LLM Group 5 Criminal Law and Criminal Administration Penology Treatment of Offenders 2019 Question Paper with Solutions

The right to be confined lawfully and no more. Sunil Batra v. Delhi Administration, (1978) 4 SCC 494, decided on 30 August 1978, is the foundation. A prisoner under sentence of death challenged section 30(2) of the Prisons Act, 1894, which permits confinement in a cell apart, and section 56, which permits the imposition of irons. The Court refused to strike either down and instead read both narrowly: section 30(2) authorises safe custody and not solitary confinement, and applies only once the death sentence is finally executable, so a convict whose appeal or mercy petition is alive may not be isolated; and bar fetters under section 56 may be imposed only for the shortest necessary period, on reasons recorded, subject to review, after hearing the prisoner, and never as a punishment.

The right to a remedy for conditions, not merely for detention. Sunil Batra (II) v. Delhi Administration, (1980) 3 SCC 488 treated a letter alleging that a warder had tortured a fellow inmate by forcing a rod into his anus to extort money from his relatives as a writ of habeas corpus, although nobody sought release. It held that the writ lies against unlawful conditions of a lawful detention, appointed amici to inspect the prison, and directed District Judges to act as visitors and hear grievances.

The right to live with dignity, and to see a lawyer and a family. Francis Coralie Mullin v. Administrator, Union Territory of Delhi, (1981) 1 SCC 608, decided on 13 January 1981, arose from the detention of a British national under the Conservation of Foreign Exchange and Prevention of Smuggling Activities Act, 1974, who was refused interviews with her legal adviser and her family on reasonable terms. The Court held that the right to life in Article 21 is not confined to physical existence but includes the right to live with human dignity, and that a detenu's right to consult a legal adviser, not merely for a defence in a criminal proceeding but to secure release from preventive detention or to file a writ petition, is part of personal liberty. It is the case to cite where the complaint is about access rather than about violence.

The right not to be fettered or handcuffed as a matter of routine. Prem Shankar Shukla v. Delhi Administration, (1980) 3 SCC 526 held handcuffing to be prima facie inhuman, unreasonable and arbitrary, and required reasons to be recorded and shown to the presiding judge. Citizens for Democracy v. State of Assam, (1995) 3 SCC 743 hardened that into a rule that a prisoner may not be handcuffed or fettered in transit or in hospital without the court's permission, and made a breach a contempt.

The right to equal treatment, and to be free of caste classification. Sukanya Shantha v. Union of India, 2024 INSC 753, decided on 3 October 2024, examined State prison manuals that allotted sweeping and scavenging by caste, segregated barracks and treated denotified tribes as born criminals. The Court declared those provisions unconstitutional under Articles 14, 15, 17, 21 and 23, gave States three months to revise their manuals, directed the Union to amend the Model Prison Manual 2016 and the Model Prisons and Correctional Services Act 2023, ordered the caste column and every caste reference deleted from undertrial and convict registers, and confined "habitual offender" in a manual to the definition in the State's own legislation.

The right to equitable wages for work. State of Gujarat v. Hon'ble High Court of Gujarat, (1998) 7 SCC 392, decided on 24 September 1998, held that hard labour lawfully imposed as part of a sentence does not create an employer and employee relationship, so the Minimum Wages Act does not apply of its own force, but that prisoners must be paid equitable wages at rates fixed by the State, and recommended that a portion be set apart as compensation for victims.

The conditions are cumulative and precise. The person must have been previously convicted by a court in India of an offence punishable with imprisonment for three years or upwards; he must be again convicted of an offence punishable with imprisonment for three years or upwards; and the conviction must be by a court other than that of a Magistrate of the second class. Where those conditions are met the court may, if it thinks fit, at the time of passing a sentence of imprisonment, also order that his residence and any change of, or absence from, that residence after release be notified for a term not exceeding five years from the date the sentence expires.

The safeguards are as important as the power. By section 394(3) the order becomes void if the conviction is set aside on appeal or otherwise. By section 394(2) the section applies also to criminal conspiracies to commit such offences and to abetments and attempts. By section 394(4) the order may also be made by an appellate court, the High Court or a Court of Session in revision. By section 394(5) the State Government may make rules to carry out the notification requirement, and by section 394(6) those rules may provide for punishment of a breach, triable by a Magistrate of competent jurisdiction in the district where the offender's place of residence is.

The direction of the modern law is nevertheless against status-based control, and this is where the note should end. Sukanya Shantha v. Union of India, 2024 INSC 753 held that references to "habitual offenders" in prison manuals must conform to the definition in the habitual offender legislation of the State legislature and declared every other manual definition unconstitutional, because such definitions had been used to brand members of denotified tribes as born criminals in violation of Articles 14, 15, 17, 21 and 23; and it directed the police to follow Arnesh Kumar v. State of Bihar and Amanatullah Khan v. Commissioner of Police, Delhi so that such persons are not arbitrarily arrested. A surveillance order that answers to section 394's conditions survives that reasoning; an administrative register of "suspects" does not.

For full marks, cover: the question asks for merits and demerits with case law, so every claim on either side must be attached to a decision or to a finding rather than asserted. The most economical way to organise it is round the five questions a legislature would have to answer before deciding whether to keep the punishment: does it deter, is it proportionate, can it be corrected, is it applied equally, and is there an adequate alternative. That structure puts merits and demerits side by side on the same issue instead of listing them separately, which is what "discuss" invites. The same subject returns at Q11 of the second paper on this scan, where the question is framed round a quotation about commutation and is answered on the commutation process.

The claim against. The 262nd Report of the Law Commission of India, submitted on 31 August 2015 under Justice A.P. Shah, recommended abolition for all offences other than terrorism-related offences and waging war, and its reasoning is the answer to this question: deterrence depends far more on the certainty of detection and conviction than on the severity of the sentence. The Commission went further and said there is no penological justification for treating terrorism differently, accepting the exception only because legislators feared for national security. Three members dissented. A punishment retained on a rationale its own official reviewers describe as political rather than penological is weakly defended on deterrence.

Question three: can it be corrected?

There is no claim for. The punishment is unique in being irreversible, and the whole weight of this question falls on the demerit side.

The claim against. Santosh Kumar Satishbhushan Bariyar v. State of Maharashtra, (2009) 6 SCC 498, decided on 13 May 2009, is the Supreme Court's own finding that it has erred. It held Ravji v. State of Rajasthan, (1996) 2 SCC 175 and six decisions following it to have been rendered per incuriam Bachan Singh, because they confined the sentencing inquiry to the crime and ignored the criminal and the possibility of reform; the six are Surja Ram v. State of Rajasthan, (1996) 6 SCC 271, Dayanidhi Bisoi v. State of Orissa, (2003) 9 SCC 310, Mohan Anna Chavan v. State of Maharashtra, (2008) 7 SCC 561, Bantu v. State of Uttar Pradesh, (2008) 11 SCC 113, Shivaji @ Dadya Shankar Alhat v. State of Maharashtra, (2008) 15 SCC 269, and State of Uttar Pradesh v. Sattan @ Satyendra, (2009) 4 SCC 736.

The claim against. The offender who is poor, unrepresented or illiterate is least able to place mitigating material before the court, and mitigation is exactly what Bachan Singh requires. That deficit has now been recognised at the highest level. Aman Singh v. State of Bihar, 2026 INSC 424, decided on 27 April 2026 by Vikram Nath, Sandeep Mehta and Vijay Bishnoi JJ., stayed an execution and directed that the trial court itself call for a report on aggravating and mitigating circumstances as a matter of course once the accused is convicted and before sentence, that such reports be comprehensive, verified and delivered within a stipulated time, that the National Legal Services Authority frame guidelines for gathering mitigating material through trained interdisciplinary teams including mental health and social science professionals, and that skilled and adequately resourced legal representation be provided.

