The Enduring Gaps and Errors In Capital Sentencing In India

Abstract: In the forty years since Bachan Singh upheld the constitutional validity of the death penalty in May 1980, there have been numerous concerns about the fate of the death penalty sentencing framework laid down by the majority. Inconsistent application, interpretational errors, and judge-centric decision making have dominated these concerns. However, this article seeks to revisit the premise of those narratives, i.e. these concerns have emerged as a result of the incorrect application of Bachan Singh. The focus is instead turned to the gaps within Bachan Singh itself and the manner in which those gaps have contributed to the subsequent fate of the sentencing framework. Demonstrating a complete collapse of what has come to be known as the ‘rarest of rare’ doctrine, the article identifies the procedural and substantive faultlines that have only widened over the last four decades. A profound lack of commitment to the rule of law and fair trial rights during sentencing proceedings lies at the heart of this doctrinal crisis where the courts are now burdened with a standard that is barely judicially maintainable. Unless we develop significant normative coherence and bring the full force of fair trial rights to bear on sentencing procedures, the constitutional crisis within death penalty sentencing will only deepen.

Introduction

In May 1980, while upholding the constitutionality of the death penalty, a five-judge constitution bench of the Supreme Court in Bachan Singh v. State of Punjab1 sought to guide sentencing discretion in capital cases by laying down a sentencing framework. This was a crucial shift in the jurisprudence on sentencing and punishment in India, as this was the first time that any law was laid down to guide the exercise of judicial discretion in sentencing, beyond the nominal guiding range provided by the legislature.

In order to understand the magnitude of the shift brought in by the sentencing framework, it is significant to understand the legislative shift in the period preceding Bachan Singh. In the 1898 Code of Criminal Procedure (the ‘CrPC’), death penalty was the default punishment for murder, requiring sentencing judges to give reasons if they wanted to impose life imprisonment instead.2 An amendment to the provision in 1955 removed the requirement of written reasons for not imposing the death penalty, reflecting no legislative preference between the two punishments.3 A substantial shift came in through Section 354(3) of the 1973 CrPC which made life imprisonment the default punishment and death sentence an exception requiring sentencing judges to give ‘special reasons’ while imposing the death penalty.4 However, there was no indication of what these ‘special reasons’ might be.

Bachan Singh filled in this gap and developed a sentencing framework applicable to Section 354(3) of the 1973 CrPC and, in doing so, created a space for individualised sentencing in capital cases. At its core, the framework in Bachan Singh was meant to guide sentencing judges in discharging their obligations under Section 354(3) of the CrPC, while choosing between the punishments of life imprisonment and death penalty.5 The focus of the framework was on individualised sentencing, as Bachan Singh required the courts to adequately consider all aggravating and mitigating circumstances, relating to both the circumstances of the offence and the offender. According to Bachan Singh, for a case to be eligible for the death sentence, the aggravating circumstances must outweigh the mitigating circumstances. In order to meet the threshold of ‘special reasons’ under Section 354(3) of the CrPC, Bachan Singh also required sentencing judges to establish that the alternative option of life imprisonment under Section 302 of the Indian Penal Code, 1860 (the ‘IPC’) was ‘unquestionably foreclosed’.

In many ways, the sentencing framework in Bachan Singh was an attempt by the Supreme Court to limit the powers of sentencing courts by laying down some guiding principles for subsequent courts. However, the majority refused to label these principles as ‘sentencing guidelines’, and observed instead that judicially mandated guidelines would go against legislative intent, since the Parliament had chosen to not cabin or fetter judicial discretion while amending the CrPC in 1973.6 Despite this caveat, the Court, nonetheless, went on to suggest a framework and laid down a list of aggravating and mitigating factors for subsequent sentencing courts to follow while deciding between life and death

The need for a framework, arguably, was an attempt by the Court to reduce the arbitrary imposition of the death penalty by providing a loose boundary within which unfettered judicial discretion could be exercised. In the absence of guidance in sentencing decisions, judicial discretion carries with it the real risk of arbitrariness and it is a concern that has received considerable attention from the Supreme Court of the United States. In Furman v. Georgia Jackson, 7 the United States Supreme Court through a five: four decision held that unguided discretion violates the Eighth Amendment as it permits juries to impose the death penalty on some defendants, while imposing life imprisonment on other similarly situated defendants convicted exactly of the same crime. Subsequently, the death penalty was reinstated in Gregg v. Georgia, 8 where the court approved legislative frameworks for limiting unfettered discretion through sentencing guidelines and automatically appealing all death cases for review.9 Bachan Singh, for its own part, upheld the constitutionality of the death penalty by holding that the discretion was not unguided, but, nevertheless interpreted ‘special reasons’ under Section 354(3) of the CrPC to create the ‘rarest of rare’ framework for guiding judicial decision-making.

Despite Bachan Singh’s attempt at creating safeguards, dominant narratives on the use of its sentencing framework over the last forty years have raised crucial concerns of inconsistent application, arbitrariness, and subjectivity of judges.10 This literature has particularly highlighted how subsequent judgments of the Supreme Court have gone against the principle of individualised

sentencing that was so crucial to the holding in Bachan Singh and essentially made it a judge-centric process. It captures the essence of the tension between the majority opinion and Justice Bhagwati’s dissent – he held that the capital sentencing system, which required ‘special reasons’ without any guidance on its meaning, essentially left decision-making to the subjective assessment of individual judges, making it arbitrary.11

As has been rightly noted, the fate of the Bachan Singh framework over the last four decades is a story of judicial error, misguided interpretations, and a lack of judicial commitment to due process concerns in sentencing. However, far less attention has been paid to be gaps within Bachan Singh itself,12 and this article seeks to identify those crucial gaps and the manner in which they have contributed to the concerns with capital sentencing over the last forty years. Without addressing fundamental due process concerns with death penalty sentencing, we cannot begin to address the close relationship between the gaps in Bachan Singh and the errors those gaps have facilitated since 1980.

