flight risk cannot be grounds that trump presumption of innocence and right to life. And fixing the maximum period that an undertrial can be detained as a period equivalent to one third of the punishment for the offence is only a small step to show that fundamental rights are taken seriously and are not to be trampled upon indiscriminately.

Similarly, there is no reason why presumption of innocence should be whittled away due to the subsequent conduct of a person who has been found guilty on a previous occasion. A law that treats a person harshly on the basis of previous conviction when he is yet to be proven guilty on a subsequent occasion is just a legal garb for the prejudice that operates against a who has been found guilty once. It may make sense to treat a person harshly on a subsequent conviction – as in the award of enhanced punishment, but not when the person is yet to be proven guilty. Therefore, the upper limit on the maximum period for which an undertrial can be detained should apply uniformly to all undertrial prisoners regardless of whether they have been convicted earlier or not.

If the clause about the maximum period an undertrial prisoner can be detained has to be implemented effectively, it has to be raised to the level of a statutory obligation on the lines of Section 187 of BNSS (which is pari materia Section 167 of CrPC), which leaves no discretionary power to the Magistrate. If the right to be freed when the investigation is not completed within the stipulated time is a fundamental right and not merely a statutory right as held in Vikramjit Singh v. State of Punjab (2020) 10 SCC 616, the right to be freed if the trial is not concluded within a reasonable time can be no less in stature. The option of continuing detention after hearing the Public Prosecutor would virtually close the doors of bail as Magistrates more often than not get persuaded by rhetoric of Prosecutors, who at times act as though they were Police Prosecutors rather than Public Prosecutors.

In the current scenario, as there is no statutory limit specified for the conclusion of trial, it is left to the Constitutional Courts to intervene and adjudicate on the question whether the right to life of the accused person has been violated due to prolonged incarceration without trial. And this would happen only in the case of those who can afford to reach up to the Constitutional Courts. As it is left unspecified, the interpretation of what ‘reasonable time period’ is varies according to the vagaries of the varied Benches. It also depends often on how soon or late the file of the accused person reaches before a sympathetic Bench. If in some cases it is a matter of a year or two, in some cases it is nine or ten years or even more. Specifying the outer time limit for the conclusion of trial in the statute would not only make it uniformly applicable, but it would also make it enforceable at the level of the trial court and thereby within the means of those who cannot afford to know the doors of the Constitutional Courts.

Conclusion

The proposal to have parity between the term of imprisonment to which sentences of imprisonment can be commuted and the maximum period for which an undertrial prisoner can be detained – both in proportion to the sentence for the offence in question – cannot be repugnant to the spirit of justice the new laws are supposed to usher in. Mandatory release of an undertrial prisoner after the stipulated time period is only part of the commitment to conclude the trial within that time frame, and thereby ensuring justice to both the accused person and the victim of the crime. Keeping an undertrial person in prison endlessly and awarding surrogate punishment is justice to neither parties. And the new criminal laws are all about ushering in an era of justice, aren’t they?

Section 167, 'Default Bail', and its Cancellation

Recently, a Division Bench of the Supreme Court delivered its judgment in Central Bureau of Investigation v. T. Gangi Reddy [Crl. Appeal No. 337 of 2023, decided on 16.01.2023; "Gangi Reddy"]. The CBI had gone up to the Supreme Court against an order of the Andhra Pradesh High Court from March, 2022, and what it wanted was to cancel the bail of the Respondent, who was one of multiple accused persons in the case concerning the murder of a former minister, Y.S. Vivekananda Reddy. 

Petitions challenging bail, either because bail orders are unreasoned or because the accused violated the terms of bail, are quite common. What made Gangi Reddy different (not unique) was that the bail order in question was what is colloquially called 'default bail' — bail granted under Section 167(2) of the Criminal Procedure Code, 1973 ["Cr.P.C."] because police failed to complete the investigation within the specified timeframe, which in cases of murder is ninety days. 

The CBI contended that the High Court was wrong in concluding that default bail could not be cancelled for considerations such as the factual merits of a case [Paras 1-2, Gangi Reddy]. The Supreme Court agreed with the CBI, invoking a line of precedent stretching back to 1977, and set aside the High Court's order directing it to reconsider the CBI's plea.