The incapacitation merit used to be decisive: only execution ended the risk absolutely. It is now largely met. Swamy Shraddananda (2) v. State of Karnataka, (2008) 13 SCC 767 created a special category between ordinary life imprisonment and death, allowing a fixed term well beyond fourteen years or life without remission. Union of India v. V. Sriharan, (2016) 7 SCC 1, a Constitution Bench, upheld it, holding that imprisonment for life means imprisonment for the rest of the convict's life and that the High Court and the Supreme Court, though not the trial court, may place such a sentence beyond remission. Ramasrey @ Fakkad v. State of Uttar Pradesh, 2026 INSC 764, decided on 30 July 2026 by Sanjay Karol and A.G. Masih JJ., rejected a fresh constitutional challenge to that category, holding that a two-judge bench cannot revisit a Constitution Bench and describing the whole-life sentence as a special category lying between ordinary life imprisonment and death.

For full marks, cover: the same topic is set at Q13 of the second paper on this scan, where the question has two express limbs and the answer is arranged accordingly. Here take a different route and use the nine problems the Supreme Court itself identified in Ramamurthy v. State of Karnataka, (1997) 2 SCC 642 as the frame, giving the history of each and its position today. That is a genuinely useful structure because it is the Court's own list, it covers the whole subject, and it lets you finish with a verdict on how many of the nine have actually been solved. Keep the committee names and dates: an answer without Cardew, Reckless, Mulla and Krishna Iyer is incomplete.

The list is also a fair summary of what a century of committees had already said, which is why it works as a frame. The Prison Discipline Committee of 1836, appointed on Macaulay's minute and reporting in 1838, had expressly rejected reformation as an object; the Indian Jails Committee of 1919-20 under Sir Alexander Cardew reversed that and made reformation the official aim; Dr Walter C. Reckless, the United Nations expert who studied Indian prisons in 1951 and 1952, urged in Jail Administration in India that jails become reformation centres and that the manuals be rewritten; the Mulla Committee, constituted in 1980 and reporting in 1983, produced the fullest programme; and the National Expert Committee on Women Prisoners under Justice V.R. Krishna Iyer, constituted in 1986 and reporting in May 1987, addressed the category the others had ignored.

In Re: Policy Strategy for Grant of Bail, 2025 INSC 239, decided on 18 February 2025 by Oka and Bhuyan JJ., attacked the convict end. Where a State has a remission policy the Government is obliged to consider every eligible convict for premature release as and when he becomes eligible, without any application from him or his relatives; States and Union Territories with no policy were directed to frame one within two months; the conditions attached to a remission must be capable of performance, not vague and not oppressive; and an order granting or refusing remission must carry brief reasons, be communicated to the convict through the prison and copied to the District Legal Services Authority, with the convict told that he may challenge a refusal.

Indian law nevertheless still diverges from the international minimum on one point. Rule 44 of the Nelson Mandela Rules, adopted on 17 December 2015 by General Assembly resolution 70/175, defines solitary confinement as confinement for twenty-two hours or more a day without meaningful human contact and prolonged solitary confinement as a period in excess of fifteen consecutive days, and Rule 45(1) provides that it shall be used only exceptionally, as a last resort, subject to independent review, and that it shall not be imposed by virtue of a prisoner's sentence. Section 11 of the Bharatiya Nyaya Sanhita, 2023 permits the sentencing court to order up to three months of solitary confinement as part of the sentence, and section 12 caps a single spell at fourteen days.

Problem six: prison vices, which is really classification

Classification is where the worst abuse was found, and the decisive decision is recent. Sukanya Shantha v. Union of India, 2024 INSC 753, decided on 3 October 2024 by Chandrachud CJI, Pardiwala and Manoj Misra JJ., examined manuals that allotted sweeping and scavenging by caste, segregated barracks by caste and treated denotified tribes as born criminals, and declared those provisions unconstitutional under Articles 14, 15, 17, 21 and 23. It gave States three months to revise their manuals, directed the Union to amend the Model Prison Manual 2016 and the Model Prisons and Correctional Services Act 2023, ordered the caste column and every reference to caste deleted from undertrial and convict registers, confined "habitual offender" in a manual to the definition in the State's own habitual offender legislation and struck down every other manual definition, and took suo motu cognizance as In Re: Discrimination Inside Prisons in India.

Problem nine: open air prisons

This is the problem on which the most has happened most recently. The Pakwasa Committee of 1949 had recommended using prisoners on road work without guard, and Rajasthan built the model. Suhas Chakma v. Union of India, 2026 SCC OnLine SC 317, decided on 26 February 2026 by Vikram Nath and Sandeep Mehta JJ., issued seven directions on Open Correctional Institutions: no reduction in the area or capacity of a functioning open camp, feasibility studies where none exist, transfer policies for women prisoners with gender-sensitive criteria and an end to blanket exclusions, eligibility based on conduct, self-discipline and reformative progress rather than on rigid sentence fractions or the gravity of the original offence, uniform minimum standards from the Union on wages, healthcare, education, family contact and discipline, vocational training beyond agriculture, and medical facilities inside the institutions so that a sick prisoner is not routinely returned to a closed jail. The Court recorded a daily cost of about Rs 49.60 per prisoner in a Rajasthan open prison against about Rs 333.12 in a closed one, and open institutions running at only 6 to 44 per cent of capacity.

The scorecard, and the reform that never arrived

The Mulla Committee recommended a National Prison Commission and an all-India Indian Prisons and Correctional Service in 1983; neither exists. The Krishna Iyer Committee recommended separate custodial facilities for women in every State in 1987; there are 34 women's jails in the country and 21 States and Union Territories have none. And the Model Prisons and Correctional Services Act, 2023, circulated to the States on 10 May 2023 to replace the Acts of 1894, 1900 and 1950, is a model only: prisons are Entry 4 of the State List, and the Centre has told Parliament that no State has confirmed its adoption.

Conclusion. Measured against the Supreme Court's own list of nine problems, Indian prison reform has solved the doctrinal ones and left the material ones. Torture, fetters, handcuffs and caste classification have all been dealt with by judicial decision, from Sunil Batra in 1978 to Sukanya Shantha in 2024; overcrowding, undertrial detention, health and staffing remain what they were, and the Cardew Committee's capacity ceiling of 1920 and the Mulla Committee's Prison Commission of 1983 are still unimplemented. The two decisions of 2025 and 2026, In Re: Policy Strategy for Grant of Bail and Suhas Chakma, attack the problem from the release end rather than by building prisons, which is the most promising direction available, and the reform still missing is a modern prisons statute actually in force.

Probation as a form of control, through the cases

Probation is a disposal in which the court convicts the offender but withholds punishment, releasing him on a bond to keep the peace and be of good behaviour, usually under supervision. It is described as new because supervision, which is what distinguishes it from a bare suspended sentence, dates only from 1841, when John Augustus, a Boston bootmaker, stood surety for a man charged with drunkenness and undertook to supervise him; Massachusetts made it statutory in 1878, England created a service by the Probation of Offenders Act, 1907, and India obtained a complete national scheme only with the Probation of Offenders Act, 1958, section 19 of which displaced section 562 of the Code of Criminal Procedure, 1898 wherever it was brought into force.

The case law establishes three things about its reach and its limits.

