Three times in four decades Parliament has rebuilt the criminal law on offences against women, and each time it has done so in the shadow of a single case that concentrated public attention. The Criminal Law (Amendment) Act of 1983 worked on evidence and on the conditions of trial. The Criminal Law (Amendment) Act of 2013 rebuilt the offence definitions and added a shelf of new crimes. The Criminal Law (Amendment) Act of 2018 worked almost entirely on the sentencing tariff and on deadlines. The question a reader comparing the criminal law amendments on women actually has to answer is not which wave was largest, or loudest, or most severe. It is which wave changed what happens in practice: what a prosecution must prove, what a complainant must endure, what a police station must do, and what a court may impose. Those are different questions, and they have different answers, because the three waves did not do the same kind of work. One changed the law of proof. One changed the catalogue of offences. One changed the numbers attached to punishment. This comparison applies a single test to all three: did the wave alter the inside of a trial or a police investigation, or did it alter only the text that is read out after guilt is established? The verdict at the end follows from that test, and the table that carries the comparison is built so that the verdict can be checked against it row by row.

The comparison sits inside a wider field. The statutes compared here are the amendments to the general criminal codes, the Indian Penal Code, the Code of Criminal Procedure and the Indian Evidence Act, and their successors of 2023. The surrounding body of protective legislation, the dowry prohibition statute, the domestic violence statute, the workplace harassment statute and the rest, is mapped in the guide to women’s rights legislation in India, and this article links down to the specialist treatments where a provision belongs to another article’s territory. Its own territory is the comparison itself: three amendment waves, measured on the same dimensions, with a defended answer to the question of which one made the greatest practical difference.

The pattern every wave repeats

Each wave began with a case, and the sequence of case, outrage and amendment is so regular in this field that it deserves to be treated as a structural feature of the law rather than as background colour. In each instance, described here in the barest terms because the legislative response is the subject, a prosecution or an incident exposed a gap between what the law punished and what the public understood the law to be for. A Supreme Court acquittal in 1978, in a case of alleged rape inside a police station, produced an open letter to the Chief Justice of India from four law teachers, Upendra Baxi, Lotika Sarkar, Raghunath Kelkar and Vasudha Dhagamwar, and a Law Commission examination that reported in 1980; the amendments followed in 1983. A crime in Delhi in December 2012 produced a committee constituted on 23 December 2012 which reported on 23 January 2013, an ordinance on 3 February 2013 and an Act assented to on 2 April 2013. Cases involving child victims in the early months of 2018 produced an ordinance on 21 April 2018 and an Act assented to on 11 August 2018. The whole cycle, from triggering case to enacted amendment, has never in this field taken longer than a few years, and in the second and third waves it took weeks and months respectively.

Speed of that order cuts in both directions, and an honest comparison has to price both. On one side, the pressure is the only reason reform happened at all. The 1980 Law Commission report on rape and allied offences made careful recommendations that waited for a political moment; the Verma Committee’s report would have joined the shelf of unread committee reports had the executive not been forced to legislate from it within days. Crisis drafting also borrows from whatever careful work already exists, which is why the 2013 Act is substantially the Verma report and the 1983 Acts are substantially the Law Commission’s 84th Report: the thinking had been done in advance, by bodies working without a deadline, and the crisis supplied only the will. On the other side, provisions drafted in weeks arrive with their edges unfinished. Courts then spend years working out what a hastily raised age threshold does to adolescent relationships, or whether a mandatory minimum leaves any room for the case that does not fit the category, and the working out is done in individual prosecutions, at individual defendants’ and complainants’ expense. The pattern is not a scandal to be exposed but a mechanism to be understood: in this field, Parliament legislates retrospectively, at the last case, and the quality of each wave depends on how much slow work the fast moment had available to borrow.

The pattern also operates below the Union level, and the 2018 wave cannot be understood without its state level prelude. Criminal law sits in the Concurrent List, and states may amend the Penal Code for their territories subject to presidential assent where the amendment repugnant to the central law is reserved for it. In the months before the 2018 ordinance, the legislatures of Madhya Pradesh and Rajasthan enacted amendments prescribing the death penalty for the rape of girls below twelve within their states, and those amendments travelled to the President for assent while the Union was still framing its own response. The Union’s ordinance and Act then generalised nationally what two states had already legislated locally, and the sequence repeated the field’s logic at two levels of the federation at once: a case concentrates attention, the legislature nearest the pressure legislates severity, and the wider legislature follows to make the position uniform. State amendments of this kind create a complication the comparison should note, since a sentencing map drawn by state amendment and then overwritten by central amendment within the same year is a map practitioners must reconstruct date by date, and the reconstruction is part of the hidden cost of legislating in layers under pressure.

One further feature of the pattern matters for the verdict. Each wave was debated almost entirely in the vocabulary of punishment. The public demand that follows a case is a demand for severity, because severity is the part of the criminal process a headline can carry. Yet the provisions that decide whether a prosecution succeeds are mostly not punishment provisions at all. They are the definition of the offence, the rules about what must be proved and by whom, the conditions under which a complainant gives evidence, and the duties laid on the police officer who first hears the complaint. A comparison that followed the headlines would rank the waves by their sentencing maxima and stop there. A comparison that follows the trial ranks them differently, and the difference between the two rankings is the argument of this article.

The 1983 wave: proof, custody and the closed courtroom

The 1983 wave consists of two Acts, and the fact that there are two is the first thing most accounts blur. The Criminal Law (Amendment) Act, 1983, Act 43 of 1983, dealt with rape: its definition at the margins, its punishment structure, the position of women in the custody of the state, the presumption on consent, the secrecy of the complainant’s identity and the closure of the courtroom. The Criminal Law (Second Amendment) Act, 1983, Act 46 of 1983, dealt with the matrimonial home: it inserted the cruelty offence, Section 498A of the Penal Code, and the corresponding presumption in the Evidence Act regarding abetment of a married woman’s suicide. Both were brought into force on 25 December 1983. Between them they contain the whole of the 1983 wave’s claim to have changed practice, and it is a claim about proof and procedure far more than about punishment.

The rape half of the wave grew out of the Law Commission’s 84th Report of 1980 and a reform bill introduced in 1980 which had travelled slowly. The gestation is worth pausing on, because it distinguishes this wave from both its successors. Between the acquittal that triggered the reform and the commencement of the two Acts on 25 December 1983, five years passed, containing a Law Commission examination, a bill, and the ordinary opportunities for the proposal to be diluted, delayed or dropped that the parliamentary process provides. The wave survived the interval with its evidentiary core intact, which the comparison reads as evidence that its contents had been argued into a durable consensus before enactment: the presumption, the closed courtroom and the identity protection were not improvised under the pressure of a news cycle but carried, substantially unchanged, from a report written three years before the vote. No later wave had that protection, because no later wave was given the time, and the difference shows in the texts, in the 2013 Act’s unexamined age line and in the 2018 Act’s absence of any definitional layer at all. Slowness is not a virtue in itself, and the interval had its own cost in cases tried under the unreformed law. But the correlation in this set between gestation and durability is complete, and it is part of the explanation for the verdict the comparison reaches. What the enacted Act did to the definition of rape was modest: it inserted a new clause into Section 375 covering consent obtained when the woman, by reason of unsoundness of mind or intoxication or the administration of a stupefying substance, is unable to understand the nature and consequences of what she consents to. The definition otherwise stayed narrow, confined to penetration in its traditional sense, and the Supreme Court’s later refusal in Sakshi v. Union of India in 2004 to widen it by interpretation, directing attention back to the legislature and to the Law Commission’s 172nd Report instead, shows how completely the definitional question was left for another generation. The 1983 Act’s real work was done elsewhere.

Its first piece of work was the punishment structure of Section 376, which it rebuilt around mandatory minima for the first time in this field. Ordinary rape attracted a minimum of seven years. The aggravated categories collected in Section 376(2), which included rape by a police officer within his station’s limits, rape by a public servant taking advantage of his position, rape by the management or staff of a jail, remand home or hospital on an inmate or patient, rape of a pregnant woman, rape of a girl below twelve, and gang rape, attracted a minimum of ten years. Both minima came with a proviso under which the court could, for adequate and special reasons recorded in the judgment, impose less. Around that core the Act built the custodial offences. Sections 376B, 376C and 376D punished sexual intercourse not amounting to rape by, respectively, a public servant with a woman in his custody, the superintendent or manager of a jail or other place of custody with a female inmate, and the management or staff of a hospital with a woman in that hospital, each carrying up to five years. Section 376A punished a husband’s intercourse with his wife during separation under a decree or custom, without her consent, with up to two years. The drafting drew a deliberate line: intercourse by a custodian that met the definition of rape was aggravated rape under Section 376(2), and intercourse procured by the abuse of a custodial position that fell short of that definition was captured by the lesser offences, so that the position of power was itself penalised rather than treated merely as a circumstance.

What separates the custody offences of 1983 from aggravated rape?

The aggravated categories in Section 376(2) punished rape where the rapist held power over the woman and set a ten year minimum. Sections 376B to 376D punished intercourse not amounting to rape but procured by a custodian’s position, up to five years, so abuse of the position was itself an offence even where rape could not be proved.