Well, this is what the proceedings look like if we take a bird's eye view. Once we opt for a close-up, things start to appear less straightforward. 

The Contentions in Gangi Reddy

The High Court's order dismissing the CBI's plea for cancellation of bail is available online. It is a lengthy order, and extracts the submissions of the CBI in detail, which were largely premised on demonstrating to the court that allowing the accused to remain on bail posed a real and perceptible threat of his tampering with evidence and harassing prosecution witnesses. The High Court dealt with these submissions at length and found that there was insufficient material to suggest that the accused was engaging in such conduct. 

So, at least upon a perusal of the order itself, the issue that the CBI placed before the Supreme Court was not directly in issue before the High Court at all. There is no sentence to suggest that High Court had, in as many words, held that an order of default bail cannot be cancelled upon considering the merits of the case at a later stage. 

At best, what we can do is infer the High Court said this by reading its discussion on the law regarding bail, where the court did not make the effort to positively state this proposition and merely noted that bail can be cancelled due to circumstances suggesting the accused misused his liberty. 

For reasons that are, therefore, unclear, the proceedings took an abrupt turn at the Supreme Court where not only did the CBI base their case upon a tangential finding in the impugned order, but even counsel for the respondent do not appear to have tried any course correction and instead justify an imaginary stand attributed to the High Court. 

Peculiar indeed.

A Subtle, and Incorrect, Shift in the Legal Position

Gangi Reddy professes to stick to precedent in arriving at its conclusion, which requires to be reproduced in full:

"The issue involved in the present appeal is answered in the affirmative and it is observed and held that in a case where an accused is released on default bail under Section 167(2) Cr.P.C., and thereafter on filing of the chargesheet, a strong case is made out and on special reasons being made out from the chargesheet that the accused has committed a non-bailable crime and considering the grounds set out in Sections 437(5) and Section 439(2), his bail can be cancelled on merits and the Courts are not precluded from considering the application for cancelation of the bail on merits. However, mere filing of the chargesheet is not enough, but as observed and held hereinabove, on the basis of the chargesheet, a strong case is to be made out that the accused has committed non-bailable crime and he deserves to be in custody." [Emphasis mine]

Again, at a bird's eye level, there is not much different in what the Court has held here to what was held in Bashir & Anr. v. State of Haryana [AIR 1978 SC 55] when the issue first came before the Supreme Court. Bashir, Raghubir Singh [AIR 1987 SC 149], Rajnikant Patel [AIR 1990 SC 71], Aslam Babalal Desai [AIR 1993 SC 1], and now Gangi Reddy, all sing in unison that the mere filing of a chargesheet is not a good enough basis to cancel bail granted under Section 167(2), Cr.P.C. So far, so good.

The problem emerges when we look at what all of these decisions prior to Gangi Reddy had said about what might actually be good grounds to cancel the bail granted under Section 167(2). It starts with Bashir, where the Court held that:

"The court before directing the arrest of the accused and committing them to custody should consider it necessary to do so under section 437(5). This may be done by the court coming to the conclusion that after the challan had been filed there are sufficient grounds that the accused had committed a nonbailable offence and that it is necessary that he should be arrested and committed to custody." [Emphasis mine]

The test, if it can be called one, was that the court "should consider it necessary" to cancel bail, and one of the circumstances which could deem it necessary was the emergence of sufficient grounds that the accused had committed a non-bailable offence and that his arrest was necessary. Not 'or' the arrest is necessary, but a twin condition, requiring separate findings on why the arrest was necessary. 

