The New Supreme Court “Guidelines” on Bail

By: Abhinav Sekhri

It is fairly well known that under the Criminal Procedure Code 1973 [Cr.P.C.], the investigating agency files a Report before court upon completing an investigation, sharing the finding of its investigation [Section 173]. What is not so well known is that, at this stage, if police concludes that there is sufficient evidence to prosecute a case, then under Section 170 the officer “shall forward the accused under custody” to the court empowered to try the case. Only where the offence is bailable, and, the accused can furnish some security to the effect that she shall appear before court when required, can the police choose not to forward the accused in custody. 

This clause is not problematic when persons are arrested during investigation as the “custody” requirement is met. But Section 170 creates a perplexing situation where no such arrest takes place, for it seemingly demands that all accused persons should be taken into custody when the investigation is complete and the case is sent to court, without elaborating more about the nature or duration of such custody. 

In August of 2021, a Two Justices’ Bench of the Supreme Court in Siddharth v. State of U.P. [Crl. 838 of 2021, decided on 16.08.2021] was presented a chance to interpret Section 170 in a case where an accused person was served with warrants for his arrest upon completion of the investigation by the police so that it could present the 173 Report in court. Troubled by the mandate of this clause and the prospect of routine arrests it entailed, the Supreme Court blessed the line of High Court cases which had interpreted Section 170 to contain a measure of discretion, so that in cases where the police officer had no reason to suspect an accused will abscond, then there was no cause to arrest the person upon completing the investigation. The Court used the opportunity to reiterate what has become the settled legal position on the issue of arrests:

We may note that personal liberty is an important aspect of our constitutional mandate. The occasion to arrest an accused during investigation arises when custodial investigation becomes necessary or it is a heinous crime or where there is a possibility of influencing the witnesses or accused may abscond. Merely because an arrest can be made because it is lawful does not mandate that arrest must be made.

The story did not end here, though, because in parallel proceedings based on seemingly similar facts [Satender Kumar Antil v. CBI & Anr., SLP (Crl). 5191/2021], the same bench of the Court on an earlier date had expressed its displeasure at the prospect of routine arrests and also felt that it was appropriate to “lay down some principles in this behalf“. Once the judgment in Siddharth was pronounced, the Court on 18.08.2021 suggested that counsels take note of the same to “assist” them in this case. On 14.09.2021, the agenda moved from principles (which, frankly, Siddharth provided) to “Guidelines”, as the Addl. Solicitor General submitted to the Court that he would want to provide suggested guidelines after consulting other counsel. That exercise culminated in the order dated 07.10.2021 in Satender Kumar Antil, where the Court blessed the Guidelines so proposed by the counsel. 

Understanding the Guidelines

If you have had a chance to read the order of 07.10.2021, then you may as well skip this section which looks at what are the Guidelines. At the outset, it must be stated clearly that these are guidelines and do not override judicial discretion available to judges to decide cases on their facts. But, as anything that comes from the Supreme Court, it will probably be given great deference by the lower courts. 

The Guidelines in Satender Kumar Antil are meant to be relevant only if two conditions coexist: (1) A person is not arrested during an investigation, and (2) She cooperated during the investigation including appearing before the police when required. The Guidelines state that for such persons, there is no need to arrest such persons and send them to court when filing the 173 Report, and then suggest how courts should decide the issue of bail for sucsuggest how a court should decide on the issue presented by Section 170 Cr.P.C. in respect of such persons, and does so by adopting an offence-specific approach. 

Offences have been divided across four Categories: (A) Those punishable with a term of seven years or less, but not falling either in categories B or D; (B) Offences punishable with death, life imprisonment, or a term of more than seven years; (C) Offences punishable with special acts which contain restrictive bail clauses, such as Section 37 NDPS, Section 45 PMLA, etc.; (D) Economic offences, not covered by the said special acts, and where emphasis ought to be given to the seriousness of the charge and associated punishment in each specific case. 

While for all such cases, persons ought not to be arrested and sent in custody while filing a 173 Report where the two conditions outlined above are met, the Guidelines suggest different future courses of action depending upon the kind of offences involved:

  • Category (A) cases are the most benign from a custody perspective, where resort to coercive process (warrants, as opposed to summons)  to secure appearance of accused persons is discouraged and bail applications may be decided without taking an accused in custody or releasing her on bail during pendency of a bail application.
  • For Category (B), (C) and (D) cases, courts are simply told that upon appearance of accused persons pursuant to issuing process (with no suggestion to first resort to summons and not coercive process), their bail applications ought to be decided on merits. For Category (C), courts must also consider the bail clauses contained in special acts while deciding cases. There is no bar on granting interim bail, but at the same time, it is not expressly commended in such cases either. 