Nagarajan v. State of Tamil Nadu, 2025 INSC 703, decided on 15 May 2025 by Dipankar Datta and Manmohan JJ., establishes the limit. Parliament reversed the result of Ishar Das on its own subject by inserting section 20AA into the Prevention of Food Adulteration Act in 1976, providing that neither the Probation Act nor section 360 of the Code applies to a person convicted under that Act unless he is under eighteen. The Court held the bar unambiguous, that probation was categorically unavailable for such offences committed between 1976 and the Act's repeal in 2006, and that neither the reformative principle nor a constitutional argument could displace an express exclusion. The composite rule is therefore that a minimum sentence does not exclude probation but an express bar does, and the other express bars are section 19 of the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989 and section 16A of the Protection of Civil Rights Act, 1955.

Asfaq v. State of Rajasthan, (2017) 15 SCC 55, decided on 11 September 2017, is the case that fixes the difference and the one to cite. The Court explained that when a prisoner is released on parole, the period of release does not count towards the total sentence, whereas when he is on furlough the period does count as sentence undergone. It set both in a reformative frame, saying that once reformation is accepted as an object of sentencing there is justification for releasing even life convicts for short periods so that they can attend to personal and family problems and maintain their links with society. It balanced that against competing public interests, holding that a person who is a habitual offender, or who may commit crime again on release, or who would be a threat to law and order, should not be released.

The distinction, set out

Two further contrasts are worth a line each. Probation is available only where the offence is not excluded and the offender qualifies, so it is a front-end device that keeps a person out of prison; parole is a back-end device that lets a person out of prison temporarily, and it therefore reaches offenders whom probation could never have helped. And probation is judicial, so it is reasoned, recorded and appealable, whereas parole is administrative, which is why the fairness of the parole decision has had to be secured by the courts rather than by the statute.

They also share a common weakness, which is administrative. Probation depends on a probation officer being available to make the report that section 4(2) and section 6(2) require, and where a State has too few officers the Act cannot be worked. Parole depends on rules that differ from State to State and on an executive decision that is often unreasoned; In Re: Policy Strategy for Grant of Bail, 2025 INSC 239, decided on 18 February 2025, dealt with the related question of premature release and held that an order granting or refusing it must carry brief reasons, be communicated to the convict through the prison and be copied to the District Legal Services Authority, which is the standard the parole decision should also be held to.

For full marks, cover: the words "with respect to their utility in the present system" are the whole question and most answers ignore them, describing the two systems at length and adding two sentences of evaluation. Reverse the proportions. Set out each system briefly and accurately, then take its distinctive ideas one at a time and ask what Indian law has done with each: kept it, transformed it, or constitutionally repudiated it. That gives you three verdicts to defend and turns a descriptive question into an argument.

The classical Hindu approach

The organising concept is danda, literally the rod, which stands both for the punishment and for the sovereign's power to inflict it. The Manusmriti treats danda as the instrument by which the king maintains dharma, and the Arthashastra of Kautilya treats it additionally as an instrument of statecraft, with a detailed tariff of fines for civil and administrative wrongs.

Hadd offences infringe the right of God. They are a small and closed list, including theft, unlawful intercourse, false accusation of unchastity, drinking, apostasy and highway robbery, and their punishments are fixed by revelation. Because the claim belongs to God, no human being may pardon or compromise a hadd offence; but the evidentiary requirements are correspondingly severe and doubt defeats the charge, so the category is narrow in practice as well as in definition.

Qisas offences are homicide and wounding, and they infringe the right of the individual and his family. The primary sanction is exact retaliation, but the heirs may forgive, or may accept diyat, blood money, in place of retaliation. Ownership of the claim therefore lies with the victim's family and not with the State.

Tazir covers everything else, where neither the offence nor the penalty is fixed by revelation and the qadi or the ruler determines the punishment according to the circumstances of the offence and the offender. It is discretionary, individualised and by far the largest category.

Utility, verdict one: what Indian law has kept

The expiatory idea has survived, and it is the most useful of the four. The proposition that punishment discharges the wrong, so that the offender returns with his standing restored, is the only classical idea that says anything about the day after the sentence ends. Its modern statutory form is section 12 of the Probation of Offenders Act, 1958, which provides that a person dealt with under section 3 or section 4 shall not suffer any disqualification attaching to a conviction under any other law unless he is later sentenced for the original offence. That is expiation drafted into a statute: the sentence, once served, is not to follow the offender through his working life.

The tazir idea of individualised discretionary sentencing is now the Indian norm. Bachan Singh v. State of Punjab, (1980) 2 SCC 684 requires the sentencer to weigh aggravating and mitigating circumstances relating to both the crime and the criminal, and Santosh Kumar Satishbhushan Bariyar v. State of Maharashtra, (2009) 6 SCC 498 held Ravji v. State of Rajasthan, (1996) 2 SCC 175 and six decisions following it to be per incuriam for looking at the crime alone. That is the tazir principle in modern dress.

Utility, verdict two: what Indian law has transformed

The duty to punish has become the duty to give reasons. The classical Hindu position that the king must punish, and sins if he does not, has no place in a system of separated powers; what has replaced it is a set of statutory obligations to explain a sentencing choice. Section 393(3) of the Bharatiya Nagarik Suraksha Sanhita requires special reasons for a death sentence; section 402 requires special reasons where a court could have used probation or the Juvenile Justice (Care and Protection of Children) Act, 2015 and did not; section 6 of the Probation of Offenders Act requires reasons for imprisoning an offender under twenty-one. The duty is no longer to punish but to justify.

Utility, verdict three: what the Constitution has repudiated

Graduation of punishment by caste is beyond recovery, and the point should be made with a case rather than as an assertion. Sukanya Shantha v. Union of India, 2024 INSC 753, decided on 3 October 2024 by Chandrachud CJI, Pardiwala and Manoj Misra JJ., examined State prison manuals that allotted sweeping and scavenging by caste, segregated barracks by caste and described denotified tribes as born criminals. The Court declared those provisions unconstitutional as violating Articles 14, 15, 17, 21 and 23, gave States three months to revise their manuals, directed the Union to amend the Model Prison Manual 2016 and the Model Prisons and Correctional Services Act 2023, and ordered the caste column and every reference to caste deleted from undertrial and convict registers. Whatever the classical texts said, a penal system that distributes labour or accommodation by varna is unconstitutional in India.

Conclusion. Both classical systems retain real utility, but not the parts most often cited. From the Hindu tradition the surviving idea is expiation, the proposition that a discharged sentence restores the offender, whose modern form is section 12 of the Probation of Offenders Act; from the Islamic tradition the surviving ideas are the victim's ownership of the claim, transformed into victim compensation and now built into section 293(a) of the Bharatiya Nagarik Suraksha Sanhita, and the tazir model of individualised discretionary sentencing, which is what Bachan Singh requires. What both systems contained and the Constitution has repudiated is graduation by status, which Sukanya Shantha struck down in 2024, and fixed bodily penalties, which could not survive the standard Sunil Batra applied to far milder measures.

The four objections

Before any theory is stated it is worth being clear what the theories are for. Deliberately inflicting suffering on a person needs justification, and four distinct objections have to be met. Why punish at all, when the harm is already done and adding to it increases the total suffering in the world? Why this much, when nothing obvious fixes the exchange rate between a wrong and a term of years? Why this person, when he is a product of circumstances he did not choose? And why in this way, when the same objective might be reached by treatment, compensation or supervision? Each theory below is strongest against one of these objections and weakest against another.

Retribution answers "why punish at all" and "why this much"

Retribution holds that punishment is deserved because a wrong has been done, is measured by the gravity of the wrong, and is owed whether or not it produces any further good. Kant's formulation is that a person is punished because he has committed a crime and never merely as a means to another end.