The second piece of work was evidentiary, and it is the provision on which this article’s verdict turns. The Act inserted Section 114A into the Evidence Act: in a prosecution for the aggravated categories of rape, where sexual intercourse by the accused is proved and the question is whether it was without the woman’s consent, the court shall presume that she did not consent if she states in her evidence that she did not. The presumption is mandatory in form and rebuttable in substance, and it reverses the practical geometry of a rape trial. Before 1983, absence of consent was an element the prosecution had to establish like any other, in cases that typically turned on the uncorroborated word of the complainant against the accused, assessed against judicial habits of suspicion that the 1978 acquittal had displayed at their starkest. After 1983, in the categories covered, once intercourse was proved and the woman said she had not consented, the evidential burden on that issue moved to the accused. The provision did not decide cases by itself, and courts continued to examine the complainant’s testimony for credibility in the ordinary way. What it changed was the starting point of the analysis, and starting points in trials are not technicalities; they are where most cases are actually won and lost.

The presumption also landed in a body of case law that the courts were rebuilding at the same time, and the two movements reinforced each other. In State of Maharashtra v. Chandraprakash Kewalchand Jain, decided in 1990, the Supreme Court held that a complainant in a sexual offence case is not an accomplice to the crime, that her testimony carries the weight of the testimony of an injured witness, and that a court requires no corroboration of it as a matter of law before acting on it, though a court may look for assurance where the testimony itself gives reason for caution. The holding attacked the same judicial habit the presumption attacked, the standing demand that a woman’s account of a sexual offence be verified against something outside itself before it could count. Read together, the statute of 1983 and the judgment of 1990 mark the moment the centre of the rape trial moved: from an inquiry that began by discounting the complainant to an inquiry that begins with her account, tests it, and in the aggravated categories presumes the element of non consent once the act is proved and she denies consenting. Neither the statute nor the judgment ended acquittals, and neither was designed to. What they ended was a particular distribution of the benefit of the doubt, and distributions of that kind are among the most consequential things a legal system can change, because they operate in every case including the ones no report ever mentions.

The scope of the presumption deserves one precise note, because summaries often widen it past its text. Section 114A applied, and its successor applies, to the aggravated categories of rape enumerated in it, not to every rape prosecution; in an ordinary prosecution under the principal provision, absence of consent remained an element for the prosecution to prove in the usual way, assisted by the corroboration jurisprudence but not by the presumption. The 2013 and 2018 waves each extended the presumption’s list as they extended the offence structure, and the recodification carried the list into Section 120 of the Bharatiya Sakshya Adhiniyam. The pattern is characteristic of the evidentiary wave’s afterlife: the device was built once, in 1983, and every later legislature that rebuilt the offences reached for the same device and lengthened its schedule rather than inventing a new one. A reform that later reformers treat as standard equipment has demonstrated its practical value in the only laboratory available, the drafting room of the next crisis.

The third piece of work was about the conditions under which a complainant entered the process at all. The Act inserted Section 228A into the Penal Code, making it an offence, punishable with up to two years, to print or publish the name or any matter that may make known the identity of a victim of specified sexual offences. The prohibition’s working life runs well beyond the courtroom, into every newsroom and research desk that handles such a case: reporting proceeds under a statutory bar on identification, and the material a journalist accumulates in covering a trial, interviews, documents, medical and personal records, carries a sensitivity that ordinary filing does not, which is why VaultBook’s account of how journalists use an encrypted workspace to protect their sources describes a discipline directly translatable to this field, a single controlled file in which material that could identify or endanger a person is kept rather than scattered across shared drives and open inboxes. It amended the Code of Criminal Procedure so that the inquiry and trial of rape and the custodial offences were to be conducted in camera, and so that the presiding officer could control access to the courtroom. Together with the presumption, these provisions addressed the three reasons a complaint was, in the reformers’ diagnosis, least likely to be made and most likely to fail: that the process publicised the complainant’s name to her community, that she would have to narrate the events in an open court, and that her denial of consent would be weighed from a standing start of disbelief. The point deserves emphasis because it is easy to misread 1983 as a sentencing amendment with procedural trimmings. The sentencing minima were the part the headlines could carry. The presumption, the closed courtroom and the protected identity were the parts that changed the experience and the prospects of the person the system depended on.

Can a court impose less than the minimum sentence in a rape case?

Under the 1983 settlement it could, where adequate and special reasons were recorded in the judgment, and the Supreme Court spent two decades insisting those reasons be genuinely special and not routine mitigation. The 2013 Act deleted the escape for the principal offence, and the 2018 Act added categories with no judicial room below the floor at all.

The jurisprudence of the proviso is part of the 1983 wave’s long tail. Sentencing courts used the adequate and special reasons route often enough that the Supreme Court returned to it repeatedly, holding in State of Karnataka v. Krishnappa in 2000, and again in later cases, that reasons such as the accused’s family circumstances or the passage of time were not the kind of reasons the proviso contemplated, and that a sentence below the floor in a rape case required justification of a different order. The exchange between a legislature that set minima and a judiciary that kept a controlled escape route, and then supervised the escape route’s use, is the template against which the later waves’ abolition of discretion has to be assessed. The 2013 Act removed the proviso from Section 376 altogether, so that the seven-year floor became absolute for the principal offence. The 2018 Act went further in kind rather than degree: it created categories carrying minima of twenty years and maxima expressed as imprisonment for the remainder of the convict’s natural life, with death available at the top of two of them, and left the sentencing court no statutory route below the floor whatever the facts. Whether that removal of discretion strengthens or weakens convictions is one of the live questions of the field, and it is taken up in the sentencing debate below rather than prejudged here.

The matrimonial half of the 1983 wave answered a different emergency, the pattern of harassment and death inside the matrimonial home that the dowry debates of the period had made impossible to ignore. The second Act’s own Statement of Objects and Reasons, as the Supreme Court has since quoted it, recorded that the increase in dowry deaths was a matter of serious concern, that the Joint Committee of the Houses appointed to examine the working of the Dowry Prohibition Act had commented on the extent of the evil, and that in a number of cases the cruelty of the husband and his relatives culminated in the suicide or murder of the woman. The diagnosis explains the Act’s two part design: a substantive offence for the cruelty itself, which the prohibition statute of 1961 had never created, and an evidentiary presumption for the case in which the cruelty ended in the woman’s death by her own hand, where the principal witness to the household’s interior was by definition unavailable. Section 498A made it an offence, punishable with up to three years and a fine, for a husband or his relative to subject a married woman to cruelty, defined to include wilful conduct likely to drive her to suicide or to cause grave injury or danger to life, limb or health, and harassment intended to coerce her or her family to meet an unlawful demand for property or valuable security. The same amending Act inserted Section 113A into the Evidence Act, under which a court may presume that a married woman’s suicide within seven years of her marriage was abetted by her husband or his relative where she had been subjected to cruelty, a discretionary presumption whose contrast with the mandatory dowry death presumption that followed in 1986 is examined in the specialist treatment of the evidence. Section 198A of the procedure code channelled cognizance of the cruelty offence through a police report or a complaint by the aggrieved woman or specified relatives, a gatekeeping choice that shaped how the offence would be used. Three years later the Dowry Prohibition (Amendment) Act of 1986, Act 43 of 1986, added the dowry death offence, Section 304B, and the mandatory presumption in Section 113B of the Evidence Act; the full record of that offence and its uneasy relationship with the prohibition statute is carried by the Dowry Prohibition Act article, and it is placed here only to fix the sequence, because accounts that assign the cruelty offence to 2013, or the dowry death offence to 1983, corrupt every comparison built on them.

Why a comparison of the three amendment waves includes the cruelty offence at all is a question worth answering directly, since the offence is not a sexual offence and sits in a different part of the Penal Code.

Why does this comparison include the cruelty offence created in 1983?

Because the 1983 wave was two Acts, not one, and its second addressed the place where the period’s documented harm was concentrated, the matrimonial home. A comparison counting only the rape provisions would halve the wave, misdate the offence, and miss its design, which paired trial law for sexual offences with a new offence and presumption for cruelty inside marriage.

The cruelty offence also carries the 1983 wave’s most contested afterlife, and the comparison must take a position on it that agrees with the record the series has already established. The charge that the provision is misused as a lever in matrimonial disputes was made in the Supreme Court itself, in Sushil Kumar Sharma in 2005, which warned that the provision could unleash legal terrorism if used as an instrument of revenge, while upholding it; it was developed in Preeti Gupta in 2010, and it produced institutional responses in Arnesh Kumar in 2014, which subjected arrest in cases punishable with up to seven years to a recorded checklist, and in Rajesh Sharma in 2017, which directed that complaints pass first through district Family Welfare Committees. A three judge bench in Social Action Forum for Manav Adhikar in 2018 set the committee directions aside as judicial legislation, and the mechanism then returned by a separate route: Allahabad High Court guidelines in Mukesh Bansal in 2022, approved by the Supreme Court in Shivangi Bansal v. Sahib Bansal in July 2025, in a judgment of a bench presided over by the Chief Justice, which directed that the safeguards remain operational without overruling the 2018 holding. On the other side of the argument stand the provision’s defenders, among them the All India Democratic Women’s Association and lawyers such as Flavia Agnes and Kirti Singh, who answer that low conviction figures reflect the ordinary failure of prosecutions dependent on family witnesses rather than falsity, and that a complaint within a household that controls the evidence is structurally the hardest complaint in the criminal calendar to prove. This article does not reopen that debate; the judicial response to the provision has its own owner in the account of the Section 498A litigation sequence. What the comparison needs from the afterlife is one observation: the cruelty offence is the only provision of the 1983 wave that generated a forty year argument about its own use, which is itself evidence that the wave created working law rather than declaratory law.