Problems Ahead?
There are general issues, and some very specific ones, that I wish to place under the spotlight.
General Issues — Understanding “Cooperation”, “Merits”, and the Pitfalls of CategoriesThe preconditions set out by the Court require cooperation during an investigation besides the fact of the accused person not being arrested. The meaning of cooperation is not clarified by the Court, but it seems reasonable to assume that the Court meant the word means something more than appearing before police as and when required, for the court refers to this as being but one facet of cooperation. I agree that pegging cooperation simply to appearance might be counterproductive to investigative needs and, even contrary to law in some cases. But then again, the problem with a qualitative understanding of cooperation is one that has already been seen in context of bail, where police do not shy away from taking a stand that till an accused parrots whatever line the agency wants, she is not cooperating with the probe. By making cooperation a precondition, it confers just the kind of discretion at the stage of 170 Cr.P.C. which the Court might have wanted to do away with. If the Court was making the effort to give guidelines, a little extra effort to spell things out could have helped.
The other general issue is that of deciding bail applications on “merits”. The Court does not use this kind of language for Category (A) cases but the other more serious kinds of cases, either punishable with imprisonment for more than seven years and / or for offences under statutes with restrictive bail clauses. There is a fair bit to unpack here. If not merits, then on what basis are courts to decide bail applications in Category (A) Cases? Unless, of course, the Court meant that merits excludes the three keystones of bail i.e., that the accused is not a flight risk, will not threaten witnesses, or tamper with evidence; and instead a court ought to be concerned only with the merits of the allegation. It is no secret that the allegations play, arguably, the most important role even at the stage of bail in the Indian criminal process. But never has the Supreme Court recognised its position of first among equals. The Guidelines can be seen as doing exactly that, which in my opinion is contrary to law.     
Finally, there is the broad issue of the adoption of yet another set of categories for aiding in the exercise of bail discretion. This is an issue about which I have written about elsewhere, suffice to state that in creating categories where the court expressly discourages use of coercive powers for some offences, it ends up creating perverse incentives for the police to try and mischievously paint cases as involving more serious offences simply to keep a hold on the coercive powers that make ordinary people so fearful of the police setup in the first place. For it is police who retain almost exclusive control on how to frame the narrative in the language of criminal law — a victim might bring a story in the form of a complaint, but the police decides which offences are made out in a case. This levelling-up already happens with the cognizable / non-cognizable category, and began to happen post 2014 in respect of offences punishable with up to seven years being painted as more serious, when the Supreme Court limited the scope of arrest powers in Arnesh Kumar [(2014) 8 SCC 273]. In this scenario, while invoking the more serious offence will not give police automatic arrest powers in a Section 170 Cr.P.C. context, it can make bail a lot harder by taking a Category (A) case into Category (B), and encourage misuse and corruption as people would be willing to go to great lengths to remain out of custody.      
Specific Issues — Category (C) and the Economic Offences In noting Section 45 of the PMLA under Category (C) as an example of clauses prescribing restrictive bail clauses, the Court set alarm bells ringing in some quarters. Recall that the restrictive bail clause of Section 45, PMLA had been struck down as unconstitutional by the Supreme Court in 2018. Since then, the state and the Enforcement Directorate have tried a great deal to resurrect the clause, with the issue now pending consideration with the Supreme Court. For some lawyers, this reference to Section 45 is being seen as a blessing in disguise, which they fear might be used by the Enforcement Directorate to support its claim that the restrictive bail clauses of Section 45 are active. The Court would have done well to simply avoid reference to Section 45 given the contentious nature of the clause, and it cannot be seen as resurrecting the conditions at a time when the issue is pending before a different bench of the Court. Nevertheless, it will be interesting to see if the Enforcement Directorate tries to make use of this reference in Category (C) the next time it is opposing a bail under the PMLA.
Lastly, I come to Category (D), which covers “economic offences not covered by special acts“. This fetish with placing economic crime on the same pedestal as crimes punishable with death has been around since 1950 — back then, death was suggested for hoarders and black-marketeers — and is a sentiment that the courts have never let go of in the seven decades since the Constitution came to be. The scrutiny required to interrogate this moral equivalence is far too much for a blog post to contend with. While I disagree with this position strongly, I will assume its validity for now but instead ask why was the court required to paint its canvas with such a broad brush — economic offences will range from measly theft till robbery, but all were lumped together in Category (D). Thankfully the Court paid heed to the suggestion given by counsel and mentioned that factors such as the seriousness of charge and severity of punishment will matter when we step into the vortex of Category (D). But even then, the Court could not help emphasise the “different nature” of these offences, and so one can only wonder just what kind of warped alternate reality might the issue of bails for economic offences be pushed into following the Guidelines.        


Of course, we will know more about the Guidelines and the benefits / problems they bring only when the dust settles and they are implemented. At this preliminary stage, one can only flag concerns down the road, and the concerns are aplenty. Honestly, it is difficult to understand the need for the Guidelines after the Court had already laid down principles in Siddharth. Even though they are not meant to circumscribe judicial discretion and are “guidelines”, they are likely to be implemented with the force of law and throw up some difficult situations. One such will probably be the limited issue of Section 45 PMLA, and that one will be visible. The other issue will be the behind-the-scenes manipulation of case files that the Guidelines will prompt, where offences will be levelled-up or levelled-down by the police keeping the Guidelines in mind; all of which will be invisible, and will further erode the credibility of the criminal process.

[The author is a criminal lawyer based in New Delhi. This article first appeared on his blog, ‘The Proof of Guilt’.]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s