Then comes Raghubir Singh, where the Court held that:

"The order for release on bail may however be cancelled under s. 437(5) or s. 439(2). Generally the grounds for cancellation of bail, broadly, are, interference or attempt to interfere with the due course of administration of justice, or evasion or attempt to evade the course of justice, or abuse of the liberty granted to him. The due administration of justice may be interfered with by intimidating or suborning witnesses, by interfering with investigation, by creating or causing disappearance of evidence etc. The course of justice may be evaded or attempted to be evaded by leaving the country or going underground or otherwise placing himself beyond the reach of the sureties. He may abuse the liberty granted to him by indulging in similar or other unlawful acts. Where bail has been granted under the proviso to s. 167(2) for the default of the prosecution in not completing the investigation in sixty days, after the defect is cured by the filing of a chargesheet, the prosecution may seek to have the bail cancelled on the ground that there are reasonable grounds to believe that the accused has committed a non-bailable offence and that it is necessary to arrest him and commit him to custody. In the last mentioned case, one would expect very strong grounds indeed." [Emphasis mine]

Again, a recognition that there is an additional ground for seeking cancellation of bail when it comes to an order under Section 167(2), but framing it as not merely being the emergence of reasonable grounds that the accused committed an offence but an additional requirement that this makes his arrest necessary. As the Court itself noted, in such cases "one would expect very strong grounds indeed".

Which brings us to the judgment in Aslam Babalal Desai, the only Three Justices' Bench decision in this line of precedent where, remarkably, all three Justices penned different opinions. On an outcome basis, it is recorded as a 2-1 split where Ramaswamy J. in his separate opinion (startlingly bereft of clarity) agrees with Ahmadi J.'s opinion, and Punchhi J. dissents. 

What exactly is the dissent regarding? It is about equating bail under Section 167 with one under 437 or 439 — the minority disagrees with treating the deeming fiction in Section 167 this way, and insists that a default bail cannot be deemed as being the same one on merits. For the minority, it is because default bail is different that there is a separate, "singularly sufficient", ground for cancelling such bails: emergence of sufficient grounds to believe the accused committed the offence and that his arrest is necessary. Therefore, the "strong grounds" referred to in Raghubir were nothing but the merits of a case.

Ahmadi J. gave the deeming fiction its fullest expression, finding that Section 167 was the manifestation of the legislative anxiety when it came to pretrial detention and personal liberty. This led him to conclude that while default bail under Section 167 could, naturally, be cancelled too, it could not be "interfered with lightly" on the grounds of filing a chargesheet. Rather, to cancel bail there must be "special reasons for so doing besides the fact that the charge-sheet reveals the commission of a non-bailable crime." [Emphasis mine]

The majority and minority both agree that the merits matter when entertaining a cancellation of bail that was granted under Section 167(2). Where they differed was whether the merits are "singularly sufficient" — the majority clearly said that the special reasons to cancel bail must be besides the fact that today there is a chargesheet which reveals the commission of a non-bailable crime. 

Finally, we can return to Gangi Reddy. Take a close look at how the Court now framed the special reasons / strong grounds issue in the extract that was reproduced above: "special reasons being made out from the chargesheet that the accused has committed a non-bailable crime ... strong case is to be made out that the accused has committed a non-bailable offence and he deserves to be in custody." This is clearly not what was held by the majority in Aslam Babalal Desai, but speaks to the minority.

This incorrect attribution of what was the actual holding in Aslam Babalal Desai is even more pronounced at Para 9 of Gangi Reddy, where while summing up this earlier judgment the Court notes that bail cannot be cancelled on the mere filing of a chargesheet but upon "making out a special and strong ground that commission of non-bailable crime is disclosed from the chargesheet." With the greatest respect, this is not  what the Court held, but only what the minority held in that case. 

Even in its reasoning, the Court in Gangi Reddy is inspired not by the majority but the minority in Aslam Babalal Desai. Where the majority saw Section 167 as an expression of legislative anxiety that merited the fullest protection of the courts, the minority opinion of Puncchi J. spoke of how bails under this clause could be "managed through a convenient investigating officer, however heinous the crime" and decried the resultant injustice if courts were denied powers to cancel such bails. Fast-forward thirty years to Gangi Reddy, where the bench was eager to emphasise the perils of limiting the powers of court to cancel bails given that there would be cases where the accused "manages through a convenient investigating officer ... not to file the chargesheet ... giving a premium to illegality and / or dishonesty." 