Deterrence answers "why this way", and fails on "why this person"

Deterrence treats the sentence as a price. Bentham's requirement is that the pain exceed the profit and that the sanction be certain and prompt. Indian law is full of the technique: minimum sentences in special statutes, enhanced punishment on a previous conviction under section 13 of the Bharatiya Nyaya Sanhita for offences under Chapter X and Chapter XVII, and security proceedings against habitual offenders under section 129 of the Bharatiya Nagarik Suraksha Sanhita, 2023.

Its empirical weakness is that certainty matters more than severity, a proposition the 262nd Report of the Law Commission of India, submitted on 31 August 2015, relied on in recommending abolition of the death penalty for all offences other than terrorism and waging war. Where detection is uncertain, raising severity converts punishment into a lottery.

Its moral weakness is decisive against the third objection: deterrence taken alone would justify punishing a person more than he deserves, or even punishing an innocent person, if the general effect were great enough, because on this theory the offender is a means to the deterrence of others. That is why it must be bounded by retribution's ceiling.

Reformation answers "why this person" differently, and struggles with "why this much"

Reformation treats the sentence as an occasion to change the offender so that he does not want to reoffend. Its instruments are classification, education, vocational training, work, remission, parole, open institutions, probation and aftercare, and its official Indian adoption dates from the Indian Jails Committee of 1919-20 under Sir Alexander Cardew, which was the first body to hold that a prison must have a reformative and not merely a deterrent effect.

Its statutory expressions are strong. Section 6 of the Probation of Offenders Act, 1958 provides that a person under twenty-one shall not be sentenced to imprisonment for an offence not punishable with life unless the court is satisfied that dealing with him under section 3 or section 4 would be undesirable, and it must record reasons after calling for a probation officer's report. Section 402 of the Sanhita requires special reasons wherever a court declines to use probation or the Juvenile Justice (Care and Protection of Children) Act, 2015. Community service became a punishment in its own right in section 4(f) of the Bharatiya Nyaya Sanhita, defined in the Explanation to section 23 of the Sanhita as work benefiting the community for which the convict receives no remuneration.

Its defect is that it asks nothing of the offender towards the victim, which is why the modern development is restorative. Section 5 of the Probation of Offenders Act allows compensation and costs to be ordered, and State of Gujarat v. Hon'ble High Court of Gujarat, (1998) 7 SCC 392 held that prisoners must be paid equitable wages fixed by the State and recommended that a portion be set apart as compensation for victims.

Why a civilised State must consider all of them

Conclusion. The theories of punishment are best understood as answers to four different objections, and no single theory answers all of them: retribution supplies the reason and the ceiling, deterrence and prevention supply the forward-looking purpose, reformation supplies the concern with the individual offender, and expiation alone addresses what happens after the sentence ends. A civilised State must consider all of them because each corrects a defect in the others, and Indian law makes that consideration a legal duty rather than a philosophical preference, in Bachan Singh for the gravest sentence, in section 402 of the Bharatiya Nagarik Suraksha Sanhita for the lightest, and in Aman Singh by requiring that the material for the weighing actually be gathered.

(a) Concept and definition of punishment

What is inflicted? A deprivation. Every item in the Indian catalogue is one: section 4 of the Bharatiya Nyaya Sanhita, 2023 lists death, imprisonment for life, imprisonment which is rigorous with hard labour or simple, forfeiture of property, fine, and community service, the last being new in 2023 and defined only in the Explanation to section 23 of the Bharatiya Nagarik Suraksha Sanhita, 2023 as work the court may order a convict to perform as a form of punishment that benefits the community, for which he is not entitled to any remuneration. Nothing on the list is a benefit, and the deprivation is not incidental but intended: it is the point.

By whom? By an authority constituted by the legal system against which the offence was committed, acting through a court. That element separates punishment from revenge, from lynching and from private discipline, and in India it is constitutionalised: Article 21 requires that a person be deprived of life or personal liberty only by a procedure established by law, which since Maneka Gandhi must be fair, just and reasonable.

Preventive detention is not punishment, because it does not respond to a proved offence; it is authorised separately by Article 22 and is measured by anticipated danger. The distinction is not merely formal: an order under section 129 of the Bharatiya Nagarik Suraksha Sanhita, requiring a habitual offender to show cause why he should not give security for good behaviour, is preventive, whereas the enhancement under section 13 of the Bharatiya Nyaya Sanhita for a person previously convicted under Chapter X or Chapter XVII is punitive, because it follows a conviction.

A civil penalty is not punishment in the constitutional sense, which is why a departmental penalty, a compounding fee and a prosecution may all follow the same conduct without offending Article 20(2).

And a measure a legislature calls something else will be judged by what it does. Mithu v. State of Punjab, (1983) 2 SCC 277 struck down section 303 of the Penal Code, which made death the only sentence for a murder by a person already serving life, because it removed the requirement of special reasons, the weighing of circumstances and the hearing on sentence from a whole class of accused. That is a proposition about the concept itself: in Indian constitutional law a punishment must be chosen, and a provision that leaves the court nothing to decide is void.

Two further limits belong in the definition rather than after it. A punishment must not be degrading, which is why Sunil Batra v. Delhi Administration, (1978) 4 SCC 494 read section 30(2) of the Prisons Act, 1894 as authorising safe custody and not solitary confinement and allowed bar fetters under section 56 only on recorded and reviewable reasons after hearing the prisoner. And it must not be discriminatory, which is why Sukanya Shantha v. Union of India, 2024 INSC 753 struck down prison manual provisions allotting labour by caste under Articles 14, 15, 17, 21 and 23 and ordered the caste column deleted from prison registers.

A rail passenger was found wearing a waistchain of 820 grams of pure gold coated with mercury to look like silver, seized under the Customs Act, 1962 and the Gold (Control) Act, 1968; both courts below acquitted, the High Court refusing the prosecution leave under section 391 of the Code to lead additional evidence proving the Mint Master's report because it was inexpedient to reopen the case six years on. The Supreme Court allowed the application and set aside the acquittal, holding that a murder may be committed in the heat of the moment when passions are aroused whereas an economic offence is committed with cool calculation and deliberate design for personal profit regardless of the consequence to the community, and that the community acting through the State is entitled to justice too.

The fine achieves nothing unless it is scaled to the gain. Section 8(1) of the Bharatiya Nyaya Sanhita provides that where no sum is expressed the amount of the fine is unlimited but shall not be excessive, which sounds generous and is in practice the opposite: courts anchor on figures written into statutes decades ago. A fine smaller than the profit is a licence fee.

What Parliament has done instead is remove discretion. Distrusting judicial leniency, it has prescribed minimum sentences and excluded the alternatives. Section 20AA of the Prevention of Food Adulteration Act, 1954, inserted in 1976, provides that neither the Probation of Offenders Act, 1958 nor section 360 of the Code applies to a person convicted under that Act unless he is under eighteen. Nagarajan v. State of Tamil Nadu, 2025 INSC 703, decided on 15 May 2025 by Dipankar Datta and Manmohan JJ., enforced it without qualification, holding probation categorically unavailable for offences under that Act committed between 1976 and its repeal in 2006, and that neither the reformative principle nor a constitutional argument could displace an express bar. A floor stops the sentence being too low; it does nothing about disparity above the floor.

And what is still missing is a scale. India has no Sentencing Council and no statutory sentencing guidelines, and the Malimath Committee's recommendation in 2003 that guidelines be framed has not been acted upon. Within one section a loss may range from a lakh to a thousand crore and nothing tells the judge how to price it.