The 2013 wave: the definition rebuilt and the catalogue enlarged

The 2013 wave is the only one of the three that began with a committee rather than a bill, and the committee’s speed became part of the story. The committee constituted on 23 December 2012 under Justice J. S. Verma, a former Chief Justice of India, sitting with Justice Leila Seth and Gopal Subramanium, reported on 23 January 2013, thirty days later, with a draft bill attached. The executive legislated first by ordinance, promulgated on 3 February 2013, and Parliament then enacted the Criminal Law (Amendment) Act, 2013, Act 13 of 2013, passed by the Lok Sabha on 19 March and the Rajya Sabha on 21 March, assented to on 2 April 2013 and deemed in force from the ordinance’s date. The full passage story, the standing reform bill that preceded the committee and the negotiations inside the budget session, belongs to the statute’s owner article, the Criminal Law (Amendment) Act 2013 article; what the comparison needs is the shape of what the wave changed, set against the wave on either side of it.

Begin with the definition, because the 2013 wave is the only one that rebuilt it. Section 375 was rewritten to describe the offence by seven descriptions of act, extending far beyond the single act the old law had named: penetration of specified kinds, with objects and with parts of the body other than the penis, manipulation causing penetration, and application of the mouth, each became rape when done in the circumstances the section lists. The circumstances were reorganised around consent, and consent was defined for the first time in the code’s history: Explanation 2 provides that consent means an unequivocal voluntary agreement, communicated by words, gestures or any form of verbal or non verbal communication, signifying willingness to participate in the specific act. The age below which consent is irrelevant was raised from sixteen to eighteen. A proviso added that a woman who does not physically resist shall not, by reason only of that fact, be regarded as consenting. Each element answered a documented failure of the older law: the narrow definition that had excluded most sexual violation from the offence’s name, the judicial inquiry into resistance that the 1978 acquittal had made notorious, and the absence of any statutory statement of what consent is. The definitional change is the 2013 wave’s strongest claim in this comparison, because a definition governs every prosecution brought under it, in every court, every day the code operates.

Around the rebuilt definition the wave constructed new offences for conduct the old code had reached awkwardly or not at all. The acid attack provisions, Sections 326A and 326B, made causing grievous hurt by acid punishable with a minimum of ten years extending to life, with a fine directed to the victim’s medical expenses, and the attempt punishable with five to seven years; the acid attacker’s victim had previously been prosecuted, if at all, under the general hurt provisions, whose categories fitted burns and disfigurement badly. Sections 354A to 354D created offences of sexual harassment, assault or criminal force with intent to disrobe, voyeurism, and stalking, each graded, with the harassment provision dividing its acts between those punishable with up to five years and those punishable with up to three. Section 370 was substituted wholesale to create a trafficking offence built on recruitment, transport, harbouring, transfer or receipt by threat, force, coercion, abduction, fraud, abuse of power or inducement, for the purpose of exploitation. Section 376B carried forward the separation offence with its range raised to two to seven years. Section 376C consolidated the position of authority offences at five to ten years. Section 376A, substantially recast, punished rape that causes death or a persistent vegetative state with a minimum of twenty years extending to life or death. Section 376D set the gang rape minimum at twenty years. Section 376E made a repeat conviction for specified rape offences punishable with imprisonment for the remainder of natural life or death. The minimum for ordinary rape stayed at seven years, but the proviso that had let a court go below it for recorded reasons was removed, which converted the floor from a supervised presumption into an absolute rule.

The procedural and evidentiary layer of 2013 is the part most often skipped in summaries, and it is the part that connects this wave to 1983. Section 166A of the Penal Code made it an offence, punishable with six months to two years, for a public servant knowingly to disobey a legal direction regulating investigation, or to fail to record information given in relation to specified offences, prominently including the sexual offences; Section 166B punished a hospital that refused treatment in violation of the treatment duty with up to a year. The procedure code was amended to require information in sexual offence cases to be recorded by a woman police officer where the complainant is a woman, to have the complainant’s statement recorded by a judicial magistrate in specified cases, to time the medical examination and to condition it on consent, and, by Section 357C, to oblige every hospital, public or private, to provide first aid and medical treatment free of cost to victims of specified offences and to inform the police. The Evidence Act received Section 53A, making evidence of a victim’s character or previous sexual experience irrelevant to consent, and a proviso to Section 146 barring questions in cross examination intended to prove that the victim was of generally immoral character; Section 114A was retained and its field adjusted to the new offence structure. Read together, these provisions are an attempt, thirty years on, to do for the police station and the hospital what 1983 had done for the courtroom: to change the conditions the complainant meets at the points where complaints are made, recorded, treated and tested.

The wave’s classification choices complete its design and are easily overlooked because they sit in the procedure code’s First Schedule rather than in the Penal Code’s text. The 2013 Act graded its new conduct offences by bail in a way that disclosed a judgment about their ordinary gravity: a first conviction for stalking was left bailable, with a subsequent conviction non bailable, while voyeurism’s first conviction was bailable and its repetition was not, and the acid attack and aggravated provisions were non bailable from the start. The sexual harassment provision split its own acts between the graver physical forms, punishable with up to five years, and the remainder, punishable with up to three. These gradations matter to the comparison because they are the one place where the 2013 legislature visibly calibrated rather than maximised: the wave that is remembered as a severity wave in fact contains the most differentiated penalty architecture of the three, built offence by offence, while the 2018 wave, remembered in the same vocabulary, worked by raising floors wholesale across a category defined by a single fact, the victim’s age. Calibration is a form of drafting care, and its distribution across the waves does not follow the distribution of the waves’ reputations.

Between the first wave and the second lies a thirty year interval that the comparison should not skip, because the interval explains why the second wave took the form it did. The definitional question that 1983 had deferred did not go away. The Law Commission’s 172nd Report of 2000, on the review of rape laws, recommended widening the offence under a gender neutral name of sexual assault, among a wider set of changes worked out with the women’s organisations that had petitioned the Supreme Court in the Sakshi proceedings; the Court in 2004 declined to rewrite the definition by judgment and pointed to the legislature. A Criminal Law (Amendment) Bill built substantially on that inheritance was introduced in 2012 and was under examination by the parliamentary standing committee when the crime of December 2012 overtook the process, and the Verma Committee’s thirty day report then displaced the slower bill as the wave’s vehicle. The sequence is the drafting pattern of this field in its purest form: a decade of careful, unglamorous work, a standing committee’s examination, and then a crisis that used the accumulated work but discarded the process that had been refining it, with the result that some of the standing bill’s considered positions, including its committee’s preference on the age of consent, were lost in the transfer. When the comparison credits the 2013 wave with careful definitional drafting, the credit belongs in large part to the interval’s slow work; when it charges the wave with a number chosen without examination, the charge belongs to the crisis that consumed the slow work without completing it.

The wave’s silences are as much a part of the comparison as its contents, and they were deliberate rather than overlooked. The Verma Committee had recommended the removal of the exception that excludes a husband’s acts with his wife from the offence of rape; the ordinance and the Act both retained the exception, and the committee’s recommendation was rejected in terms. The committee had recommended retaining the age of consent at sixteen; Parliament set it at eighteen, aligning the code with the child protection statute enacted the year before and against the committee’s reasoning. The committee had recommended against the death penalty for rape, reasoning from the evidence on deterrence and from the danger to victims; the Act introduced death for the case of death or persistent vegetative state and for repeat offenders, against the recommendation. The committee’s structural recommendations, on the sanction required to prosecute members of the armed forces under special legislation, on command responsibility, and on disqualification of candidates charged with such offences, were not taken up at all.

Two features of the wave’s passage deserve separate record because they show the drafting pattern at work. The first is the distance between the ordinance and the Act on the offence’s name and structure. The ordinance of February 2013 had replaced rape with a gender neutral offence of sexual assault, in which perpetrator and victim could be of any gender; the bill as enacted restored the offence’s name as rape and its gendered structure, a man as perpetrator and a woman as victim, while extending the definition’s acts as the committee had proposed. The episode matters to the comparison because it marks the boundary of what the fast process could carry: the committee’s own report had been cautious on full gender neutrality, Parliament was more cautious still, and the question of offences against male and transgender victims was left to the older, narrower provisions, a gap the recodification of 2023 would reproduce rather than close. The second feature is the overlap the wave created with the child protection statute. The Protection of Children from Sexual Offences Act, 2012 had been in force for only a few months when the committee sat, and once the Penal Code’s age of consent also stood at eighteen, the same act against a person below eighteen could be charged under either code, each with its own courts, presumptions and punishments. The 2013 Act answered by amending the child statute itself, providing in its Section 42 that where an act is an offence under both statutes the offender is liable to the punishment greater in degree, and inserting Section 42A to declare the child statute additional to and not in derogation of other law. The answer managed the overlap without resolving the underlying duplication, and practitioners in this field have worked inside two parallel codes for every complainant below eighteen ever since, a structural cost of legislating a wave on top of a statute barely a year old. The pattern in what was filtered out is consistent enough to be stated as a rule of this field: recommendations that widen offences, raise punishments or discipline procedure pass; recommendations that redistribute power, between wife and husband, between the citizen and the security forces, or between the accused legislator and the electorate, do not. The 2018 wave would confirm the rule by legislating at the very point where the committee had advised restraint.

The 2018 wave: the tariff rebuilt by age

The 2018 wave is the narrowest of the three and makes no pretence of being otherwise. Its vehicle was the Criminal Law (Amendment) Ordinance, 2018, promulgated on 21 April 2018, replaced by the Criminal Law (Amendment) Act, 2018, Act 22 of 2018, passed by the Lok Sabha on 30 July 2018 and the Rajya Sabha on 6 August 2018, assented to on 11 August 2018 and deemed in force from the ordinance’s date. No committee reported before it. No definition was touched. The wave’s entire substantive content sits in the sentencing structure of the rape offences, regraded by the age of the victim, together with a set of procedural deadlines and a restriction on bail. Describing it accurately means resisting both the temptation to inflate it into a general reform and the temptation to dismiss it, because within its chosen dimension it changed the law more radically than either predecessor.

