By November 2024, all persons arrested in the high-profile ‘liquor scam’ prosecution under the dreaded Prevention of Money Laundering Act 2002 (PMLA) had been released on bail. Until then, the insurmountable ‘twin-test’ for bail under the PMLA, where the accused had to demonstrate her innocence at the bail stage, had come in the way of bail applications.

When the courts – both the Delhi High Court and the Supreme Court – finally granted bail, their decision did not  come from a finding that the accused had a good case. Rather than deal with merits, the courts cited delays in the process, concluding that prolonged incarceration would effectively erode whatever meaning was left of the presumption of innocence.

In legal terms, the argument is simple. Article 21 of the Constitution protects the right to life and personal liberty. While undertrials are innocent until proven guilty, they can yet be detained in custody pending trial if certain parameters are met. In some laws such as PMLA, the Unlawful Activities (Prevention) Act 1967 (UAPA), or the Narcotic Drugs and Psychotropic Substances Act 1985 (NDPS), the parameters include clauses requiring persons to establish their innocence at the bail stage. However, since undertrials are yet to be proven guilty of any crime, there arrives a point in the time-spectrum of their detention after which such detention becomes an affront to Article 21.

Let’s call this moment in time as an inflection point. Once this inflection point is crossed, any supposed justifications for custody are rendered meaningless regardless of the statutory text, and a constitutional court ought to release persons on bail.

What is on display is not constitutionalism. Rather, it is a jurisprudence of apology.

As an outcome, grant of bail in a system where persons languish as undertrials is something we (usually) can rally behind. As a processual matter though, I find it rather hard to muster the same enthusiasm for these orders where courts evoke fidelity to Article 21 and personal liberty.

For what is on display is not constitutionalism. Rather, it is a jurisprudence of apology, a by-product in a system where courts are neither discharging their discretion to decide bail as soundly as they ought to, nor are they conducting trials as they ought to.

An approach where courts can grant bail where they feel it has been too long barely contributes to a fairer system of bail. At the same time, with any approach where some kind of premium gets attached to the idea of delay — here, the premium being possible release from custody — there is the possibility of this delay slowly becoming normalised over time. Whichever way we look at it, the more one scrutinises this inflection point logic in bail, the more worrying it gets.

Stating the problem

My concerns with the inflection point jurisprudence in bail are broadly threefold. The primary issue is one of arbitrariness and the evil consequences that it begets. The second issue is narrower. Locating the source of this power within the Constitution suggests that bail on such grounds can only be granted by constitutional courts, which entirely casts out the most crucial site for bail decision-making: trial courts. The last issue is more a fear than an issue: will the consequence of normalising this inflection point approach result in making longer custody a norm? Would courts will now find it possible to evade a merits-based determination altogether knowing that a ‘right’ to bail will crystallise down the road?

Will the consequence of normalising this inflection point approach result in making longer custody a norm?

In a fair system, there is a common time-spectrum for a category of undertrials, which judges can refer to decide when further detention is egregious. An example is default bail at the stage of investigations (section 167 of the old Cr.P.C., and 187 of the new BNSS). The law prescribes 60 days as the time limit for completing investigations for one category of persons, 90 for another, and in some special laws 180 days.

What we have with the inflection point logic, however, is no such common guideline but each judge’s personal sense of right and wrong deciding the fate of an undertrial pleading for bail on grounds of delay. This breeds the most obvious and invidious form of arbitrariness in decision-making, where like cases are not treated alike at all. Personal liberty becomes reduced to pure chance and litigation to a game. After what feels like a long enough time, litigants may begin trying their luck at securing judicial pity. In some cases, with highly persuasive lawyers, a court may even be convinced to grant bail if shown that the status of trial is unlikely to change in the foreseeable future and that it is futile to wait till the inflection point arrives.

Because the trial will never end and merits-based scrutiny on bail is never going to be favourable, the choice is not between guilt or innocence, but between early release with a guilty plea or bail after spending the better part of one’s life in custody.

The invidious consequences have wider ramifications for the sanctity of the criminal process. For instance, it might driving up guilty pleas. Undertrials trying their luck at bail on grounds of delay are not merely sitting idle waiting for the mythical inflection point to arrive while the unending trial drags on. Such persons usually have had previously rejected bail applications, especially where alleged offences are heinous or concern socioeconomic or national security affairs. And all through this while, certain ideas and expectations around the system naturally arise for those living inside it (Mayur Suresh’s excellent study on this aspect in the context of terror trials gives a ringside view to this). Undertrials verily believe, at the instance of agencies or fellow-prisoners or well-wishers, that because the trial will never end and merits-based scrutiny on bail is never going to be favourable, the choice is not between guilt or innocence, but between early release with a guilty plea or bail after spending the better part of one’s life in custody.