(a) Collective fine

The actor who imposes: the State Government. The power is in section 10A of the Protection of Civil Rights Act, 1955, inserted by Act 106 of 1976 with effect from 19 November 1976, and it is extended to caste atrocities by section 16 of the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989, which applies section 10A so far as may be. Four conditions govern the exercise. There must be an inquiry in the prescribed manner. The Government must be satisfied of one of five matters: that the inhabitants of an area are concerned in an offence under the Act, or are abetting it, or are harbouring persons concerned in it, or are failing to render all the assistance in their power to discover or apprehend the offenders, or are suppressing material evidence. The imposition must be by notification in the Official Gazette. And the apportionment must be made according to the Government's judgment of the respective means of the inhabitants, a portion being assignable to a Hindu undivided family. The notification must then be proclaimed in the area by beat of drum or otherwise brought home to the inhabitants.

Why the device exists at all. State of Karnataka v. Appa Balu Ingale, AIR 1993 SC 1126, decided on 1 December 1992 by Kuldip Singh and K. Ramaswamy JJ., supplies the factual justification. Members of a Scheduled Caste were prevented, on the threat of a gun, from taking water from a newly dug borewell on the ground that they were untouchables; the Magistrate convicted under section 4 of the Protection of Civil Rights Act and awarded a month's simple imprisonment and a fine of Rs 100 each; a Single Judge of the Karnataka High Court acquitted in revision; the Supreme Court restored the conviction. Ramaswamy J. treated Article 17 as an absolute prohibition carrying no reasonable-restriction clause, and emphasised that the practice is sustained by a community rather than by an individual. Where the offence is committed under the shelter of a locality that will not give evidence, a sanction directed only at the hand and never at the shelter cannot work.

An older ancestor is worth a sentence: section 15A of the Police Act, 1861 allows an additional punitive police force to be quartered on a disturbed area at the inhabitants' cost, and the colonial state used it against whole villages. The 1955 Act took the technique, tied it to a defined constitutional evil, and added the hearing and the victim's exemption the Police Act never had.

Death in custody, which is the commonest. The duties designed to prevent it belong to the Medical Officer. Section 13 puts the prisoners' health in his charge under the Superintendent's orders. Section 14 requires him to report whenever the mind or body of any prisoner is likely to be injured by continued imprisonment or by any treatment or discipline, which is the statutory route by which a condition inside can be brought to authority. Section 15 requires him, on the death of any prisoner, to record the day of admission, discharge or death, the disease, the duration of the illness and its apparent cause. Section 29 requires him or a Medical Subordinate to visit at least once a day any prisoner kept in a cell in solitary confinement for more than twenty-four hours. Rule 46 of the Nelson Mandela Rules, adopted on 17 December 2015, reinforces the design by keeping health-care personnel out of the imposition of disciplinary sanctions altogether.

Torture and unlawful punishment. These are the Superintendent's province. Section 11 makes him responsible, subject to the Inspector General's orders, for discipline, labour, expenditure, the punishment of prison offences and the general administration, and section 12 requires him to keep the prescribed records. Sunil Batra v. Delhi Administration, (1978) 4 SCC 494 loaded that office with the safeguards that make the Act constitutional: confinement apart under section 30(2) means safe custody and not solitary confinement and is lawful only once a death sentence is finally executable, and bar fetters under section 56 may be imposed only for the shortest necessary period, on reasons recorded, subject to review, after hearing the prisoner, and never as a punishment. Sunil Batra (II), (1980) 3 SCC 488 supplied the remedy by treating a prisoner's letter alleging that a warder had tortured an inmate as a writ of habeas corpus and directing District Judges to visit prisons and hear grievances, so the staff owe a duty of access as well as of restraint.

Discrimination inside the prison. Sukanya Shantha v. Union of India, 2024 INSC 753, decided on 3 October 2024 by Chandrachud CJI, Pardiwala and Manoj Misra JJ., held that manual provisions allotting sweeping and scavenging by caste, segregating barracks and treating denotified tribes as born criminals were unconstitutional under Articles 14, 15, 17, 21 and 23; ordered the caste column and every reference to caste deleted from undertrial and convict registers; confined "habitual offender" in a manual to the definition in the State's own legislation; required States to revise their manuals within three months and the Union to amend the Model Prison Manual 2016 and the Model Prisons and Correctional Services Act 2023; and put the District Legal Services Authorities and the Board of Visitors under a duty of joint inspection reporting through the State authorities to NALSA. Custodial staff therefore now owe a duty to administer a manual purged of caste and to facilitate that inspection.

Conclusion. A collective fine under section 10A of the Protection of Civil Rights Act, 1955 is a rebuttable presumption against the inhabitants of an area, imposed by notification after inquiry, resistible by a free petition with a hearing, and inapplicable to victims, which is what keeps it inside Article 14 despite being an exception to personal liability. The duties of custodial staff are best read as the answers to the four failures of custody: the Medical Officer's reporting and death-recording duties against death, the Superintendent's recorded and reviewable orders against torture, his statutory application under section 479(3) against over-detention, and, since Sukanya Shantha, a duty to administer a manual free of caste against discrimination.

Its boundaries are best fixed against its neighbours. Criminology studies crime and the criminal, its causes, incidence and patterns. Criminal law defines offences, defences and liability. Victimology studies the person harmed and what is owed to him. Penology takes the conviction as given and studies the response. On one set of facts, criminology asks why the accused stole, criminal law asks whether the ingredients of theft are made out, victimology asks what the owner has lost, and penology asks whether he should be fined, imprisoned, ordered to do community service or released on probation, and what should be done with him if he is imprisoned.

Its scope has five parts: the justifications, meaning the theories of punishment, retributive, deterrent, preventive, reformative and expiatory, with restorative justice as the modern addition; the forms of punishment the law recognises; the institutions in which they are executed; the alternatives to institutional punishment; and the aftermath, meaning remission, parole, premature release, aftercare and the removal of disabilities.

The ideal penal system: the criteria

An ideal penal system is not one that pursues a single theory. It is one that satisfies a set of criteria simultaneously, and the criteria can be stated and then tested.

It must be certain rather than severe. Deterrence depends far more on the probability of detection and conviction than on the length of the sentence, a proposition the 262nd Report of the Law Commission of India, submitted on 31 August 2015, relied on in recommending abolition of the death penalty for all offences other than terrorism and waging war. A system that raises penalties while conviction remains uncertain converts punishment into a lottery.

It must be proportionate and capped by desert. Retribution supplies the only internal ceiling any theory offers, and without it deterrence and prevention would justify punishing a person more than he deserves. Indian law expresses proportionality through graduated maxima in section 4 of the Bharatiya Nyaya Sanhita and through the requirement of special reasons for a death sentence in section 393(3) of the Bharatiya Nagarik Suraksha Sanhita.

It must repair as well as condemn. State of Gujarat v. Hon'ble High Court of Gujarat, (1998) 7 SCC 392, decided on 24 September 1998, held that prisoners doing lawfully imposed hard labour must be paid equitable wages at rates fixed by the State, and recommended that States legislate to set apart a portion of those wages as compensation for the victims of the offence. Section 5 of the Probation of Offenders Act allows compensation and costs to be ordered as part of a probation order.

It must treat the prisoner as a person throughout. Sunil Batra v. Delhi Administration, (1978) 4 SCC 494 established that a prisoner retains every right not necessarily lost by confinement, and Sukanya Shantha v. Union of India, 2024 INSC 753, decided on 3 October 2024, held unconstitutional under Articles 14, 15, 17, 21 and 23 those prison manual provisions that allotted labour by caste, segregated barracks and treated denotified tribes as born criminals, ordering the caste column deleted from prison registers and confining "habitual offender" to the definition in a State's own legislation.

And it must restore the offender at the end. Section 12 of the Probation of Offenders Act, 1958 provides that a person dealt with under section 3 or section 4 shall not suffer any disqualification attaching to a conviction. A system in which a served sentence continues to punish through collateral consequences has not finished punishing at all.