The regrading worked as follows. The minimum sentence for rape under Section 376(1) rose from seven years to ten. A new Section 376(3) punished rape of a woman under sixteen with a minimum of twenty years, extendable to imprisonment for the remainder of the convict’s natural life, with a fine directed to be just and reasonable and paid to the victim for medical expenses and rehabilitation. A new Section 376AB punished rape of a woman under twelve with the same twenty year minimum, the same natural life maximum, or death. For gang rape, a new Section 376DA punished the gang rape of a woman under sixteen with imprisonment for the remainder of natural life and a fine, and a new Section 376DB punished the gang rape of a woman under twelve with natural life imprisonment or death. The identity protection of Section 228A and the recording offence’s field in Section 166A were extended to the new sections, and Section 114A of the Evidence Act was extended to prosecutions under them. The child protection statute was amended in parallel, its provision awarding the greater of the punishments under the two codes being enlarged to cover the new categories, so that the two statutes did not undercut each other for complainants below eighteen.

The wave’s drafting technique deserves a credit that summaries rarely give it. The rape offences are enumerated, section by section, in lists scattered across the three codes: in the recording offence’s schedule of specified offences, in the identity offence, in the magistrate statement provision, in the explanation that removes the sanction requirement for prosecuting public servants for these offences, in the in camera rule, in the compensation provision that operates alongside the acid attack fine, and in the presumption’s own list in the Evidence Act. A new offence category that is inserted into the Penal Code but omitted from any of those lists arrives half built, punishable but outside the recording duty, or inside the trial rules but outside the presumption. The 2018 Act combed the codes for the lists and inserted its new sections into each of them, amending Sections 166A and 228A of the Penal Code, Sections 164, 173, 197, 309, 327 and 357B of the procedure code, and Section 114A of the Evidence Act, in a single consequential sweep. The technique is the opposite of the wave’s reputation: whatever the merits of its sentencing choices, its legislative carpentry was thorough, and the comparison should record that the most criticised wave in the set is also the most carefully consequentialised, a combination that the field’s arguments about drafting quality seldom accommodate.

The procedural layer was a layer of clocks. Section 173(1A) of the procedure code was amended so that investigation of the listed rape offences must be completed within two months of the information being recorded; the sub-section had existed since 2008 in a narrower form, directing that investigation into the rape of a child be completed within three months, and the 2018 Act at once widened its field to the whole run of rape offences and cut its period by a third. The same Act reached Section 309, the trial provision under which the inquiry or trial of the principal rape offences was to be concluded, as far as possible, within two months of the filing of the charge sheet, and deleted the qualifying words for the new child categories, converting a direction of effort into a direction in terms. A new provision in the appellate chapter set six months as the period within which appeals in the listed cases are to be disposed of. Section 438 of the procedure code, the anticipatory bail provision, was amended to bar anticipatory bail altogether for persons accused under Sections 376(3), 376AB, 376DA and 376DB. Each of these is a real change to the working process rather than to the sentencing text alone, and the comparison must credit them: a two month investigation deadline, even one honoured unevenly, alters the incentives of an investigating officer in a way a sentencing maximum does not. But the character of the wave is set by where its weight lies. Its definitional contribution is nil, its evidentiary contribution is the extension of an existing presumption to new sections, and its procedural contribution is deadlines within a system whose other constraints, forensic capacity, court strength, witness protection, it did not touch. What it did supremely was raise the price of conviction, category by category, to the highest levels in the code, including the introduction of capital punishment for the rape of a child below twelve, a step the expert committee of 2013 had examined and advised against. The wave is therefore the cleanest possible test of the proposition that severity is what reform consists in, and the sentencing debate below examines that proposition on its merits.

Why did the 2018 amendments begin as an ordinance?

Because the executive chose to legislate between sessions of Parliament, as it had in 2013, so that the enhanced punishments were in force from 21 April 2018 rather than after a bill’s passage. The ordinance route delivered speed at the known price: a text framed inside the executive, without a committee stage, which Parliament enacted in substantially the same form.

Three waves, one table

The table states the comparison on the dimensions that decide the verdict, with a final row classifying each wave’s principal contribution. The classification is derived from the rows above it rather than asserted independently: a wave whose heaviest entries sit in the offences row is a substantive wave, a wave whose heaviest entries sit in the evidence and procedure rows is an evidentiary and procedural wave, and a wave whose heaviest entries sit in the sentencing row is a sentencing wave.

Dimension	The 1983 wave	The 2013 wave	The 2018 wave
What triggered it	A Supreme Court acquittal in 1978 in an alleged custodial rape case, an open letter from four law teachers, and the Law Commission’s 84th Report of 1980	A crime in Delhi in December 2012 and sustained public protest	Cases involving child victims in early 2018 and sustained public protest
Drafting process	A reform bill pending from 1980, built on the Law Commission report; two Acts passed in the ordinary course	A committee constituted on 23 December 2012 reporting on 23 January 2013, an ordinance on 3 February 2013, then an Act deemed in force from the ordinance	No committee; an ordinance on 21 April 2018, then an Act deemed in force from the ordinance
Offences created or rebuilt	Aggravated custodial categories within Section 376(2); Sections 376A to 376D for separation and custodial intercourse; Section 498A cruelty; the identity offence in Section 228A	Section 375 definition rebuilt; acid attack, sexual harassment, disrobing, voyeurism, stalking and trafficking offences; Sections 376A to 376E recast; recording and treatment offences in Sections 166A and 166B	No new offence of conduct; new sentencing categories by age of victim in Sections 376(3), 376AB, 376DA and 376DB
Evidentiary and procedural changes	Section 114A presumption on consent inserted; trials in camera; Section 113A presumption on abetment of suicide; cognizance discipline in Section 198A	Section 53A character bar and Section 146 proviso; Section 114A retained; woman officer recording, magistrate statement, timed medical examination; free treatment under Section 357C	Section 114A extended to the new categories; investigation within two months; appeals to be disposed of within six months; anticipatory bail barred for the child categories
Sentencing changes	First mandatory minima in the field: seven years ordinary, ten years aggravated, with a recorded reasons escape below the floor	Ordinary minimum held at seven years but the escape below the floor abolished; twenty year minimum where death or a vegetative state results; death for that case and for repeat offenders	Ordinary minimum raised to ten years; twenty year minima for victims under sixteen and under twelve; death for rape and gang rape of a child under twelve
Recommended and not adopted	The definitional widening the reformers sought was largely deferred; the escape below the minimum was retained against the reform’s logic	The Verma Committee’s recommendations on the marital exception, on retaining sixteen as the age of consent, and against the death penalty were rejected; its structural recommendations were not taken up	No committee existed to be rejected; the wave legislated death at the point where the 2013 committee had advised against it
Principal contribution	Evidentiary and procedural	Substantive, with a substantial procedural layer	Sentencing, with procedural deadlines attached

The row on recommendations made and not adopted is worth reading down the columns rather than across, because it discloses the filter through which every wave passed. In 1983 the deferred item was the definition itself, the widening the reformers most wanted, postponed for thirty years while the trial law around the unwidened offence was rebuilt. In 2013 the rejected items were the structural ones, the marital exception, the security forces sanction, the electoral disqualification, together with the two exercises of restraint the expert committee had urged, on the death penalty and on the age line. In 2018 there was no committee and therefore nothing to reject, which is itself the datum: the wave that legislated most severely was the wave with no institutional filter at all, and its content shows it, a tariff assembled without any body charged to ask what the tariff would do to the process beneath it. A reader who wants to predict the content of the next wave could do worse than ask, at its first announcement, whether a committee has reported, because the presence or absence of that slow filter has determined the character of every wave in the set.

The table also exposes the asymmetry the verdict rests on. Only the 1983 column is heaviest in the rows that govern what happens before conviction. Only the 2013 column rebuilt the definition, the element of the law that every subsequent prosecution must use. Only the 2018 column is confined almost entirely to the row that operates after conviction, on the small fraction of complaints that reach that point. A reader who ranks the waves by their effect on the largest number of cases, weighted by the stage of the process each case actually reaches, will not produce the same ordering as a reader who ranks them by the severity of their maxima.

One complaint, three waves: the comparison at the level of a case

The dimensions in the table can be tested by walking a single complaint through the system and asking, at each stage, which wave is doing the work. Take the ordinary run of a case in this field in its most general form, described by its legal stages rather than by any facts. At the police station, the governing provisions are 2013’s: the duty to record, enforced by the offence in Section 166A when an officer fails to record information about a specified offence; the requirement that a woman complainant’s information in a sexual offence case be recorded by a woman police officer; and, after 2018, the two month clock that starts running on the investigation from the moment of recording. Nothing in the 1983 wave speaks to this stage at all, because the reformers of 1983 located the system’s failure in the courtroom; the station house layer of the law is entirely a creation of the later waves, and within it the operative provisions are the recording offence and the deadline, one from each of 2013 and 2018.

Between recording and trial sits the question of the accused’s liberty, and here the governing law is a layered product of the waves and of the case law they provoked. For the cruelty offence, arrest practice was reshaped not by any amendment but by the judicial sequence the offence generated, from the Arnesh Kumar checklist of 2014, which required recorded reasons against statutory conditions before arrest for offences punishable with up to seven years, to the committee and cooling period mechanisms that the courts built and rebuilt over the following decade. For the gravest sexual offences, the 2018 wave intervened directly, barring anticipatory bail for the child categories and leaving the accused to seek regular bail after arrest in provisions whose sentencing floors colour every bail court’s assessment. The stage illustrates a general truth the comparison relies on: the practical content of an amendment wave includes the judicial responses it triggers, and a wave that creates very high floors changes bail practice across its field without ever mentioning bail, because no court assesses the risk of flight or interference independently of the sentence the accused faces.