This paper focuses on the concerns plaguing capital sentencing jurisprudence in India in the period following Bachan Singh. Part I traces the judicial evolution of the capital sentencing framework leading up to the Bachan Singh framework. Part II discusses the sentencing framework developed in Bachan Singh and the manner in which the gaps in the framework have led to its distortion. It focuses particularly on foundational questions on collection of mitigation evidence, the constitutional threshold for fair trial requirements in, and the appropriate judicial approach to, aggravating and mitigating factors which were not answered in Bachan Singh. Part III lays down the developments in the law on capital sentencing post-Bachan Singh, and discusses the manner in which several Supreme Court pronouncements have completely misread and misapplied the law laid down in Bachan Singh. This section also reflects upon the lack of clarity within Bachan Singh itself which has, in no small measure, contributed to these errors resulting in a complete collapse of the Bachan Singh framework. Part IV concludes by highlighting that any attempt towards repairing the broken nature of capital sentencing in India first requires bridging normative and procedural gaps within Bachan Singh and in the developments thereafter, so as to effectively protect the right to a fair trial.

II. Capital Sentencing Prior to Bachan Singh

The legislative shift from the death sentence being the norm for murder to it eventually being the exception also resulted in a corresponding judicial evolution. Before the 1973 amendment came into force, the Court in Jagmohan Singh v. State of UP13 in 1972 looked into the constitutional concerns surrounding judicial discretion in capital cases under the 1955 CrPC. Ediga Anamma v. State of AP14 in 1974 was the first decision that emphasised on the role of personal and social factors relating to the accused in sentencing. The recognition of death sentence as an extraordinary punishment in law in the 1973 CrPC eventually led the courts to interpret the meaning of ‘special reasons’ under Section 354(3). The meaning of ‘special reasons’ was first attempted by the Supreme Court in Rajendra Prasad v. State of UP15 While the idea of individualised capital sentencing saw emergence in Ediga Anamma in 1974 through its emphasis on a bifurcated trial under Section 235(2) of the CrPC, it took concrete shape and was completely embraced only in 1980 in Bachan Singh.

This section traces the judicial evolution of the death penalty sentencing framework starting from the Supreme Court decision in Jagmohan Singh in 1972. This shows the movement within the courts towards adopting the legislative change was brought in through the amendments to the CrPC. This allows us to better understand the judicial context in the lead up to the judgment in Bachan Singh.

The Supreme Court in Jagmohan Singh (1973) emphasised the need to maintain judicial discretion and observed that the amount of discretion vested in sentencing judges under the 1955 amendments to the CrPC was not excessive, and therefore, the arbitrariness of outcomes was not a concern.16 In 1974, the Supreme Court in Ediga Anamma commenting on the lack of statutory sentencing guidelines under the 1955 CrPC amendment highlighted the need for introducing “facts of a social and personal nature”, especially at the sentencing stage, to help judges focus on “not only the crime, but also the criminal”.17 It was observed that the bifurcation of a trial into conviction and sentencing stages under Section 235(2) of the CrPC would enable the collection of social and personal data of the offender, thereby aiding judges in “hearing the accused on the point of sentence” before imposing the appropriate punishment.

In 1979, the Supreme Court first discussed the meaning of ‘special reasons’ under Section 354(3) of the 1973 CrPC in Rajendra Prasad. The three-judge bench, headed by Justice Krishna Iyer, observed that ‘special reasons’ under the CrPC meant factors relevant “not to the crime, but to the criminal” and recommended judges to consider “the personal and social, the motivational and physical circumstances” of the criminal.18 The Court also discussed in further detail circumstances that justify the imposition of death, noting that a “callous criminal … jeopardising social existence by his act of murder” is deserving of the death sentence. This category was extended to include technologists, manufacturers, and white collar criminals who will fully jeopardise the lives of others to maximise their self-interest, as well as hardened criminals and dacoits who cannot be rehabilitated. Attempting to set out the justifications for imposing the death sentence, the Court ruled that ‘special reasons’ would exist “only if the security of State and society, public order and the interests of the general public compel that course as provided in Art. 19(2) to (6)”.19

Effectively, Rajendra Prasad failed to provide clarity for sentencing judges. On the one hand, while it did seek to introduce individual circumstances of the offender into sentencing, it also created crime categories that the judges seemed to think deserved the death sentence more. Creating crime categories to suggest that certain crimes deserve the death sentence irrespective of individual circumstances is to take away from the commitment to individualised sentencing. Further, the considerations of “security of State and society, public order and the interests of the general public” added even more uncertainty. Those considerations would open up a very dangerous set of consideration in death penalty sentencing where judges would be driven by utilitarian considerations rather than individual culpability. The risk of unguided discretion in death penalty sentencing was also increased by interpreting ‘special reasons’ to include factors like national security and public order.

III. Bachan Singh and Its Inherent Gaps

Far more than any judgment before it, individualised capital sentencing was embraced by a Constitution Bench in Bachan Singh, which moved away from the problematic prescription of death-eligible categories in Rajendra Prasad. Bachan Singh categorically warned against the standardisation of categories warranting death sentences20 as it severely undercuts individualised sentencing. It offered a framework with mitigating circumstances at its core and with the recognition of life imprisonment as the default sentence under Section 302 of the IPC, in line with the legislative policy underlined in Section 354(3) of the CrPC21. However, the inherent lack of procedural and normative clarity around these key aspects allowed for the distortion of the Bachan Singh framework by later decisions.