Testing the Indian system against the criteria

Conclusion. Penology is the study of punishment and of the treatment of offenders, standing between criminology, which explains crime, and criminal law, which defines it, and its Indian content runs from section 4 of the Bharatiya Nyaya Sanhita through the Prisons Act, 1894 and the Probation of Offenders Act, 1958 to the remission provisions in sections 473 to 477 of the Sanhita. The ideal penal system is not the purely reformative one but the one that is certain rather than severe, proportionate and capped by desert, individualised, sparing of imprisonment, restorative to the victim, decent to the prisoner, and equipped with an automatic exit and a genuine restoration at the end. Measured against those criteria India performs best where the courts have intervened and worst where the failure is administrative, and its most serious defect is that the largest group in its prisons is not being punished at all.

Limb one: the place of probation in penal policy

Penal policy offers a court a ladder of responses, and probation sits at a precise rung on it. Read from the lightest, the ladder is: admonition, under section 3 of the Probation of Offenders Act, 1958, where the offender is reprimanded and discharged; fine and community service, both punishments in their own right under section 4(e) and the new section 4(f) of the Bharatiya Nyaya Sanhita, 2023; probation, under section 4 of the 1958 Act or section 401 of the Bharatiya Nagarik Suraksha Sanhita, 2023; and only then imprisonment, simple or rigorous, followed by life and death.

Probation therefore occupies the rung immediately below imprisonment, and its policy function is to make that last step avoidable. Three features fix its place.

It is a disposal, not an acquittal. The offender is convicted; what is withheld is the sentence. That is why probation can be offered for a wide range of offences: section 4 is available for any offence not punishable with death or imprisonment for life, which is very wide, while section 401 of the Sanhita is narrower for an adult, requiring the offence to carry fine only or imprisonment for seven years or less, though for a person under twenty-one or for any woman it extends to any offence not punishable with death or life.

Where penal policy has decided that probation is inappropriate it has said so expressly, and those exclusions define its place from the other side. Section 19 of the Scheduled Castes and the Scheduled Tribes (Prevention of Atrocities) Act, 1989 excludes both section 360 of the Code and the Probation Act; section 16A of the Protection of Civil Rights Act, 1955 excludes the Act for anyone above fourteen; and section 20AA of the Prevention of Food Adulteration Act, 1954, inserted in 1976, excludes both unless the offender is under eighteen. Nagarajan v. State of Tamil Nadu, 2025 INSC 703, decided on 15 May 2025 by Dipankar Datta and Manmohan JJ., held that bar unambiguous and categorical, so that probation was unavailable for offences under that Act committed between 1976 and its repeal in 2006, and that neither the reformative principle nor a constitutional argument could displace it.

The point of the whole scheme is section 12: a person dealt with under section 3 or section 4 shall not suffer any disqualification attaching to a conviction unless he is later sentenced for the original offence. Probation's place in penal policy is to preserve an offender's civil standing where imprisonment would destroy it.

Rattan Lal v. State of Punjab, AIR 1965 SC 444, decided on 10 April 1964, shows how far that policy is carried. A boy of sixteen was convicted on 31 May 1962 and sentenced to rigorous imprisonment and a fine; the Act was extended to Gurgaon only on 1 September 1962, so the trial magistrate had no power under it. The Supreme Court held that a later law reducing punishment is a beneficial ex post facto law which may be applied to an earlier conviction, and that the appellate court could make the order the trial court could not.

Limb two: the major functions of the probation officer

Section 13 says who he may be: a person appointed or recognised as a probation officer by the State Government, a person provided by a society recognised by the State Government, or, in an exceptional case, any other person whom the court considers fit in the special circumstances. He works under the control of the District Magistrate of the district in which the offender for the time being resides, and by section 15 he is a public servant.

Section 14 sets out five duties.

Four, aftercare. He must advise and assist, in such cases and in such manner as may be prescribed, persons who have been released under section 4, which extends his role beyond the currency of supervision.

Five, such other duties as may be prescribed by rules made by the State Government under section 17. To these the Act adds a reporting function under section 9, since it is on the report of a probation officer that a court may act on a suspected breach of a bond.

The weakness is administrative. Where a State has too few officers the report cannot be obtained, sections 4(2) and 6(2) become formalities, and probation is refused for want of machinery rather than on the merits. That is why section 402 of the Sanhita matters: it puts the court under a duty to explain itself when it does not use the Act.

Limb three: how probation differs from parole

The two are frequently confused and they belong to different stages of the criminal process.

Probation happens instead of the sentence. It is ordered by a court at the moment of conviction, under the Probation of Offenders Act, 1958 or section 401 of the Sanhita, and if the bond is kept no sentence is served at all.

For full marks, cover: the same quotation is set on the second paper of the 2018 scan, where the answer opens on the Supreme Court's own admission of error and then sets out a balance sheet. Take a different route here. A death sentence can be reduced at four separate stages, and the quotation is really an observation about what happens at each; work through them, show why the outcome varies at each stage and what has been done about it, and derive the merits and demerits from that analysis rather than listing them separately. That produces an answer that uses the quotation instead of merely quoting it.

Stage one: the trial court's choice of sentence

The first opportunity to avoid a death sentence is not to impose one. Section 393(3) of the Bharatiya Nagarik Suraksha Sanhita, 2023, the successor of section 354(3) of the Code of 1973, requires the judgment to state the special reasons for a death sentence, and section 258(2) requires the convicting judge to hear the accused on the question of sentence before passing it.

Stage two: confirmation and appeal, where the doctrine itself produces variation

Machhi Singh v. State of Punjab, (1983) 3 SCC 470 tried to give the rarest of rare test content by grouping the cases under five heads, the manner of commission, the motive, the anti-social or socially abhorrent nature of the crime, its magnitude and the personality of the victim, and asked whether the collective conscience of the community would expect the death sentence. That last criterion is the mechanism by which the quotation becomes true. It cannot be proved by evidence, cannot be tested in cross-examination and cannot be reviewed on appeal, so two benches applying it faithfully may legitimately differ.

Bhagwati J. predicted exactly this. Dissenting in Bachan Singh and delivering his reasons separately on 16 August 1982, reported as (1982) 3 SCC 24, he held that no legislative principle distinguishes the case that attracts death from the case that does not, so the outcome turns on the composition of the bench, and that such arbitrariness offends Articles 14 and 21 in a punishment that cannot be recalled.

Stage four: commutation by the executive, and by the court on supervening grounds

The executive route. Articles 72 and 161 confer the power to pardon, reprieve, respite, remit, suspend and commute, and section 474 of the Bharatiya Nagarik Suraksha Sanhita confers the statutory power, beginning with the commutation of a sentence of death to imprisonment for life; section 476 gives the Central Government a concurrent power in death cases. Maru Ram v. Union of India, (1981) 1 SCC 107, a Constitution Bench, held that the constitutional powers, though wide, are neither absolute nor unfettered and are exercised on the aid and advice of the Council of Ministers, and it upheld section 433A of the Code of 1973, the fourteen-year floor now found in section 475 of the Sanhita. Kehar Singh v. Union of India, (1989) 1 SCC 204 held that the President may examine the merits of the case and is not confined to the record as the courts found it, but is not obliged to give reasons and need not grant an oral hearing. Epuru Sudhakar v. Government of Andhra Pradesh, (2006) 8 SCC 161 established that the exercise is subject to judicial review where there was no application of mind, bad faith, wholly irrelevant considerations or arbitrariness.