Post-Script: Course Correction in an Alternate Reality

It is unclear whether the Supreme Court in Gangi Reddy even had to decide the issue of whether the merits of the case disclosed in the chargesheet cannot be considered in a plea for cancelling a default bail order that is passed under Section 167(2). Nevertheless, it took up the issue, and reiterated the existing position that merely filing a chargesheet was not good grounds to cancel a default bail order. 

However, much like all the earlier occasions on when it considered this issue, the Supreme Court in Gangi Reddy was remarkably unclear in what can be good grounds to cancel an order of default bail. If anything, it appears that the Court has preferred a subtle, yet incorrect, shift in the legal position by relying upon the view of a minority opinion in an earlier decision as against what the decision had actually held. Therefore, Gangi Reddy requires reconsideration.

In an alternate reality, where such a reconsideration does occur, one would hope that the Court takes a long hard look at what prompts this confusion — locating the power to cancel bail under Section 437(5) and not only under Section 439(2). By extension, it would mean not equating a bail order with granting bail under Section 437, but under Section 436. This is not outlandish, but what giving fullest expression to the deeming fiction would look like, and a view that was endorsed by the Division Bench of the Delhi High Court in Noor Mohammad [ILR 1978 Del 442]. 

Unlike the approach in Gangi Reddy which views Section 167(2) bail orders as a mere technicality that should be treated with suspicion and cast aside at the first available opportunity, Noor Mohammad gives us a glimpse of just how significant the introduction of this clause was within the criminal justice landscape at the time, and why courts stood up to ensure that it was given the fullest protection. 

That is the subtle, and necessary, shift in the legal position we require.

Muddied Waters: The Supreme Court's Decision in Gautam Navlakha's plea for Statutory Bail

The decision of a Division Bench of the Supreme Court in Gautam Navlakha v. National Investigation Agency [Crl. Appeal No. 510 of 2021 (Decided on 12.05.2021) ("Navlakha")] continues the rich tradition of poorly authored judgments which frequently emerge from the Supreme Court of India and other High Courts on matters of criminal procedure and personal liberty. The judgment is unclear, to say the least, and peppered with needless extracts throughout. The result is a laborious 206 page exercise which judges, lawyers, and law students will struggle with for some time to come (and, curious laypersons would be well-advised to avoid reading it altogether).  

The Facts

The Appellant was arrested in August 2018 and transit remand was granted to the police to take him from Delhi to PS Vishrambaug. But the arrest and remand were challenged in a habeas corpus petition filed on the same day before the Delhi High Court. Since it was unable to decide the petition on the same date, the High Court did not order the Appellant's release but directed that he be detained under "house arrest" till the petition could be decided. Later, in a separate public interest litigation, the Supreme Court extended this "house arrest" of the Appellant till it could decide that litigation. 

What we have then is a clear chain of events: the Appellant was arrested and remanded to custody by the trial court, and the terms of this custody were modified by the High Court till it decided a habeas corpus petition. The terms of this detention were clearly spelt out: the Appellant was not permitted to leave or interact with persons save his lawyers or doctors; his house was guarded by police, and crucially; he was not available for interrogation either. This detention was then extended by the Supreme Court. Ultimately, on 01.10.2018, the High Court ruled that the remand order was illegal. 

This did not spell the end of the litigation, of course, and as is well known the Appellant was arrested later and taken into custody in April 2020. In June, the Appellant raised a plea of statutory bail (or default bail) before the trial court in Maharashtra. This is a concept critical to criminal procedure and personal liberty in India. Once custodial detention during investigation crosses certain time-periods and the investigation is yet to be concluded, it triggers a right to bail for an accused under Section 167 of the Code. This "statutory bail" under Section 167 is distinct from the liberty that accused persons have to file bail applications and plead for release, because while judges retain vast discretion to grant or refuse ordinary bail applications, there is no such discretion when it comes to statutory bail. 

The Appellant claimed this right to bail because, according to him, the time spent by him in custody had to include the 34 days of custody he suffered in 2018 as well. Both the trial court and the Bombay High Court disagreed because the "house arrest" suffered by the Appellant could not been as custodial detention of the kinds envisioned under Section 167 of the Code. Even if that kind of detention could be seen as the kind of detention contemplated by Section 167, in this case the Delhi High Court had anyway held that the remand order and subsequent custody was illegal and thus stood wiped out. 