This section describes the framework laid down by Bachan Singh on capital sentencing and proceeds to discuss the impact of inconsistencies within Bachan Singh’s formulation, which has given rise to much confusion over four decades of its interpretation. It is argued that Bachan Singh’s own ambiguities have led to the lack of a meaningful capital sentencing process.

In considering the constitutional validity of the death penalty under Section 302 of the IPC, the Supreme Court in Bachan Singh had to decide two issues: first, the constitutionality of providing for the death penalty under Section 302 of the IPC; and second, the sentencing procedure articulated within ‘special reasons’ under Section 354(3) of the CrPC. Answering the first question in the negative, the majority opinion held that Section 302 of the IPC met the standard of reasonableness in Articles 19 and 21 of the Constitution.22 For the second question, it was argued that Section 354(3) vested unguided discretion with courts, leading to the arbitrary imposition of the death penalty. The court, however, held that the 1973 amendments to the CrPC addressed the concerns raised in its prior ruling in Jagmohan, and said that a rigid formulation of ‘special reasons’ would be impractical as judges would not be able to take account of variations in culpability.23 Therefore, the court was chose to lay down only very broad guidelines consistent with the legislature’s policy indicated in Section 354(3) of the CrPC.24 This led to the formulation of the sentencing framework, which required the weighing of aggravating and mitigating circumstances relating to both the circumstances of the offence and the offender, and deciding if the alternative option of life imprisonment was unquestionably foreclosed.

According to Bachan Singh, while deciding between life imprisonment and the death sentence, a sentencing court has to give due regard to both the circumstances of the crime and the criminal. Relative weight to be attached to aggravating and mitigating factors is dependent on the facts and circumstances of each case.25 However, one aspect that has been paid very little attention is the guidance from the majority opinion in Bachan Singh that mitigating factors must receive a ‘liberal and expansive’ instruction, while notably omitting such an approach for aggravating factors. This approach was appropriate given the legislative preference that life imprisonment is the default punishment and death sentence the exception. The majority opinion also seems to suggest that determination of ‘special reasons’ under Section 354(3) of the CrPC requires sentencing judges to establish that the alternative option of life imprisonment is unquestionably foreclosed.26 Thus, under the Bachan Singh framework, the death penalty can be imposed not only when the aggravating factors outweigh mitigating ones, but also when the alternative of life imprisonment is unquestionably foreclosed.

Bachan Singh itself, however, has many ambiguities, which have given rise to re-interpretations that do not sit comfortably with its original framework. At the core of the uncertainty surrounding the Bachan Singh framework is the lack of normative clarity on sentencing factors. Though the framework requires judges to consider aggravating and mitigating factors, Bachan Singh did not provide any conceptual clarity on the reasons for such a requirement and neither is there any normative clarity to be found on the relationship between these two set of factors. A reading of Bachan Singh also does not reveal the penological considerations behind judges being required to consider sentencing factors like age, socio-economic background, mental state, and reformation.27 By baldly asserting that these are relevant factors and no more, it left future sentencing judges the discretion to fill this normative gap with their own considerations. Further, the lack of a theoretical basis for the framework developed in Bachan Singh has impacted the procedural fairness of sentencing proceedings. A failure to indicate the integral role of sentencing factors subjects the collection, presentation, and consideration of these factors to a very low threshold. With no real judicial discourse on these aspects of sentencing, implications on the fairness of trials with very poor quality sentencing proceedings remain unexplored.

Crucial procedural and substantive aspects of capital sentencing that Bachan Singh leaves unanswered are discussed below.

A. Why are Mitigating Circumstances Relevant?

While Bachan Singh provides an indicative, not exhaustive, list of aggravating and mitigating circumstances, it does not clarify why these factors are relevant in a sentencing hearing. Mitigating circumstances provide insight into an individual’s historical, social, biological, and psychological context. Such information pertaining to their life history enables sentencing courts to meaningfully locate the individual in their unique context by providing a cohesive narrative of their life. This contextualisation allows for the courts to understand the implication of these life experiences on the individual and take them into account while deciding the quantum of punishment to be imposed.28 However, the absence of an underlying normative understanding of mitigation and its role in sentencing leaves the field open for judges to arbitrarily discard sentencing factors, or not accord appropriate weight to those factors.

Ediga Anamma,29 and later Santa Singh,30 indicate some penological rationale behind considering mitigating circumstances, observing that these factors personalise the punishment so that the reformist component is as much operative as the deterrent element. However, Bachan Singh fails to build upon this idea. The sentencing framework does not clarify that the list of aggravating and mitigating factors cannot be exhaustive, since the purpose of the sentencing exercise is individualising punishment, and in that exercise of individualisation, there are numerous possibilities in constructing the social, personal, and psychological history of the individual.31

Given this lack of normative explanation, subsequent judgments of the Supreme Court have found a way to give a go-by to mitigating evidence, invoking penological justifications such as deterrence and retribution instead.32 By imposing death sentences citing criminal justice policy goals, the Supreme Court has effectively substituted the original capital sentencing framework developed in Bachan Singh with these justifications, and made it possible to impose a death sentence on the basis of broader penological goals, without adhering to the framework at all.33

B. Onus to Produce Sentencing Material

Bachan Singh, in its sentencing framework, highlighted the need for mitigating factors to be considered while deciding between life imprisonment and the death penalty. However, it did not specify how those mitigating circumstances would be produced before sentencing courts. Aggravating factors of the crime available from case file and from the papers are presented before courts. Thus, courts have ready access to circumstances of the crime. Mitigating factors of the offender, on the other hand, involve building a social, personal and psychological history of the individual, which need to be gathered from repeated personal meetings with the offender, their family and members of their community.34 Unless this is done, courts cannot have an accurate understanding of the individual before them. It is a cardinal principle of criminal law that punishment must be individualised, and the socio-economic vulnerability of death-row prisoners often means that barely any sentencing information is presented before the judge.35 However, Bachan Singh did not clarify as to who would bring such mitigating evidence before sentencing courts.