At the hospital and in the early evidence, the position is again 2013’s: the obligation on every hospital to treat free of cost and inform the police, the punishment for a hospital that refuses, the timed and consent conditioned medical examination, and the early recording of the complainant’s statement before a judicial magistrate, designed to fix her account before the pressures of a long trial could work on it. At the trial, the governing provisions change hands: the complainant’s identity is protected by a provision of 1983, the proceedings are held in camera under a rule of 1983 that later waves extended to new offences, her character and previous sexual experience are excluded from the inquiry into consent by provisions of 2013, and, if the charge is in the aggravated categories, the presumption on consent that decides the evidential posture of the trial is the device of 1983, extended by each successor. At sentencing, the governing provisions are 2018’s for any case involving a victim below sixteen, and a layered combination elsewhere: the floors of 1983 as rebuilt by 2013, the removal of the escape below the floor in 2013, and the capital provisions that entered at the gravest points in 2013 and widened in 2018. On appeal, the six month direction is 2018’s.

The walkthrough yields the comparison’s central observation in concrete form. The stages through which every case must pass, recording, treatment, statement, trial, are governed predominantly by the evidentiary and procedural work of 1983 and its extension in 2013. The stage that only a minority of cases reach, sentencing after conviction, is where the 2018 wave’s work is concentrated, and where the public argument about all three waves has always been loudest. A complainant’s practical position, the likelihood that her complaint is recorded, that her evidence is taken in conditions she can survive, that the presumption assists her on the element hardest to prove, was built by the waves that the headlines of their day treated as technical. The walkthrough also shows why the waves are complements in operation even where they are rivals in this comparison: remove any one of them and a stage of the ordinary case loses its governing reform, which is precisely why the verdict that follows ranks their marginal contributions rather than their absolute value.

The recodification: three waves carried into the Bharatiya Nyaya Sanhita

The Bharatiya Nyaya Sanhita, 2023, Act 45 of 2023, assented to on 25 December 2023 and brought into force on 1 July 2024, replaced the Penal Code, and with it every provision this comparison has tracked was renumbered and re-enacted. The general question of what the recodification changed across the whole code is owned elsewhere, in the comparison of the Penal Code and the Bharatiya Nyaya Sanhita; the narrower question here is what happened to the three waves’ work, and the answer is that it was carried forward essentially intact, which makes the recodification a fourth data point in the comparison rather than a fourth wave. Nothing in the sentencing architecture of 2018 was diluted, nothing in the offence catalogue of 2013 was removed, and nothing in the evidentiary settlement of 1983 was disturbed in its new home in the Bharatiya Sakshya Adhiniyam, where Section 114A’s presumption sits as Section 120, and the presumptions of Sections 113A and 113B sit as Sections 117 and 118.

The mapping a practitioner needs can be stated compactly. The definition of rape, Section 375, is Section 63 of the Sanhita, exceptions included. Punishment for rape under the old Section 376(1) and the aggravated categories descend into Sections 64 and 65, with Section 65 gathering the age categories the 2018 wave created, under sixteen in its first sub-section and under twelve in its second, death retained in the second. Section 66 carries the offence of rape causing death or a persistent vegetative state. Section 67 carries the separation offence at two to seven years. Section 68 consolidates the intercourse by a person in authority, the descendant of the 1983 custody offences by way of the 2013 recast, at five to ten years. Section 70 carries gang rape, setting the ordinary minimum at twenty years and providing imprisonment for the remainder of natural life or death where the victim is under eighteen, absorbing the 2018 gang rape categories into a single age line. Section 71 carries the repeat offender provision. Section 72 carries the identity offence of the old Section 228A. The conduct offences of 2013 sit at Sections 74 to 78, assault with intent to outrage modesty, sexual harassment, disrobing, voyeurism and stalking; the acid attack offences sit at Section 124; trafficking sits at Section 143; the cruelty offence sits at Sections 85 and 86, and dowry death at Section 80. The one substantive addition in this field is Section 69, which punishes sexual intercourse by deceitful means, including a promise of marriage made without intention of fulfilling it, with up to ten years; the one notable omission is that the Sanhita contains no equivalent of the Penal Code’s Section 377, an omission whose consequences are owned by the criminal law cluster’s treatment of that provision and are noted here only because the comparison would be incomplete without recording that the recodification, which diluted nothing the three waves built, nonetheless left that gap open.

The recodification also carried the procedural layer across, and the carrying matters for the comparison’s arithmetic. The two month investigation direction for the rape offences reappears in the successor procedure code’s investigation provisions, the in camera rule and the identity protections continue, and the presumptions continue in the Sakshya Adhiniyam in the terms already stated. For offences committed before 1 July 2024 the old codes continue to govern by the ordinary savings logic of repeal, so the two settlements, the Penal Code’s and the Sanhita’s, will run side by side in the courts for the life of the pending caseload, which in this field is measured in years. The practitioner consequence is that the three waves must be known in both numberings at once, and the academic consequence is that the waves’ case law, accumulated over four decades of Penal Code adjudication, remains the working commentary on provisions that were re-enacted in substantially identical words: a recodification that changes numbers and keeps texts conserves the jurisprudence, and the conservation was plainly deliberate, since a Parliament intending substantive revision of this field in 2023 would have had to reopen the marital exception, the age of consent and the capital provisions, and it reopened none of them.

The exception that survived every wave

Across the three waves and the recodification, one provision was examined repeatedly and changed only at its margin: Exception 2 to the definition of rape, under which sexual intercourse or sexual acts by a man with his own wife, the wife not being under eighteen years of age, is not rape. The age line within the exception moved, from fifteen in the old code as enacted, read up to eighteen by the Supreme Court in Independent Thought v. Union of India in 2017, which held that the exception could not protect intercourse with a wife below eighteen, the age the child protection statute and the 2013 amendment had made general. The exception for adult wives survived the Law Commission’s examination in 1980, which left it; survived the Sakshi litigation; survived the Verma Committee, which recommended its removal in the strongest terms the committee used anywhere in its report; survived the ordinance, the 2013 Act, the 2018 Act, and reappeared, restated at the eighteen year line, as Exception 2 to Section 63 of the Sanhita. A provision that outlasts four legislative opportunities to remove it is not an oversight. It is a settled legislative position, restated by silence each time, and the honest description of the position is that Parliament has repeatedly chosen to retain it while repeatedly enlarging every offence around it.

The case for removal, as the Verma Committee made it and as the petitioners in the constitutional challenge repeat it, is that the exception rests on a doctrine of irrevocable consent given at marriage which the rest of the code has abandoned: consent under the 2013 definition is act specific and revocable, the exception carves the marital relationship out of that principle, and the result is that the law protects a woman’s refusal everywhere except in the relationship where refusal is most likely to be overridden. Justice Rajiv Shakdher of the Delhi High Court, in the split verdict in RIT Foundation v. Union of India delivered on 11 May 2022, held the exception unconstitutional on that reasoning, finding it in violation of Articles 14, 15, 19 and 21 of the Constitution. The case for retention, as Justice C. Hari Shankar held in the same case, is that marriage is a relationship of a qualitatively different kind, involving a legitimate expectation and a web of reciprocal obligations that make the classification between marital and non marital intercourse an intelligible one for a legislature to draw, that the creation of a rape offence within marriage would carry consequences for the institution and for the criminal process that a court is not equipped to weigh, and that the decision belongs to Parliament. The Union government defended the exception in that litigation on substantially the parliamentary ground. The challenges, including appeals from the split verdict, are before the Supreme Court, where the matter remains undecided; this article states the position, records both arguments at the strength their holders give them, and predicts nothing.

One further feature of the position needs recording, because both sides of the argument use it. The exception removes forced intercourse within marriage from the offence of rape; it does not place the conduct outside the criminal and civil law altogether, and the distance between those two propositions is where much of the practical argument lives. A wife subjected to sexual coercion can invoke the cruelty offence, where the conduct meets its definition; she can plead sexual abuse as domestic violence under the Protection of Women from Domestic Violence Act, 2005 and obtain protection, residence and monetary relief through its civil machinery; and where the parties live apart under a decree of separation, the separation offence applies in terms. Defenders of the exception point to this apparatus and argue that the conduct is remedied without being named rape, and that the naming is the true stakes of the dispute, with the consequences for arrest, bail and social destruction that the name carries. The petitioners answer that a civil remedy and a cruelty prosecution of up to three years are not the law’s recognition of the wrong, that every other legal system element, the definition of consent included, treats the wrong as rape, and that a protection order is not a substitute for the criminal law’s judgment on the act itself. The comparison takes no side in that exchange beyond recording that the exception’s survival has never rested on a claim that the conduct is harmless; it has rested, in every forum that has sustained it, on the allocation of the decision to the legislature and on the distinctness of marriage as a legal institution.