Finally, there is the issue presented by limiting bail on delay grounds to the constitutional court. For the forlorn undertrial, this means that getting bail on grounds of delay is an expensive effort requiring petitions before a high court and then Supreme Court — no mean feat for an ordinary accused person. Trial courts judges may sympathise with the accused but will shrug their shoulders because the recent spate of judgements insists that the power to grant bail overriding statutory constraints by invoking Article 21 is one reserved for constitutional courts. A high court or the Supreme Court may be able to disintegrate the stricter conditions on bail by invoking Article 21 and the rhetoric of fundamental rights. However, that remains a bridge too far for courts at the ground level.

The Indian experience

As I briefly alluded to in the previous section, there are a few statutorily identified inflection points that govern how undertrials are treated in the Indian criminal process.

The idea of having time limits on custody is one that has been part of the reform imaginary in India since the 1950s. The first notion of a time limit came in 1955, with the law on bail before magistrates being amended to prescribe a 60-day time-limit under Section 497(3A) to conclude evidence after framing of charges, failing which bail would be granted. This has remained part of the law ever since, first as Section 437(6) Cr.P.C. and currently as Section 480(6) BNSS.

The default bail regime has become a cautionary tale [...] specifying a time-limit would send the wrong signal that an investigation — and its concomitant investigative custodial detention —should inevitably take two to three months. 

It was not long before this time-limit came to be accompanied by the most well-known time-limit: the rule that when persons are arrested and in custody, investigations must be completed within a fixed period. This was Section 167 of the 1973 CrPC, retained under Section 187 of the BNSS.

This initially was deemed quite radical, even by the Supreme Court, and the critics were mollified by creating gradations. The blanket rule of 60 days gave way in 1978 to the present regime of having two classes of undertrials divided across the category of alleged offence, and having those accused of more serious cases treated differently from others.

Rather than seeming radical, the default bail regime has become a cautionary tale. Its lived experience of 50 years has confirmed the fears held by the Law Commission at the time of its introduction, that specifying a time-limit would send the wrong signal that an investigation — and its concomitant investigative custodial detention —should inevitably take two to three months. In other words, the maximum became the minimum.

The only omnibus time-limit in place today – in the sense that it applies regardless of the stage at which proceedings are at – was mooted by the Supreme Court in the early 1980s and firmed up in 1994 into a clear directive to grant bail where people remained as undertrials in prison for longer than half the sentence for the crime they were charged with. It received statutory recognition in 2005 through introduction of Section 436A.

The accused had served enough time, and continued incarceration as an undertrial had become an embarrassment. The language in the relevant cases [...] is not an energised reclamation of personal liberty, but an apology from the highest court of the land.

This time-limit, though, was fundamentally different from the preceding ones. It was not a timer which was designed to make things work for the trial process to have meaning and work speedily towards conclusion, but it was a time limit where the court recognised that all these principles and ideals had failed. There was no longer any point to the trial because the accused had served enough time, and continued incarceration as an undertrial had become an embarrassment. The language in the relevant cases – Hussainara Khatoon, SC Legal Aid Committee – is not an energised reclamation of personal liberty, but an apology from the highest court of the land to the most downtrodden for having left them behind.

Making bail fairer

To be extremely clear, none of the above criticisms mean that bail on grounds of egregious delay should not be granted. It absolutely should.

What we need is a fairer process to administer bail so that the undertrial is not left at the mercy of unbridled judicial discretion and or saddled with filing expensive petitions before superior courts. The bail process cannot be moored within the exceptionalism of constitutional litigation; it has to be rooted in the ordinariness of statutory remedies. At the same time, we have to guard against bail on grounds of delay morphing into a poisoned chalice, where this exceptional remedy slowly normalises extended custodial detention till the inflection point.

The bail process cannot be moored within the exceptionalism of constitutional litigation; it has to be rooted in the ordinariness of statutory remedies.

As an example – by no means a perfect one – consider the Custody Time Limits that are part of the process in the United Kingdom. Introduced in the late 1980s, custody time limits are the ambitious outer limits for concluding various phases of the criminal trial, which operate if the accused is in custody. If the time limit is not met, the accused has a right to claim release on bail.