In March 2019, the Supreme Court in Khushwinder Singh v. State of Punjab, 36 confirmed the death sentence imposed on the accused, while acknowledging that the defence had not presented any mitigating material before the court. The death sentence, therefore, was imposed only by taking into consideration aggravating circumstances. This raises significant questions as to the onus of producing, and eliciting aggravating and mitigating circumstances, relevant at the time of sentencing.

In a few instances, Supreme Court judges have also played an active role in eliciting relevant mitigating material favouring the accused, from the defence counsels.37 The Delhi High Court, in Bharat Singh v. State (NCT of Delhi), 38 got the probation officer to present a social investigation report showing the conduct of the accused in prison, before commuting the death sentence. In Mukesh, the Supreme Court decided to look at relevant mitigating material that had not been presented before the trial judge, acknowledging its relevance to sentencing.39 However, none of the judgments saw the courts elaborating upon the role of sentencing judges in cases where the defence fails to produce mitigating circumstances favouring the offender.

For a sentencing hearing to meet acceptable fair trial standards, the threshold cannot be a perfunctory conversation with the accused, or shallow statements as to their age and socio-economic status. The American Bar Association notes that the process of eliciting sentencing material is complex, and should necessarily involve skills of social workers and mental health professionals.40 Yet, inadequate legal representation resulting in production of very poor sentencing material in death sentence cases is a reality of the Indian criminal justice system. Death penalty case law (both confirmations and commutations) is rife with superficial references to sentencing factors —a consequence of both the lack of standards on collection of materials and the absence of a normative foundation for considering such materials. Despite this, however, in an adversarial system, a proactive role for the judge to elicit and seek sentencing information raises institutional concerns, as the ability of and resources available to judges, in this regard, are far from certain.41 Hence, instead of requiring the judge to undertake a roving exercise, institutional coherence might nudge us in the direction of addressing this through robust standards of legal representation for capital cases.

C. Evidentiary Standards

Bachan Singh provides no guidance on the standard of proof that is to be used for considering sentencing materials. An examination of procedural rules in other jurisdictions shows that most commonwealth countries require aggravating circumstances to be proved beyond a reasonable doubt. For instance, this is mandated in Canada by statutory requirements, and in Australia and England by judicial precedent.42 The United States sees a wide variety of prescribed evidentiary standards at sentencing across the different states, but the preponderance standard has become the most prominent alternative to the no-prescribed-burden approach. The United States Supreme Court, while deciding the right to jury trial under the Sixth Amendment in Blakely v. Washington, held that certain kinds of sentencing facts must be tried before a jury and proven beyond a reasonable doubt.43 Blakely mandates, however, attach to a narrowly defined category of aggravating facts.

The Supreme Court of India, in Santa Singh, suggested that affidavits could be used to place a wide variety of material (distinguished from ‘evidence’) that have a bearing on sentence. It clarified that if the parties disagree on the veracity of the materials, then evidence can be led as per the requirements of the law of evidence.44 This method of using affidavits to place sentencing material has been affirmed subsequently in cases like Dagdu45 and Mukesh. 46

However, evidentiary concerns at the sentencing stage rarely arise in India because sentencing submissions before courts are mostly perfunctory, and limited to the economic background of the accused, number of dependants, or the lack of criminal antecedents. However, even in the current state of sentencing, questions of reformation hold out the potential for very significant evidentiary concerns. Yet, the shallow judicial discourse on reformation has not raised any significant evidentiary questions.

D. Remedying Sentencing Errors

Bachan Singh unfortunately has nothing to say on constitutional and due process thresholds for a sentencing hearing, or the creation of remedies for deficient sentencing hearings. A prominent example of this is the issue of same day sentencing, which has received differential treatment by different judgments of the Supreme Court. While one line of cases47 recognises that same day sentencing on the date of conviction is a procedural impropriety, another set of cases holds that a sentencing hearing on a separate date is not a mandatory requirement as long as an opportunity is given to the accused to furnish evidence on sentencing.48 Further, there is significant divergence on the course of action to be adopted by appellate courts when sentencing hearings are found deficient at trial. While one line of cases has directed the remand of the case for re-trial,49 another set of appellate courts has taken it upon themselves to cure sentencing defects.50

In Mohd Arif v. Supreme Court of India, 51 the Supreme Court mandated an oral hearing of death sentence reviews, justifying it on grounds of different judicially trained minds coming to different conclusions. The idea was to add two layers of protection by directing that death sentence cases be heard by three judges in the Supreme Court and that even after the confirmation

of death sentence by the Supreme Court, the option of an oral hearing would be available. Given this context, it becomes difficult to appreciate the constitutional logic behind remedying trial court sentencing errors at the appellate stage. Moreover, this approach also deprives the accused of the right to have sentencing facts be considered by the trial court as well as two judges of the High Court. It also takes away the right to appeal against the sentencing decision of trial court and the High Court.52 Yet, the failure of Bachan Singh to anticipate such problems arising in the course of capital sentencing has given subsequent courts free rein to adopt varied approaches when confronted with procedural sentencing errors at the level of lower courts.