The sentencing debate, at full strength on both sides

The 2018 wave forces the field’s deepest disagreement into the open, and the stance this series requires is that both sides be given at full strength before any weighing. The case for severity, as the Union government made it in Parliament in 2018 and in its answers to parliamentary questions since, is expressive and deterrent at once: sentences of this gravity mark the community’s condemnation of an offence against a child in the only vocabulary a sentencing provision has, and the prospect of death or of imprisonment for the remainder of natural life raises the expected cost of the offence for any offender who calculates at all. State legislatures have legislated on the same conviction: Andhra Pradesh enacted its Disha legislation in 2019 prescribing death for rape, and Maharashtra’s Shakti legislation followed the same course, each presented by its sponsors as a response their electorates demanded. Underlying the legislative position is a democratic argument that deserves to be taken seriously rather than patronised: where a community’s moral judgment about an offence is this settled and this intense, a sentencing law that refuses to register it does not elevate the community; it merely transfers the decision to a forum the community did not choose.

The case against, made by the Verma Committee in 2013 when it recommended against extending the death penalty, by the Law Commission in its 262nd Report of 2015, which recommended abolition of the death penalty for all offences other than those connected with terrorism and waging war, and by sentencing researchers such as Mrinal Satish of the National Law School, whose work on rape sentencing documents how discretion removed at the front of the system reappears as arbitrariness elsewhere in it, runs as follows. Detection and conviction, not sentence length, are the binding constraints in this field, so additional severity attaches to the small minority of cases that end in conviction and does nothing for the complainant whose case is never registered, never investigated to standard, or never proved. Very high mandatory minima change the behaviour of every actor upstream: witnesses turn hostile under family pressure where the consequence of testimony is a death or a whole life sentence for a relative or acquaintance, and since the overwhelming majority of offenders in the Bureau’s tables are persons known to the victim, that pressure is the rule rather than the exception; courts, faced with a floor they regard as disproportionate to the case in front of them, acquit on doubts they might otherwise have resolved the other way, because an acquittal is reversible in appeal and a disproportionate mandatory sentence is not correctable in conscience. Child rights organisations, among them the HAQ Centre for Child Rights, added the argument specific to 2018: where the victim is a child and the offender is, as the statistics show offenders against children usually are, a person known to and trusted by the family, a death penalty for the rape of a living victim removes the offender’s incentive to leave the victim alive, since the punishment can go no higher if she dies and the principal witness is gone. That argument is not a rhetorical device. It is the perverse incentive point, made first by the Verma Committee and repeated in the parliamentary debate by members who voted against the capital provisions, and the government’s answer to it, that the death penalty remains discretionary and reserved for the gravest cases, concedes the incentive exists while disputing its weight.

One measurable footprint of a wave does exist, and it belongs to the procedural rather than the sentencing side of the ledger. The Bureau recorded 24,923 cases of rape in 2012 and 33,707 in 2013, a rise of over a third in a single year, and it recorded offenders as known to the victim in 31,807 of the 2013 cases. Nobody seriously maintains that incidence rose by a third in twelve months; what rose was registration, driven by the publicity of the wave and by the new offence of a public servant’s failure to record, which converted the police station’s oldest defensive practice, the unrecorded complaint, into a prosecutable act. That is what a process provision looks like when it works: it moves the number of cases entering the system, which is the precondition of every other number the system produces, including convictions.

The empirical record available to both sides is thinner than either admits, and the comparison should say so. The National Crime Records Bureau’s Crime in India 2022 recorded 31,516 cases of rape and 4,45,256 cases of crime against women in that year, figures that measure registration rather than incidence, and which the Bureau’s own explanatory material treats with caution. What the Bureau’s disposal tables show across the period of the three waves is a system in which the large losses occur before sentencing: at registration, in investigation, in trials that run for years and end, in a large share of completed trials, in acquittal. Article 142’s record of the dowry offences shows the same shape in a neighbouring field. Neither side’s strongest claim is established by the data. The deterrence claim cannot be isolated from the reporting changes the same amendments produced, and the acquittal claim, though supported by the reasoning of judges in individual cases, has never been measured across the system. The severity side’s criminological answer, which completes its case and is too rarely stated alongside its expressive one, is incapacitation and signal combined: an offender serving imprisonment for the remainder of his natural life commits no further offences against the public during that term, whatever the deterrent position, and a tariff that the community regards as proportionate sustains the willingness of complainants and their families to invoke the system at all, since a process believed to end in a trivial sentence is a process many families will not enter. On this view, severity is not a substitute for enforcement but a condition of the reporting on which enforcement depends, and the 2013 registration figures can be read in its support: complaints rose when the state was seen to treat the offence as grave. The abolitionist answer is that incapacitation arguments justify long imprisonment, not death, and that the reporting argument proves too much, since the registration rise of 2013 coincided with procedural provisions that made reporting safer and with a recording offence that made refusal punishable, either of which explains the rise without recourse to severity. Both answers are live, and the field’s honest position is that the experiment that would separate them, a jurisdiction raising severity while holding procedure constant, has never been run in India, because every wave has changed both at once or changed severity alone for too short a period to read.

The disagreement is therefore genuinely two sided: it is partly empirical, about what severity does to behaviour inside a weak enforcement system, and partly moral, about what a community may express through punishment regardless of effect. This article’s verdict on the waves does not require resolving it, because the verdict turns on a prior question: not whether severity is justified, but where in the process each wave did its work.

What no wave attempted

A comparison gains definition from what its objects share as absences, and the three waves share a striking list of them. No wave created a witness protection mechanism, though every inquiry into failed prosecutions in this field identifies the unprotected witness, in a case where the accused is usually known to the complainant and often connected to her household, as a central cause of collapse; when a Witness Protection Scheme arrived in 2018, it came as an executive scheme endorsed by the Supreme Court in Mahender Chawla v. Union of India, not as an amendment to any code. No wave created a victim compensation architecture; the state funded schemes that operate in this field descend from Section 357A of the procedure code, inserted in 2009 by an amendment unconnected with any of the three waves, supplemented by the fine direction device the later waves built into individual sentences. No wave addressed the forensic and medical infrastructure on which its own timelines depend, the laboratories whose reports the two month investigation deadline waits for, or the supply of trained examiners whose evidence the medical provisions assume. No wave legislated for police accountability beyond the single recording offence of 2013, and none touched the size of the trial judiciary before which its deadlines would run. Fast track courts for these offences were announced as administrative and funding measures, and a dedicated fund for women’s safety projects was created by the executive in the budget that followed the 2013 wave; neither entered the statute book as part of an amendment. The fast track idea itself has a history that proves the point: special courts for rape cases were sanctioned as a centrally sponsored scheme after 2013, allowed to lapse into the ordinary court system in many states, and then revived in a dedicated scheme for rape and child protection cases from 2019, with the courts’ existence, staffing and continuation dependent on scheme funding and state uptake rather than on any statutory entitlement a complainant could invoke. A deadline written into a code is permanent until amended; a court created by a scheme lasts as long as the scheme, and the three waves’ silence on the institutional side is precisely the difference between those two kinds of legal instrument. The field’s procedural gains therefore rest on a foundation the waves never built, and the durability of the gains is hostage to administrative decisions that no amendment controls.

The absences are not a list of oversights to be laid at any one legislature’s door. They mark the boundary of what the crisis amendment form can do. A wave can rewrite a code because a code is text, and text can be rewritten in thirty days by a committee or in a week by an ordinance. A wave cannot appropriate money with the same ease, cannot recruit forensic scientists, cannot build courtrooms, and cannot change the incentives of a police force whose accountability runs through state governments and an unreformed structure that the Supreme Court’s directions in Prakash Singh v. Union of India in 2006 addressed without legislative follow through. The consequence for the comparison is a qualification that strengthens rather than weakens its verdict. Even the best located reform, the evidentiary and procedural work of 1983 and its extension in 2013, operates inside a system whose binding constraints sit outside the codes altogether, in capacity. The waves changed what the law says to the actors. The actors’ ability to comply is a different reform, one that no triggering case has ever generated, because it cannot be dramatised as punishment and cannot be enacted as an amendment to a section. A reader who understands why the waves share their absences understands the field’s politics better than a reader who knows only their contents.

The age of consent: the problem none of the waves solved

One provision of the 2013 wave has generated a difficulty of a kind none of its drafters intended, and the comparison should record it because it illustrates the cost side of crisis drafting with unusual clarity. By raising the age of consent from sixteen to eighteen while the child protection statute of 2012 already fixed eighteen as the line below which consent is legally impossible, the 2013 Act made every sexual act involving a person below eighteen a serious offence irrespective of the factual consent of the parties. The consequence, documented in case after case, is the prosecution of adolescent relationships: a couple of seventeen and eighteen, a marriage the families opposed and the statute does not recognise for this purpose, an elopement converted by a parent’s complaint into a rape and kidnapping prosecution in which the supposed victim testifies for the accused and the machinery runs on regardless. Judges in individual cases in the Madras High Court, among them the court in Vijayalakshmi v. State in 2021, and in other High Courts, have recorded the difficulty from the bench, acquitting or quashing where the statute permitted and remarking where it did not that the law’s protection was being used as an instrument against the person it was drafted to protect. The Law Commission returned to the question in its 283rd Report of 2023 and recommended retaining eighteen as the age, against the argument for restoring sixteen, while recommending guided judicial discretion in sentencing for cases involving adolescents between sixteen and eighteen where the facts show tacit approval, precisely because an absolute floor in such cases punishes what the evidence does not support treating as predation. The mechanism of the difficulty repays one more paragraph, because it is a mechanism and not merely a grievance. A prosecution of this kind is typically set in motion not by the person the statute casts as the victim but by her parent, on the breakdown of a relationship the parent opposed; the young woman then testifies, in the large majority of such cases that reach trial, that she acted of her own will, and the prosecution is left attempting to prove an offence whose every material fact except age is denied by its own principal witness. Age, however, is the one fact the statute makes decisive, so the case reduces to documentary proof of a birth date, often contested, in records that are themselves unreliable in the districts where such prosecutions cluster. Courts have responded at the only points the statute leaves open, in the assessment of age evidence and in sentencing, and acquittals in this class of case are correspondingly common, consuming the time of the special courts the protection scheme depends on. The Law Commission’s proposed sentencing discretion addresses the second of those points and not the first, which is why its report, careful as it is, reads as a partial answer: the difficulty is not in the tariff but in the offence’s threshold, and a threshold is a definitional matter, in the keeping of the legislature that fixed it at eighteen against expert advice in the middle of a crisis.