Prescribing ambitious time limits ensures that the outer limit does not result in normalising delays — in fact, extensions of the time limits during the Covid-19 pandemic drew sharp criticisms citing such fears. At the same time, grant of bail for breaching the custody time limit is not an automatic process, which allays fears of the crime control variety. Prosecutors can seek extensions of the time limit, and judges can reject bail requests if they find cogent reasons to do so. Nevertheless, it is a detailed system putting every stakeholder on notice, where bail on grounds of lengthy undertrial incarceration is not a mercy plea. Such a process is missing in India and is needed to make bail adjudication fairer.

Conclusion

How long is too long, as a recent article by two of India’s most famous undertrial detenus put it, is a policy which is actually not a policy and a practice contributing to wildly varying results. Using Article 21 to dilute strict bail conditions due to lengthy custodial incarceration, which the Supreme Court currently does, speaks to the apologetic legacy of Hussainara Khatoon even as it evokes the idea of enforcing a time-limit to have a meaningful trial process. Yet, these are fundamentally different ideas, which is why there is such instability in the enforcement of this remedy.

It is fascinating to note that much before the UK came up with custody time limits, a nascent Indian republic flirted with the idea in 1950s. What ended up as a rather narrow rule with Section 497(3A) limited to one phase of the process, had begun as a proposal for an omnibus rule for a time limit covering the entirety of criminal proceedings in the draft legislation. Committee deliberations whittled it down to a narrow rule, as it was felt that an omnibus remedy was too radical and even unnecessary as other measures existed to make magistrate trials run faster.

A practice where a different king retains powers to dole out mercy based on whoever pleads most effectively, without any rules or principles, is the flipside of the very same coin. Just because it is the side we like does not make the coin any better.

As a glance at the Law Commission’s 1958 Report on Reforms in Judicial Administration will show, circumstances have changed dramatically since the 1950s. Trials take  incomparably longer today. In such circumstances, the core idea that persons should not be detained as undertrials for so long as to render conducting trials to find them guilty and sentencing them to prison meaningless has become unremarkably obvious. It is heartening that the Supreme Court has begun to consistently acknowledge this.

Yet, the Supreme Court needs to break out of a trajectory that has codified the moment at which a trial is rendered meaningless and bail is awarded as an apology. It must double down on the idea of having a meaningful trial for undertrials and develop it further by giving guidance on the kind of time limit that would be needed to make it work. If it genuinely wants to try and begin solving the problem, the court must break away from the past models of using empathetic constitutional litigation to beg forgiveness from those who are so condemned while presumed innocent. Instead, it must shed this reactive posture and give a concrete and active meaning to fairness by reinvigorating the trial process, simultaneously empowering and nudging the trial courts to slowly restore a functional criminal process. 

To start, we can adopt a more systematic approach to the problem.

There are three time-consuming stages involved in every criminal case: (i) the time it takes for investigations to end, (ii) the time it takes for courts to decide whether charges should be framed or not, and (iii) holding the trial and passing judgement. Today, there is statutory clarity on how much time it should take for investigations to end and prosecutions to reach court. There is also some clarity on how much time it should take for trials to end.

But there is no clarity on how much time it should take to reach the trial itself. This phase in the prosecution, where police have filed their chargesheets and the case is being prepped and heard on the point of charge, becomes a black hole for the process. Courts must focus on identifying what is unfair delay at this stage and reinvigorate the already existing statutory guidance for the other two parts of the process. Care must be taken to not prescribe overly generous – but ultimately aspirational – time limits, to avoid the maximum limit becoming the norm, as has been the case with default bail during investigations.

We have some guidance in the BNSS 2023, which offers a fresh perspective on investigations and the stage of charge. Unlike the 1973 Code, the BNSS prescribes time limits both for the completion of investigation and for arguing the point of charge. Why not treat these time-limits as guidance for the delay that can render custody onerous for this part of the process?

In a recent judgment concerning the notorious practice of bulldozing houses of persons accused of offences, the Supreme Court issued paeans to the rule of law, reminding society that the Republic of India could not move away to a rule where might is right and the king can administer whatever manner of justice he deemed fit. A practice where a different king retains powers to dole out mercy based on whoever pleads most effectively, without any rules or principles, is the flipside of the very same coin. Just because it is the side we like does not make the coin any better.

Abhinav Sekhri is a lawyer practising in New Delhi. When not working or writing about criminal law, he can be found trawling through the archives in search of curious stories from Indian legal history.