E. Weighing Aggravating and Mitigating Factors

At the core of the Bachan Singh framework is the identification of aggravating and mitigating factors followed by the application of judicial mind to these factors. However, Bachan Singh has very little to offer in terms of guiding judicial discretion on this aspect. Perhaps the only real assistance appears in one line of the majority opinion that requires sentencing judges to give mitigating factors a ‘liberal and expansive’ reading (the absence of such an approach for aggravating factors is instructive).53 However, it is interesting to note the miniscule attention this guiding factor in Bachan Singh has received in subsequent decisions of the Supreme Court.

The lack of any real guidance on weighing aggravating and mitigating factors has led to a crime-centric focus in sentencing, and has also resulted in some judgments outrightly dismissing any role for mitigating factors, as discussed above. Some judgments have even gone to the extent of dismissing a whole class of mitigating factors, before attempting to weigh them against aggravating factors. For instance, the Supreme Court in State of Karnataka v. Krishnappa54 held that socio-economic status, religion, race, caste, or creed of the accused or the victim are irrelevant considerations in sentencing policy. This has been further exacerbated by poor quality of sentencing material presented by the defence and lack of engagement on meaningful fair trial rights during sentencing. The role of the prosecution in leading evidence to show that the accused is beyond reformation has also not received any clarification in Bachan Singh, leading to inconsistent and arbitrary compliance by subsequent courts.

Bachan Singh, through its sentencing framework, aspired to create room for individualised capital sentencing, requiring judges to consider the role of each individual accused within their social context. However, it did not throw light on the methods for doing so, or on the normative requirement for the same. Resultantly, the very foundations of the Bachan Singh framework have been unsettled by subsequent decisions.

F. Considering the Alternative of Life Imprisonment

Besides the weighing of aggravating and mitigating circumstances, the Bachan Singh framework requires sentencing judges to consider the alternative option of life imprisonment. Noting that “a real and abiding concern for the dignity of human life postulates resistance to taking a life through law’s instrumentality”, the Court in Bachan Singh held that the death penalty could be imposed only when the alternative option of life imprisonment was ‘unquestionably foreclosed’.55 However, the question as to how this determination could be made was left open, without any further clarification. Three years later, the Supreme Court in Machhi Singh v. State of Punjab56 attempted to resolve this query and in the process introduced a much lower standard of ‘inadequacy’ for consideration of life imprisonment as opposed to the ‘unquestionably foreclosed’ standard in Bachan Singh. This has been discussed in further detail below in Section IV(e).

IV. Machhi Singh And After: A History Of Errors

The three-judge bench Supreme Court judgment in Machhi Singh that was delivered three years after Bachan Singh was the first attempt to further develop the original death penalty sentencing framework.57 Reflecting on the question of death penalty, the Court in Machhi Singh delved into reasons for why the community as a whole does not endorse the humanistic approach of ‘death sentence-in-no-case’. Machhi Singh introduced ‘collective conscience’ into the capital sentencing framework and laid down five categories where the community would expect the holders of judicial power to impose death sentence, because collective conscience was sufficiently outraged. These five categories include motive of the crime, manner of its commission, anti-social or socially abhorrent nature of the crime, magnitude of the crime, and personality of the victim of the murder.58 Instances were listed under each of these categories. Notably, the examples cited by the court point towards specific crime categories, an approach that Bachan Singh specifically guarded against.

Particularly, crime-centric categories warranting the imposition of death sentence were laid down, public opinion was introduced as a relevant sentencing factor, the role of mitigating factors was substantially diluted by the Court, the manner of consideration of aggravating and mitigating factors was altered, and the standard to rule out the alternative punishment of life imprisonment was significantly lowered.

This section discusses broadly the formulations laid down by Machhi Singh, and then goes on to thematically discuss the different arguments which flow from a misreading of Bachan Singh’s framework by several judgments of the Supreme Court post-Machhi Singh.

Details of how each of these aspects has been dealt with by the Supreme Court in subsequent judgments are discussed below.

A. Crime-Centric Framework

The five categories which Machhi Singh introduced to guide sentencing courts while imposing the death penalty include motive of the crime, manner of its commission, anti-social or socially abhorrent nature of the crime, magnitude of the crime, and personality of the victim of the murder.59 This completely undermines the Bachan Singh framework which explicitly warned against categorization of offences qualifying the death sentence,60 and is at odds with the determination of culpability based both on the circumstances of the offence and the offender.

Following Machhi Singh’s crime-centric approach to death penalty sentencing, there have been a spate of cases that have suggested that death sentence can be imposed only on the basis of circumstances of the crime. In Ravji, the Supreme Court held that the nature and gravity of the crime, and not the criminal, were central to the question of deciding appropriate punishment.61 The Court opined that punishment should conform to, and be consistent with, the atrocity and brutality of the crime, as well as the public abhorrence it warrants, and that courts should respond to society’s cry for justice against the criminal.62 Ravji was examined subsequently by a division bench of the Supreme Court in Santosh Kumar Satishbhushan Bariyar v. State of Maharashtra and rendered per incuriam Bachan Singh, for its exclusive focus on crime.63 The court in Santosh Bariyar listed six more cases which had relied on the incorrect precedent in Ravji.64

In Shankar Kisanrao Khade v. State of Maharashtra, 65 the Court doubted the imposition of the death penalty in several cases for their failure to appreciate the circumstances of the accused, including that of Dhananjoy Chatterjee, 66 which had resulted in execution of the accused. Similarly in Sangeet v. State of Haryana, 67 the Court mentioned four cases, Shivu v. High Court of Karnataka, 68 Rajendra Pralhadrao Wasnik v. State of Maharashtra, 69 and Mohd Mannan v. State of Bihar, 70 where determination of sentence was made only on the basis of the circumstances of the crime, without taking the circumstances of the accused into consideration.