The episode does not discredit the 2013 wave; the same wave’s definitional work was careful and necessary. It shows what the drafting pattern described at the outset costs when a number chosen under pressure meets a social fact the pressure never examined, and it is the strongest available evidence that the quality of a wave depends on the slow work beneath it.

The waves measured against their own stated purposes

Before ranking the waves against each other, fairness requires measuring each against the purpose its own authors stated, since a reform can succeed at its aim and still rank below another whose aim was larger. The 1983 Acts stated their purposes in their objects clauses and in the reports behind them: to make the law respond to custodial abuse and to the pattern of cruelty and death in the matrimonial home, by strengthening punishment and by correcting the trial’s treatment of the complainant. Against that purpose the wave substantially succeeded on paper and partially in practice: its provisions entered the working law, its presumption and its courtroom rules are used daily, and its cruelty offence generated both the enforcement and the controversy that mark a living provision. Its stated purpose was the narrowest of the three, and it is the only wave whose enacted text matches its stated purpose almost exactly, because its authors, working from a Law Commission report inside the ordinary legislative process, promised only what a code amendment can deliver.

The 2013 committee’s terms of reference, as the owner article records, were to recommend amendments providing for quicker trial and enhanced punishment for sexual assault. Measured against the punishment half, the wave delivered in full and beyond the committee’s advice. Measured against the speed half, the wave delivered text, the magistrate statement, the medical timelines, the trial completion direction, without delivering the institutional capacity that speed requires, and trial duration in this field remained governed by the constraints catalogued above, which no wave addressed. The gap between the wave’s stated purpose and its achievement is therefore the largest of the three on the dimension its authors were actually asked about, a finding that should temper the wave’s reputation as the field’s complete reform. The 2018 wave stated the narrowest purpose of all, to enhance punishment for rape, above all of children, and to accelerate investigation and appeal; against that purpose it is the most completely successful wave in the set, because it did exactly and only what it announced. The comparison’s ranking does not follow from success against stated purposes. It follows from the size of the practical difference each purpose, achieved, makes to the people inside the process, and a fully achieved sentencing purpose touches fewer of those people, later, than a fully achieved evidentiary one. Stating the distinction matters, because much public argument in this field proceeds by sliding between the two measures: a wave is praised as successful, when the praise is earned only against its own announcement, and the announcement was drawn to fit what a crisis legislature can pass.

The verdict: the wave that changed the trial changed the most

The deciding factor can be applied. A reform makes a practical difference in proportion to two things: the number of cases it reaches, and the stage of the process at which it intervenes. Every registered complaint passes through recording, investigation and trial; a small fraction of complaints end in conviction, and only those reach the sentencing stage at all. A change to the law of proof or to the conditions of trial therefore operates on the whole flow of cases, while a change to the sentencing tariff operates only on the residue that survives it. That is not an argument against sentencing law, which does necessary expressive and punitive work. It is an arithmetic point about where legislative effort lands, and it orders the three waves without any assistance from rhetoric.

The cruelty offence supplies an independent confirmation of the test from within the set. Section 498A is, on its face, a modest provision, a three year maximum with no minimum and no severity features at all, and by the standards of the sentencing debate it should have been a footnote. Instead it became the most litigated, most argued over and most institutionally productive provision any of the waves created, generating arrest jurisprudence, a committee experiment, a three judge correction and a restoration, across four decades. The reason is that the offence changed behaviour inside the process: it made a complaint inside a household actionable at a police station, and every one of its controversies, about arrest, about screening, about misuse, is a controversy about process. A provision’s practical weight is measured by the behaviour it changes, and the cruelty offence changed the behaviour of complainants, police officers, magistrates and appellate courts, which is the test’s whole point stated in a single example.

On that test the 1983 wave made the greatest practical difference. Its subject matter was the trial itself. The presumption in Section 114A altered what a prosecution for the gravest categories had to prove, in every such trial, from the day the Act commenced. The closed courtroom and the protected identity altered what a complainant had to endure to reach a verdict, and those provisions reached cases far beyond the aggravated categories, because the identity offence and the in camera rule were not confined to custodial prosecutions in their practical effect on reporting and on court practice. The cruelty offence and its presumption worked the same change inside the matrimonial field, giving a complaint about conduct inside a household, the hardest evidence environment in the criminal law, a statutory form and an evidentiary assist. And the wave did all of this while attracting a fraction of the public attention its successors received, which is itself part of this article’s claim: in this field, attention and effect have been inversely related, because attention demands severity and effect lives in procedure. The verdict does not require denying 1983 its limits. Its definition stayed narrow, its minima carried an escape that sentencing courts used until the Supreme Court disciplined it, and the custody offences were drafted around a distinction between rape and intercourse not amounting to rape that later Parliaments would have to rebuild. The wave changed the trial more than any other wave changed anything.

The 2013 wave stands second, and its position is the one most likely to be argued with, so its claim should be given precisely. The rebuilt definition of rape is the single most consequential provision any of the three waves produced, because it governs the threshold question in every prosecution, and the new offences brought conduct within the criminal law’s reach that had previously been prosecuted as lesser wrongs or not at all. Its procedural layer, the recording offence, the treatment duty, the character evidence bar, genuinely extended the 1983 method into the police station and the hospital. Why, then, second rather than first? Because the comparison measures change against what existed. The 2013 wave built on a trial system whose evidentiary centre had already been moved in 1983; its marginal contribution, though large, was the enlargement of a working structure, where 1983 had created the structure’s decisive feature out of a system that began from suspicion of the complainant. A reader who weights the definition above the presumption will invert the order, and the table is built to let that reader see exactly which rows carry the disagreement. The verdict offered here weights what the prosecution must prove, and the complainant must survive, above what the offence is called, on the ground that most prosecutions in this field have historically failed on proof and attrition rather than on the absence of an offence name.

The 2018 wave stands third on this test, and the ranking is a consequence of its design rather than of its intentions. Its sentencing categories are the most severe in the field’s history, and its deadlines address real failures, but its provisions intervene at the end of the process and at its margins: the tariff after conviction, the pace of investigation and appeal. Where it will matter most is in the cases it most concerns, the gravest offences against children, and the comparison does not pretend those cases are few or unimportant. It records only that a wave which changes what follows conviction cannot, on the evidence of how cases are actually lost, have changed the practical position of complainants as much as waves that changed what happens before conviction is possible.

The test also disciplines prediction, which is the use a comparison article should finally make of it. A fourth wave in this field will come; the pattern guarantees the occasion even though nothing can say when. The test says what to look for when it arrives. If the fourth wave’s weight sits, as the pattern of public demand predicts, in new offences named for the triggering case and in higher floors for existing ones, its practical yield will be small whatever its severity, because the stages at which cases are lost will be untouched. To outrank the 1983 wave, a future amendment would have to do the unglamorous work: strengthen the presumption architecture where proof still fails, protect witnesses by statute rather than by scheme, fund the forensic chain its deadlines presuppose, and build the court capacity its timelines assume. Those items share a property that explains their absence from every wave so far: none of them can be announced as a punishment, and the politics of the triggering moment rewards only what can be announced as one. The comparison’s verdict is therefore also a forecast about the field’s reform politics, and it will be falsified, happily, by the first wave that spends its capital on the process instead of the tariff.

The verdict has edges, and a comparison that hides them is not to be trusted. For the individual child complainant in a case that does reach conviction, the 2018 categories are the difference between a sentence measured in years and one measured in a lifetime, and no system level arithmetic should be allowed to dissolve that. For the complainant whose assailant’s conduct fell outside the old definition, only the 2013 wave made her case a rape case at all. And the whole ranking is hostage to enforcement: a presumption that trial courts apply grudgingly does less work than its text promises, just as a deadline that investigating officers treat as directory does less than its drafters announced. What the comparison establishes is where each wave’s work is located, and the ranking follows from the location. That is the reusable rule this article leaves with the series: to assess any criminal law amendment, ask first whether it changed what happens inside the process or only what is written in the punishment clause, and be suspicious of any reform whose entire weight sits in the punishment clause, because that is the clause the fewest cases ever reach.

For the reader’s own use, the ranking resolves differently by purpose, and a decision article should say so. The advocate or law student needs, above all, the carrying forward: the provisions live in the Sanhita’s Sections 63 to 72, the presumptions in the Sakshya Adhiniyam’s Sections 117, 118 and 120, and the older case law remains the working commentary on texts that were re-enacted rather than replaced, so the three waves are studied as one continuous settlement rather than as three closed chapters. The researcher needs the classification in the table’s final row, because it converts a familiar narrative of escalating reform into a testable claim about kinds of reform, which can be applied to the next amendment in this field or any other. The journalist needs the anchors the headlines omit: that the cruelty offence dates from 1983 and not from the wave that dominates coverage, that the expert committee of 2013 opposed the death penalty that the wave of 2018 enacted, and that the exception for husbands has been retained by every legislature that has looked at it. Each of those anchors is checkable against the table in minutes, which is the use a comparison artifact is built for in a field where the sentencing headline is written first and the record is consulted afterwards, if it is consulted at all. This article is general legal information about a body of law, not advice on any individual case, and anyone facing a proceeding under these provisions needs advice on the facts of that case from an advocate who can see them.