The merits and demerits, read off that analysis

Merits. Retributive proportionality is the strongest, because it depends on no empirical claim: for the gravest offences any lesser sentence may be thought to undervalue the victim's life. Parliament has accepted it, retaining the punishment in section 4(a) of the Bharatiya Nyaya Sanhita, 2023 and making it available for rape of a woman under twelve under section 65(2), for rape causing death or a persistent vegetative state under section 66, for gang rape of a woman under eighteen under section 70(2), and for a repeat offender under section 71. Deterrence is asserted but not established: the 262nd Report of the Law Commission of India, submitted on 31 August 2015, recommended abolition for all offences other than terrorism and waging war and said there was no penological justification for the terrorism exception either, accepting it only because legislators feared for national security. Incapacitation was once decisive and is no longer, because Swamy Shraddananda (2) v. State of Karnataka, (2008) 13 SCC 767 and Union of India v. V. Sriharan, (2016) 7 SCC 1 created and upheld a whole-life category, reaffirmed in Ramasrey @ Fakkad v. State of Uttar Pradesh, 2026 INSC 764 on 30 July 2026.

Conclusion. The quotation is an accurate description of Indian practice and not a rhetorical flourish, and the reason is structural rather than personal: at the trial stage judges differ because the mitigating material available to them differs, at the appellate stage because Machhi Singh invites an unprovable inquiry into the community's conscience, and at the commutation stage because delay and executive discretion enter. Bariyar is the Court's own proof, having held seven of its judgments per incuriam. Weighing the merits against the demerits, the strongest case for the punishment is proportionate condemnation and the strongest case against it is that it is irreversible in a system that has admitted error, and the most important development is that Indian law now offers a sentence which is final without being irrevocable.

The theories, in brief

Retribution holds that punishment is deserved because a wrong has been done, is measured by the gravity of the wrong, and is owed whether or not it produces further good. Its indispensable contribution is the ceiling: nobody may be punished beyond desert. Indian law expresses it through the graduated punishments in section 4 of the Bharatiya Nyaya Sanhita, 2023 and through the requirement of special reasons for a death sentence in section 393(3) of the Bharatiya Nagarik Suraksha Sanhita, 2023.

Deterrence treats the sentence as a price to be set above the profit of the offence, and works on the offender himself and on everybody else. Its Indian devices include minimum sentences, enhanced punishment on a previous conviction under section 13 of the Bharatiya Nyaya Sanhita, and security proceedings against habitual offenders under section 129 of the Sanhita.

Prevention disables rather than persuades, through imprisonment, forfeiture of property under section 4(d), disqualification, and address notification under section 394 of the Sanhita for up to five years after a second qualifying conviction.

The deterrent theory and heinous crimes against women: what the law now provides

The legislative record since 2012 is a sustained experiment in deterrence, and it should be stated precisely.

The Criminal Law (Amendment) Act, 2013, following the Delhi gang rape of December 2012, redefined the offence, created new offences of acid attack, stalking and voyeurism, and introduced death or life imprisonment where rape causes the death of the victim or leaves her in a persistent vegetative state, and for repeat offenders. The Criminal Law (Amendment) Act, 2018, following the Kathua case, added the death penalty for the rape of a girl below twelve and raised the minimum sentences across the board. The Protection of Children from Sexual Offences (Amendment) Act, 2019 extended the death penalty to aggravated penetrative sexual assault on a child.

Sentencing procedure was strengthened too: Mukesh v. State (NCT of Delhi), (2017) 6 SCC 1 confirmed the death sentences in the Delhi gang rape case, and the convicts were executed on 20 March 2020, seven years and three months after the offence, which is itself a fact about how deterrence operates in practice.

What the expert committee actually advised

This is the part of the answer that separates a good script from an average one. The Committee on Amendments to Criminal Law under Justice J.S. Verma, appointed after the December 2012 case, reported on 23 January 2013. It recommended a substantial increase in sentences, from seven years to ten, twenty and life, but expressly opposed the death penalty for rape, taking the view that it does not act as a deterrent and that a graduated scheme of enhanced imprisonment was the appropriate response. The Government did not accept that recommendation, and Parliament has since extended the death penalty for sexual offences on four separate occasions.

A very high minimum can suppress reporting and conviction. Where the offender is a family member or a neighbour, a mandatory sentence of twenty years may make a family unwilling to report at all, and may make a court reluctant to convict where the evidence is imperfect. That is the standard objection to mandatory minimums and it applies with particular force here.

Where the sentence for rape approaches the sentence for murder, the marginal deterrent against killing the victim falls. This is the gravest structural objection to the 2018 model, and it is why the Verma Committee preferred graduated imprisonment: section 66 already provides death for rape causing death, so the offender who has committed an offence under section 65(2) faces a sentence that can rise no further.

Conclusion. The five theories divide the labour of sentencing: retribution supplies the ceiling, deterrence the price, prevention the disablement, reformation the concern with the offender, and expiation with restoration the position after the sentence ends. On heinous crimes against women the deterrent theory has driven every legislative change since 2013, and the Bharatiya Nyaya Sanhita now provides for the death penalty in four separate provisions, sections 65(2), 66, 70(2) and 71. But the expert body Parliament itself appointed advised against exactly that in January 2013, and the Law Commission repeated the reasoning in 2015: deterrence turns on certainty rather than on severity. The relevance of the deterrent theory in this field is therefore real but misdirected, and the honest position is that the sentences are now as high as they can usefully go and that the remaining work lies in detection, investigation and trial.

Limb one: prison reforms in India, as they stand

The statute has not been reformed at all. Indian prisons are still administered under the Prisons Act, 1894, supplemented by the Prisoners Act, 1900 and the Transfer of Prisoners Act, 1950 and by State prison manuals. The Act has chapters on maintenance and officers, duties of officers, admission and discharge, discipline, food and clothing, employment, health, visits, prison offences and punishments, and none on reformation, education, classification for treatment or release. The Model Prisons and Correctional Services Act, 2023, circulated by the Ministry of Home Affairs to all States and Union Territories on 10 May 2023 to replace the three colonial statutes, is a model only: prisons are Entry 4 of the State List, so it binds nobody until a State legislature enacts it, and the Centre has told Parliament that no State has confirmed its adoption, though 21 States and all Union Territories have adopted the Model Prison Manual 2016.

What has been reformed is what the courts have reached. Sunil Batra v. Delhi Administration, (1978) 4 SCC 494 read sections 30(2) and 56 of the 1894 Act down so that confinement apart means safe custody and not solitary confinement, lawful only once a death sentence is finally executable, and bar fetters may be imposed only for the shortest necessary period on recorded and reviewable reasons after hearing the prisoner. Sunil Batra (II), (1980) 3 SCC 488 made habeas corpus available against conditions and put District Judges into prisons as visitors. State of Gujarat v. Hon'ble High Court of Gujarat, (1998) 7 SCC 392, decided on 24 September 1998, held that prisoners doing lawfully imposed hard labour must be paid equitable wages fixed by the State, though the Minimum Wages Act does not apply of its own force, and recommended that States legislate to set apart a portion as compensation for victims. State of Andhra Pradesh v. Challa Ramkrishna Reddy, (2000) 5 SCC 712 rejected sovereign immunity where a remand prisoner was killed after warnings of danger were ignored, holding a prisoner's right to life enforceable in damages.

Two decisions have attacked overcrowding from the exit. In Re: Policy Strategy for Grant of Bail, 2025 INSC 239, decided on 18 February 2025 by Oka and Bhuyan JJ., held that where a State has a remission policy the Government must consider every eligible convict for premature release without any application, directed States and Union Territories without a policy to frame one within two months, required conditions on remission to be capable of performance and not vague or oppressive, required orders granting or refusing remission to carry brief reasons and to be communicated through the prison with a copy to the District Legal Services Authority, and held that a remission once granted cannot be cancelled without a hearing and reasons.