At the core of this misinterpretation of Bachan Singh lies the lack of any normative clarity in the judgment on the relevance of mitigating factors in deciding punishment beyond merely saying that the “scope and concept of mitigating circumstances must receive liberal and expansive construction”.71 Placing immense value on the concept of mitigation without offering any doctrinal basis for the same along with the lack of procedural clarity on whose duty it is to present this evidence and how it is to be collected, presented, and weighed, has incentivised sentencing judges to rely on Machhi Singh’s easy-to-implement crime-centric framework. Though the guidelines in Machhi Singh are at odds with Bachan Singh, its five categories with numerous instances under each category offer a very practical guide to sentencing judges. The popularity of Machhi Singh’s framework is a direct outcome of the highly ambitious yet extremely ambiguous framework in Bachan Singh.

B. Weighing v. Balancing of Aggravating and Mitigating Circumstances

As per the Bachan Singh framework, sentencing courts are required to weigh both aggravating and mitigating circumstances of the offence and the offender before deciding the punishment. Relative weight is to be attached to each of these factors and has to depend on the facts and circumstances of the particular case.72 Before Bachan Singh laid down this framework, the judgment in Jagmohan had held that judges should decide the appropriate punishment after ‘balancing’ circumstances of the crime as well as of the criminal.

Building on Jagmohan’s concept of ‘balancing’ of aggravating and mitigating circumstances, Machhi Singh introduced a balance-sheet approach, and required courts to draw up a balance-sheet, giving full weightage and striking a just balance between aggravating and mitigating factors.73 The lack of clarity regarding treatment of aggravating and mitigating factors by Bachan Singh itself, discussed in the previous section, has further compounded the problem by rendering ambiguous the treatment of mitigating and aggravating factors by sentencing courts.

The apparently minor change of vocabulary from Jagmohan to Bachan Singh, requiring the ‘weighing’ and not ‘balancing’ of aggravating and mitigating circumstances had grave implications when Machhi Singh attempted to build on the Bachan Singh framework. An exercise of ‘balancing’ gives the option to balance out mitigating and aggravating factors against each other, discharging sentencing courts of the duty to assign reasons for apportionment of weight to each relevant factor. ‘Balancing’ takes away from the judicial rigour required in a weighing exercise. The tragic impact of this has been courts, at all levels, imposing death sentences after a listing of aggravating and mitigating factors and stating that mitigating factors are balanced out by the brutality of the crime. ‘Balancing’, essentially, has allowed courts to skip the very important aspect of apportioning weight to each factor along with reasons for the same.

The use of words ‘balance sheet’ in Machhi Singh also paved the way for subsequent sentencing courts to simply list aggravating and mitigating circumstances in a tabular format, in two columns against each other, followed by the conclusion that aggravation outweighs mitigation. This has rendered the exercise meaningless where the focus is more on meeting the technical requirements, rather than an actual meaningful consideration of aggravating and mitigating factors. The fact that courts generally have access to aggravating factors of the crime through the case records as opposed mitigating factors of the offender which need further investigation makes this extremely problematic. Given this reality, aggravating factors are numerically higher than mitigating factors in most cases, making it easier for courts to ‘balance’ away mitigation towards the imposition of the death sentence. In Sangeet, 74 the Supreme court critiqued the balance sheet approach. It was of the view that the circumstances of the crime and the criminal are completely distinct and different elements and cannot be compared with one another. It further noted that Bachan Singh had discarded this proposition in Jagmohan Singh but Machhi Singh revived it.75

The approval of the ‘balancing approach’ can partly be attributed to the absence of any procedural clarity on the weighing of aggravating and mitigating factors in Bachan Singh. Firstly, this requires clear identification of roles of defence lawyers and the courts in the collection, presentation, and consideration of mitigating factors. Secondly, courts need much clearer guidance on what the process of ‘weighing’ entails, and what the Court in Bachan Singh meant when it observed that the ‘relative’ weight to be attached to aggravating and mitigating factors depends on the facts and circumstances of the case. Absent a clear normative foundation for the very idea of mitigation, Bachan Singh left it open to future sentencing judges to adopt a wide variety of rationale to undermine the very purpose of mitigation. Further, procedural implication of what it means to give liberal and expansive construction to mitigating and not aggravating factors needs to be spelt out. The Court in Santosh Bariyar did attempt to build on the concept of weighing by stipulating that “the weight which is accorded by the court to particular aggravating and mitigating circumstances may vary from case to case in the name of individualized sentencing, but at the same time reasons for apportionment of weights shall be forthcoming.”76 However, this was not picked up or further developed by subsequent decisions.

Balancing out the mitigating factors by the aggravating ones without having to go through the rigorous exercise of assigning weight to every relevant aggravating and mitigating factor along with stating the reasoning for the same is a very convenient escape for courts without any consequences given the lack of clarity on the due process and its violation.