In essence, then, these were the two issues presented before the Supreme Court.

The Supreme Court's Verdict: A Dishonesty of Convenience 

Navlakha holds that the period of custody undergone by an accused during investigation into commission of non-bailable offences must be counted towards computing the total time spent in such custody as under Section 167 of the Criminal Procedure Code 1973, even if a superior court later rules that the accused had been illegally remanded. Furthermore, it holds that confinement in "house arrest" after being apprehended by the police can be a form of custodial detention that is sanctioned in law under Section 167. In doing so, the Court expands the notion of custody beyond the usual dichotomy between "police custody" and "judicial custody", which was traditionally seen as detention in police lock-up and being sent to jail. This custodial detention at one's home would, by extension, also be included towards computing time spent in custody by an accused. Not only this, the judgment in Navlakha goes on the offensive and pushes for looking at house arrest more seriously at all stages of the process as an alternative to traditional judicial custody so as to reduce the prison population [Paras 137-140]. 

In the facts of the case before it, though, the Supreme Court was compelled to conclude that this particular house arrest suffered by the Appellant for 34 days could not have been ordered under Section 167 of the Code. As a result, it could not be said that his custodial detention during investigation had crossed the time-limit which entitled his release on bail.

Why does the Court conclude that the house arrest suffered by the Appellant, though certainly a kind of custodial detention, was not of the kind contemplated under Section 167 of the Code? The best chance at understanding this comes from closely reading the judgment from Para 131 onwards where reasons are offered — which are neither compelling nor consistent. It would be unhelpful to go through each of these individually and instead I'll highlight the underlying point, which is that the terms of this house arrest were such that it would not fit within the framework of Section 167 of the Code. No case diaries were inspected before directed house arrest and, furthermore, there was no possibility of interrogating the accused given the directions of the High Court. Thus, as the judgment puts it at Para 134.

That house arrest, in turn, involved, deprivation of liberty and will fall within the embrace of custody under Section 167 of the CrPC, was not apparently in the minds of both this Court and the High Court of Delhi. This is our understanding of the orders passed by the court.

With due respect, this is bogus. It does not matter what was "apparently in the minds" of judges when they passed certain orders. In fact, because it is impossible to fathom with any degree of certainty what might have been in the mind of an authority when it took a decision, judicial review always implores judges to refrain from engaging in such an inquiry. Instead, judges must locate the decision within the framework of law, and consider whether the exercise of discretion which lay behind the decision was not arbitrary. Thus, once the Court in Navlakha concluded that house arrest is custodial detention within the meaning of Section 167 of the Code, it cannot conveniently revisit this conclusion in the facts of a particular case and locate a specific instance of house arrest outside of Section 167, because of what it thinks was "apparently in the minds" of judges. If the detention was not in terms of Section 167, then it was illegal, but that would not stop the clock for statutory bail under Section 167 as the Court concludes. 

Conclusion

As convenient as it may have seemed, it was still wholly impermissible for the Court to locate the 34 days of custody suffered by the Appellant in a no-man's land which is both within the law and beyond it. This kind of judicial exceptionalism in the face of hard cases not only unmoors the concept of judicial review from its very foundations, but flips it entirely to permit a pick-and-choose of the worst sort imaginable. 

At this point, one would be forgiven to think this is too harsh a criticism of the judgment in Navlakha. After all, and this is a point that the judgment itself puts forth, even the Appellant and other accused persons did not imagine house arrest as a liberty-depriving measure at the time when the orders were passed. I agree; and if the Court was still unconvinced of its conclusion on this front, it could very well conclude that house arrest was not custodial detention within the framework of law. It could do this while also airing its concerns about prison overcrowding to suggest that perhaps the legislature should start to look at house arrest more seriously as an alternative to existing judicial custody detention. 

But the Court chose to not adopt this restrained course of action, which is why its conclusions are a perfect exercise in adopting what I label a dishonesty of convenience. Or, to put it more plainly, an effort to have its cake and eat it too.