What has not been reformed is the establishment. The Mulla Committee recommended a National Prison Commission and an all-India Indian Prisons and Correctional Service in 1983 and neither exists; the Krishna Iyer Committee recommended separate custodial facilities for women in every State in 1987 and 21 States and Union Territories still have none; and the Cardew Committee recommended a fixed maximum intake for every jail in 1920 and occupancy stands at 112.7 per cent.

Limb two: a note on the history of prison reforms in India

Before 1836. Confinement was a way of holding a person until punishment rather than a punishment in itself. The classical Indian sanctions were fine, corporal punishment, banishment, confiscation and death.

1836 to 1894: the committees of discipline. The Prison Discipline Committee of 1836, appointed on Lord Macaulay's minute and reporting in 1838, was concerned with corruption, indiscipline and mortality; it recommended central prisons and rigorous discipline and expressly rejected reformation as an object. An Inspector General of Prisons was appointed in 1844. Further inquiries in 1864, at a conference of experts in 1877 and by a Fourth Jail Commission in 1888 produced the Prisons Act, 1894.

1980 to 1987: the two great committees. The Mulla Committee, constituted in 1980 under Justice A.N. Mulla and reporting in 1983, recommended a National Prison Commission, an all-India correctional service, a ban on housing juveniles with hardened criminals, and separate prisons for undertrials, women and young offenders. The National Expert Committee on Women Prisoners, constituted in 1986 under Justice V.R. Krishna Iyer and reporting in May 1987, recommended a National Policy on Custodial Justice to Women, separate custodial facilities in every State for convicted and undertrial women, a statutory body to oversee implementation, and much greater reliance on non-custodial options.

1978 to 1997: the judicial phase. Sunil Batra and Sunil Batra (II), Hussainara Khatoon v. Home Secretary, State of Bihar, (1980) 1 SCC 81 on speedy trial and the release of over-detained undertrials, Sheela Barse v. State of Maharashtra, (1983) 2 SCC 96 on women in custody, and Ramamurthy v. State of Karnataka, (1997) 2 SCC 642, which identified nine chronic problems and directed the Union to prepare a model prison manual, from which the manuals of 2003 and 2016 descend.

For full marks, cover: the paper asks in terms for the Code of Criminal Procedure sections, so the answer must work through Chapter XXIA, sections 265A to 265L, section by section. It must also say plainly that those sections no longer apply: the Code of 1973 was repealed with effect from 1 July 2024 and the chapter is now Chapter XXIII, sections 289 to 300, of the Bharatiya Nagarik Suraksha Sanhita, 2023, which changed four things of substance. Giving both numbering systems in a mapping table, and identifying what actually changed, is what separates a current answer from one written out of an old textbook. The same subject is set at Q7 and Q9(a) of the 2018 papers on this site, comparatively and as a short note respectively; this answer is organised round the sections.

What plea bargaining is, and why India resisted it

Plea bargaining is the disposal of a criminal case by agreement, in which the accused surrenders his right to a trial in exchange for a reduced consequence. Comparative writing distinguishes charge bargaining, sentence bargaining and fact bargaining; Indian law permits only a controlled form of the second.

Thippaswamy v. State of Karnataka, (1983) 1 SCC 194, decided on 25 November 1982, shows the constitutional objection at its sharpest and is the case most worth citing on the danger the statutory scheme had to guard against. The appellant pleaded guilty to a charge under section 304A of the Penal Code and the Magistrate imposed only a fine of Rs 1,000; on the State's appeal the High Court maintained the fine and added one year's rigorous imprisonment. The Supreme Court held that it would be clearly violative of Article 21 to induce or lead an accused to plead guilty on a promise or assurance that he would be let off lightly and then, in appeal or revision, to enhance the sentence; in such a case the appellate or revisional court should set aside the conviction and sentence and remand the case so that the accused may defend himself if he wishes. It allowed the appeal and remanded the case to the Judicial Magistrate First Class. The decision explains why the Indian statutory scheme, when it finally came, made the resulting judgment final and kept the negotiation under the court's own eye.

Section 265B, application for plea bargaining. The accused filed an application in the court in which the offence was pending trial, containing a brief description of the case and accompanied by an affidavit that he had preferred the application voluntarily, after understanding the nature and extent of the punishment provided by law, and that he had not previously been convicted by a court in a case in which he had been charged with the same offence. The court then issued notice to the Public Prosecutor or the complainant and to the accused, and examined the accused in camera, the other party not being present, to satisfy itself that the application was voluntary. If satisfied, it gave time to work out a mutually satisfactory disposition; if not satisfied, or if the accused had previously been convicted of the same offence, it proceeded with the case from the stage the application had been filed.

Section 265C, guidelines for a mutually satisfactory disposition. In a police report case the court issued notice to the Public Prosecutor, the investigating police officer, the accused and the victim to participate in a meeting; in a complaint case, to the accused and the victim. It was the court's duty to ensure throughout that the process was completed voluntarily, and the accused or the victim might participate with an advocate.

Section 265F required the judgment to be delivered in open court and signed by the presiding officer. Section 265G made the judgment final, with no appeal, save a special leave petition under Article 136 or a writ petition under Articles 226 and 227. Section 265H gave the court all its ordinary powers as to bail, trial and disposal for the purposes of the chapter. Section 265I applied section 428 so that the period of detention already undergone was set off against the sentence. Section 265J gave the chapter overriding effect. Section 265K provided that the statements or facts stated by the accused in his application shall not be used for any other purpose. Section 265L excluded the chapter altogether in the case of a juvenile or child.

What the Sanhita changed, and the mapping

The Code of 1973 stood repealed with effect from 1 July 2024, and the chapter is now Chapter XXIII of the Bharatiya Nagarik Suraksha Sanhita, 2023.

The age qualification has gone. Section 265A excluded offences committed against a child below fourteen; section 289 excludes offences committed against a woman, or a child, without qualification, so every offence against any child is now outside the chapter.

A time limit has been introduced. Section 290(1) requires the application to be filed within thirty days from the date of framing of charge. The Code prescribed no limit at all, which meant an application could be made at any stage and used to delay a trial.

A time limit has been placed on the negotiation too. Section 290(4)(a) allows the court, once satisfied that the application is voluntary, to give the parties a period not exceeding sixty days to work out a mutually satisfactory disposition.

First offenders receive a further concession. Section 293(c) provides that where a minimum punishment is prescribed the court may sentence the accused to half of it, and to one-fourth where he is a first-time offender never previously convicted; section 293(d) provides that in any other case the sentence may be one-fourth of the punishment provided or extendable, and one-sixth where he is a first-time offender. The one-fourth-of-minimum and the one-sixth were not in section 265E.

Its weakness is take-up. A mechanism designed to reduce arrears has been little used, so most accused persons and many practitioners have no working experience of it, and the machinery of notices, meetings and reports adds work that a busy magistrate's court has little reason to prefer to a summary disposal.

Conclusion. Plea bargaining entered Indian law by the Criminal Law (Amendment) Act, 2005 as Chapter XXIA of the Code of Criminal Procedure, 1973, sections 265A to 265L, in force from 5 July 2006, after the Supreme Court had condemned the informal practice in Murlidhar Meghraj Loya and Chandrika and after the Law Commission and the Malimath Committee had recommended a statutory form. Those sections are now sections 289 to 300 of the Bharatiya Nagarik Suraksha Sanhita, 2023, and the four substantive changes are the removal of the age qualification for offences against a child, the thirty-day limit from the framing of charge, the sixty-day limit on the negotiation, and the additional first-offender concessions of one-fourth of the minimum and one-sixth of the punishment provided. Throughout, the Indian scheme differs from the American one in that the bargain is struck under the court's own supervision and the sentence is a statutory fraction rather than a negotiated figure.