C. Uncertain Role of Mitigating Factors

Machhi Singh stated that two questions have to be considered before imposing the death sentence—first, whether the sentence of life is inadequate; and second, whether there is no alternative but the death sentence despite maximum weightage to the mitigating factors. In his dissenting opinion in Manoharan v. State,77 Justice Sanjiv Khanna stated that the five categories elucidated in Machhi Singh, if carefully analysed, relate to the first question to be posed and answered. But this is not the only question that the court must answer, for the second question has to be also answered in order to direct or uphold the death penalty. According to Justice Khanna, the second question can be answered with reference to Bachan Singh which lays down a non-exhaustive list of mitigating factors.

However, the status assigned to mitigating circumstances of the offender in Machhi Singh remains conspicuous given Machhi Singh’s own treatment of mitigating factors of the offender in the criminal appeals before it. The lack of clarity in Bachan Singh on methods of collecting mitigating factors and on the normative requirement for the same in deciding outcomes in capital cases also enabled the Court in Machhi Singh to rely only on the five crime categories it laid down to confirm the death sentences. The judgment did not even remotely make a mention of the circumstances of the offenders, leave alone meaningfully consider them. Further, the meaning assigned to Machhi Singh by Justice Khanna was neither adopted by the majority in that case, nor has been relied upon in subsequent decisions in death penalty cases.

D. Introduction of Public Opinion in Death Penalty Sentencing

The introduction of ‘collective conscience’ into the capital sentencing framework by Machhi Singh made way for the entry of public opinion in deciding on the question of sentence, completely in contradiction to the formulation in Bachan Singh. It paved the way for other similarly amorphous standards into the scheme. In Dhananjoy, the Supreme Court, while imposing a death sentence, held that appropriate punishment enables courts to respond to ‘society’s cry for justice’.78 This measure of punishment, in turn, must depend upon the atrocity of the crime, the conduct of the criminal, and the defenceless and unprotected state of the victim.

In Santosh Bariyar, the Supreme Court observed that public opinion was incompatible with the Bachan Singh framework, since the constitutional role of the judiciary mandates placing individual rights at a higher pedestal than majoritarian aspirations.79 Another inherent problem it identified with this approach was the difficulty to precisely define what public opinion on a given matter actually is.80

In Rameshbhai Rathod v. State of Gujarat, 81 the Supreme Court was of the view that the expression ‘rarest of rare’ was used in Bachan Singh to read down and confine the imposition of capital punishment to extremely limited cases. Hence, the significance of this expression could not be watered down on a perceived notion of a ‘cry for justice’.82 In Om Prakash v. State of Haryana, 83 the Court observed that there was a significant tension between responding to society’s cry for justice and Bachan Singh’s sentencing framework, and held that courts are bound by precedent and not by the incoherent and fluid responses of society.84 Recently, in 2018, the Supreme Court commuted death sentences in MA Antony v. State of Kerala85 and Chhannu Lal Verma v. State of Chhattisgarh, 86 noting problems with imposing punishment based on collective conscience.

Despite these concerns, however, the Supreme Court, in some cases, continues to impose death sentences invoking public opinion as a justification. Collective conscience found its most recent endorsement in the Supreme Court judgment in the December 2012 gang rape case of Mukesh v. State (NCT of Delhi), and continues to be used rampantly across trial and appellate courts in India.87

The Court in Bachan Singh did not delve into the role of public opinion while sentencing. While reflecting on public opinion in the context of the constitutionality of the death penalty, the Court noted that judges should not become oracles of public opinion.88 However, the increasing public clamour for the death penalty in response to heinous crimes, along with the uncertain role of mitigating factors, has resulted in an ever expanding determinative role of public opinion in capital sentencing.

E. ‘Unquestionably Foreclosed’ to ‘Inadequacy’ of Life Imprisonment

Bachan Singh, reflecting the legislative intent behind making the death penalty an exceptional punishment, mandated that death sentence can only be imposed if the alternative of life imprisonment is ‘unquestionably foreclosed’.89 The judgment in Bachan Singh did not, however, clarify how such determination was to be made. Three years later Machhi Singh used the vocabulary ‘inadequacy’ in the context of life imprisonment instead of reiterating the Bachan Singh’s unquestionably foreclosed standard.90 This seemingly subtle shift had the very consequential impact of lowering the standard for consideration of life imprisonment from it being ‘unquestionably foreclosed’ to one of ‘inadequacy’. The Court in Machhi Singh observed, “death sentence must be imposed only when life imprisonment appears to be an altogether inadequate punishment having regard to the relevant circumstances of the crime, and provided, and only provided, the option to impose a sentence of imprisonment for life cannot be conscientiously exercised having regard to the nature and circumstances of the crime and all the relevant circumstances.”91 Such a framing of this standard by Machhi Singh makes it seem like one of a much lower threshold, and dependent on the circumstances of the crime. It has pushed the death penalty sentencing jurisprudence towards judges examining whether life imprisonment would be adequate for the crime in question. The extensive use of Machhi Singh by sentencing judges has exacerbated this error, with judges often relying on the description of the crime to come to the conclusion that life imprisonment would be ‘inadequate’, rather than establishing that life imprisonment was ‘unquestionably foreclosed’.

In Santosh Bariyar, the Supreme Court interpreted the question of life imprisonment within the context of reformation and clarified that “life imprisonment can be said to be completely futile only when the sentencing aim of reformation can be said to be unachievable.”92 The Court also imposed a duty on the court “to provide as to why the convict is not fit for any kind of reformatory and rehabilitation scheme”.93

A significant development came in 2015, with the widening of the ‘unquestionably foreclosed’ standard, by a constitution bench of the Supreme Court in Union of India v. V Sriharan. 94 Reaffirming its ruling in Swamy Shraddanada (2) v. State of Karnataka, 95 the Supreme Court held that it is open to the appellate courts to impose a life sentence for the rest of the prisoner’s natural life, without any possibility of review or parole, in cases where death is one of the statutorily prescribed punishments. The court held that the State government’s power of remission under Section 432 of the CrPC could be ousted while determining the sentence in an appellate court. Constitutional powers of pardon of the Governor and President under Articles 161 and 72 of the Constitution, respectively, remained untouched. Two dissenting judges in the case found the formulation to be a violation of the separation of powers. The Sriharan sentencing formulation is supposed to be a middle ground between death and a normal life sentence (which makes a prisoner eligible for consideration for remission after fourteen years). Interestingly, in Sangeet, the court had expressly disagreed with the formulation of the Supreme Court in Swamy Shraddananda, which subsequently found affirmation with the constitution bench.96 Notably, only appellate courts have the power to impose this sentence under Sriharan.

Machhi Singh’s standard of ‘inadequacy’ has altered the sentencing courts’ duty fundamentally, requiring them to answer the question of lifeimprisonment in light of the circumstances of the crime. This essentially makes the Bachan Singh framework redundant, since all crimes punishable with death are likely to involve significant levels of brutality and heinousness. A discussion in Bachan Singh on the normative and procedural components of the ‘unquestionably foreclosed’ standard would have gone a long way in avoiding this problem.

V. Repairing India’s Broken Capital Sentencing Framework

The sentencing framework developed in Bachan Singh offered a transformative potential for the death penalty jurisprudence in India which was not sufficiently utilised by the subsequent judgments. The essence of this framework lies in the crucial embracing of the spirit of individualised justice under Section 235(2) of the CrPC by emphasising on the questions of culpability of an individual and proportionate punishment, and stressing on the relevance of mitigating factors with a liberal and expansive construction. The incredibly high standard for ruling out life imprisonment and imposing death sentence in Bachan Singh truly embodied the legislative mandate in Section 354(3) of the CrPC. Despite its inherent limitations, the decision was a very significant and rich addition to death penalty jurisprudence in India.

It is quite unfortunate that rather than attempting to further develop the Bachan Singh framework by resolving the doctrinal and normative concerns that afflict the framework, future benches of the Supreme Court have distorted it by offering varied interpretations, some of which go against the very grain of Bachan Singh. Forty years since its origin, the original framework has been twisted partly because of its own normative and procedural deficiencies. The lack of a coherent doctrinal framework has allowed for the sidelining of mitigating factors. Without any normative clarity on the scope and concept of mitigating factors, subsequent benches have not been able to appreciate their role and relevance in determining culpability and, therefore, the punishment. Despite the significant role assigned to mitigating factors in Bachan Singh, various sentencing courts have indulged in serious fair trial rights violation of the accused by imposing death sentence without any, or meaningful, consideration of mitigating factors. Absence of any indication on the elements of the ‘unquestionably foreclosed’ standard to rule out life imprisonment as the alternative has diluted the legislative mandate of life imprisonment being the normal punishment under Section 302 of the IPC. Death sentences have often been imposed without completely foreclosing the option of life imprisonment.

Besides the normative deficiency, there is procedural ambiguity in the different processes involved. Sentencing courts have no clarity on identification, presentation, and weighing of individual aggravating and mitigating factors and on meaningful consideration of the option of life imprisonment. Later judgments by the Supreme Court have not made any attempts towards bridging these normative and procedural gaps. While some decisions have recognised the relevance of individual mitigating factors, the lack of a doctrinal basis to appreciate the collection, presentation, and consideration of mitigating factors continues to adversely affect the capital sentencing framework.97

Undoubtedly arbitrary application of the Bachan Singh framework is a major concern. But we need to unpack the meaning of arbitrariness in this context. An understanding that mainly invokes the framework that ‘similar cases resulting in different outcomes’ does not sufficiently capture the nature of arbitrariness in death penalty sentencing. ‘Similar cases’ in our legal and judicial discourse tends to lean heavily towards just looking at the crime and whether there have been dissimilar outcomes. The attempt in this paper has been to demonstrate that there are deeper normative and procedural concerns that we must pay a lot more attention to in death penalty sentencing. Perhaps the very first step would be to examine closely the manner in which sentencing information is brought before courts, and the minimum constitutional thresholds they must adhere to. In that context, the systemic reality of most death row prisoners being extremely poor and further burdened by abysmal quality of legal representation presents a huge challenge at the very first step.98

Absence of a meaningful judicial discourse on capital sentencing has contributed to the arbitrary and unpredictable imposition of death sentences. In its current shape and form, the capital sentencing framework poses a serious threat to the right to fair trial of the accused. Any effort towards repairing the broken nature of capital sentencing in India to ensure meaningful realisation of fair trial rights of the accused requires filling the normative and procedural gaps that have been highlighted in this paper. While addressing the normative deficiencies might be a long and drawn out process, the serious and irreversible nature of the death sentence should drive the Supreme Court to take immediate steps towards establishing clear sentencing procedures and identifying the remedies available to the accused when the due process is violated. The majority opinion in Bachan Singh gave very sound constitutional advice when it said that “a real and abiding concern for the dignity of human life postulates resistance to taking a life through law’s instrumentality”. The taking of life through a legal process ought to be very tough and must adhere to the highest standards of fair trial rights and the rule of law. The endeavour cannot be to make it easy and convenient to impose the death sentence.

Anup Surendranath, Neetika Vishwanath and Preeti Pratishruti Dash are with Project 39A at National Law University, Delhi in India. This article first appeared in the National Law School of India Review. Views are personal only.