Frequently Asked Questions

Q: How do the three waves of criminal law amendment on women compare?

The 1983 wave changed the law of proof and the conditions of trial, inserting the presumption on consent, closing the courtroom, protecting the complainant’s identity and creating the cruelty offence. The 2013 wave rebuilt the definition of rape, created new offences including acid attack, stalking, voyeurism and trafficking, and disciplined recording and treatment. The 2018 wave raised sentences by the victim’s age, adding twenty year minima and the death penalty for a child under twelve, with deadlines for investigation and appeal. Measured by where each intervened in the process, the comparison ranks the evidentiary wave first, the substantive wave second and the sentencing wave third.

Q: Which criminal law amendment introduced custodial rape provisions?

The Criminal Law (Amendment) Act, 1983, Act 43 of 1983, in force from 25 December 1983. It placed rape by police officers, public servants, and the management or staff of jails, remand homes and hospitals into the aggravated categories of Section 376(2) with a ten year minimum, and it created the lesser offences in Sections 376B to 376D for sexual intercourse by a person in custody or authority not amounting to rape. The architecture answered a documented pattern of abuse by custodians, and it survives, recast, in Sections 64, 65 and 68 of the Bharatiya Nyaya Sanhita.

Q: When was cruelty by a husband first made an offence?

In 1983, by the Criminal Law (Second Amendment) Act, 1983, Act 46 of 1983, the second of the two Acts that make up the 1983 wave. It inserted Section 498A into the Penal Code, punishing cruelty by a husband or his relatives with up to three years and a fine, and it inserted Section 113A into the Evidence Act, letting a court presume abetment where a married woman died by suicide within seven years of marriage after cruelty. The offence is frequently misdated to 2013 in popular accounts; the 2013 amendment did not create it. It sits in Sections 85 and 86 of the Bharatiya Nyaya Sanhita.

Q: Which Verma Committee recommendations were not adopted?

The committee’s structural and restraining recommendations were the ones rejected. Parliament retained the exception that excludes a husband’s acts with his adult wife from rape, rejecting the committee’s recommendation to remove it. Parliament fixed the age of consent at eighteen rather than the sixteen the committee recommended. Parliament introduced the death penalty for rape causing death or a persistent vegetative state and for repeat offenders, against the committee’s reasoned opposition. And the committee’s wider recommendations, on the sanction protecting armed forces personnel, on command responsibility, and on disqualifying candidates charged with sexual offences, were not taken up in any form.

Q: Why is the death penalty for child rape criticised?

The criticism, made by the Verma Committee in 2013, the Law Commission in its 262nd Report, child rights organisations and sentencing researchers, has three strands. Severity does not deter where detection and conviction are the binding constraints, so the penalty attaches to few cases. Very high mandatory punishments increase pressure on a child to retract and on courts to acquit, since most offenders are known to the family. And where the punishment for rape equals the punishment for murder, the offender gains an incentive to kill the principal witness. The government’s answer is that the penalty is discretionary and expresses the gravity of the offence.

Q: How did the Bharatiya Nyaya Sanhita carry these amendments forward?

By re-enacting them, renumbered, without substantive dilution, from 1 July 2024. The definition of rape became Section 63, exceptions included; punishment sits in Sections 64 to 66; the separation offence is Section 67; intercourse by a person in authority is Section 68; gang rape is Section 70, with the 2018 age categories absorbed at an eighteen year line; repeat offenders are Section 71; the identity offence is Section 72. The cruelty offence became Sections 85 and 86 and dowry death Section 80. The presumptions moved to the Bharatiya Sakshya Adhiniyam as Sections 117, 118 and 120.

Q: What is the current position on marital rape?

Sexual intercourse by a man with his adult wife is not rape, under Exception 2 to Section 63 of the Bharatiya Nyaya Sanhita, as it was not under the Penal Code. The exception does not protect intercourse with a wife below eighteen, after the Supreme Court read the age line up in Independent Thought in 2017, and it has never protected a husband separated under a decree, whose non consensual intercourse is an offence under Section 67. A constitutional challenge to the exception produced a split verdict in the Delhi High Court in 2022 and is before the Supreme Court, undecided.

Q: Which criminal law amendment made the biggest practical difference?

On the test this comparison applies, the 1983 amendments. They changed what happened inside prosecutions: the presumption on consent in Section 114A of the Evidence Act altered what the prosecution had to prove in the gravest cases, in camera trials and the protection of identity changed what a complainant had to endure, and the cruelty offence gave the matrimonial home’s harms a statutory form. The 2013 wave ranks second, on the strength of its rebuilt definition and new offences, and the 2018 wave third, because its changes operate mainly on the sentencing text that only convicted cases reach.

Q: How did the three waves differ in the way each was drafted?

The 1983 wave was drafted in the ordinary way, out of a Law Commission report of 1980 and a reform bill that waited for its political moment. The 2013 wave was drafted by an expert committee in thirty days, with a bill attached to its report, and reached force first as an ordinance before Parliament amended the text in session. The 2018 wave had no committee at all: the executive drafted it, promulgated it as an ordinance on 21 April 2018, and Parliament enacted it in substantially the ordinance’s form. The drafting route predicts each wave’s character: borrowed committee work produced the 2013 definitions, and its absence in 2018 produced a sentencing amendment with no definitional layer.

Q: What did the 2018 amendment change in the law of bail for rape cases?

It closed the door on anticipatory bail for the new child categories. The 2018 Act amended Section 438 of the Code of Criminal Procedure to provide that its anticipatory bail provisions do not apply to a person accused under Section 376(3), Section 376AB, Section 376DA or Section 376DB of the Penal Code, the provisions covering rape and gang rape of girls under sixteen and under twelve. Regular bail after arrest remained a matter for the courts under the ordinary provisions, and the bar did not extend to the other rape categories. The change sits in the Bharatiya Nagarik Suraksha Sanhita’s corresponding bail provision in the recodified procedure.

Q: How long do the police get to investigate a rape case after the 2018 amendment?

Two months from the recording of the information. The 2018 Act amended Section 173(1A) of the Code of Criminal Procedure so that investigation of the listed rape offences, including the aggravated and child categories, must be completed within two months. The deadline joined the obligation the 2013 Act had created in Section 166A of the Penal Code, under which a public servant who fails to record information about specified offences commits an offence punishable with six months to two years. The two provisions work as a pair: one criminalises the refusal to start the process, and the other time limits its conduct. Compliance with the deadline in practice depends on forensic and medical reporting, which the amendment did not address.

Q: What sentence does the gang rape of a child attract after the 2018 amendment?

It depends on the child’s age, because the 2018 Act graded gang rape by age for the first time. The gang rape of a woman under sixteen, under Section 376DA, attracted imprisonment for the remainder of the convict’s natural life and a fine directed to the victim’s medical expenses and rehabilitation. The gang rape of a girl under twelve, under Section 376DB, attracted the same natural life term or the death penalty. Ordinary gang rape of an adult remained under Section 376D, with its twenty year minimum from 2013. In the Bharatiya Nyaya Sanhita the categories were consolidated in Section 70, which sets a twenty year minimum and provides natural life or death where the victim is under eighteen.

Q: Why can very high minimum sentences make convictions harder to obtain?

Because every actor in the process prices the consequence of a conviction before it happens. Where the floor is twenty years, natural life or death, family witnesses in cases involving a known offender face pressure to turn hostile that a shorter tariff would not generate, and the Bureau’s tables show most offenders are known to the victim. A trial court that regards the mandatory floor as disproportionate to the case before it has no sentencing route available, and the only instrument left is the assessment of proof, which is why judges and sentencing researchers argue that doubt at the margin resolves into acquittal more often under very high minima. The argument is about the system’s behaviour, not about the gravity of the offence.

Q: What did the 1983 amendments do to protect a rape complainant’s identity?

Two things that had not existed before. The 1983 Act inserted Section 228A into the Penal Code, making it an offence punishable with up to two years to print or publish the name of, or any matter that may make known the identity of, a victim of the specified sexual offences, subject to limited exceptions such as authorisation by the victim or, where she is dead or a minor, by her next of kin to a recognised welfare organisation. The same Act amended the procedure code so that the inquiry and trial of rape and the custodial offences be conducted in camera. The identity offence survives as Section 72 of the Bharatiya Nyaya Sanhita, and later waves extended its list of covered offences rather than altering its design.

Q: How did the 2013 amendment change the punishment for repeat offenders?

It created a specific repeat offender provision. Section 376E, inserted by the 2013 Act, provided that a person previously convicted of rape under the specified provisions who is subsequently convicted of any of them again shall be punished with imprisonment for life, meaning the remainder of that person’s natural life, or with death. Before 2013 the code had no dedicated enhancement for a second rape conviction, and a repeat offender was sentenced within the same ranges as a first offender, with the prior record relevant only as a matter of judicial assessment. The provision was carried into Section 71 of the Bharatiya Nyaya Sanhita in the same terms, and it was one of the two points at which the 2013 Act introduced capital punishment against the Verma Committee’s advice.

Q: When was the offence of acid attack created, and what does it punish?

The 2013 Act created it, by inserting Sections 326A and 326B into the Penal Code. Section 326A punishes whoever causes permanent or partial damage or deformity, burns, maims, disfigures or disables, or causes grievous hurt, by throwing or administering acid, with imprisonment of not less than ten years extending to life, and a fine that must be just and reasonable to meet the victim’s medical expenses and must be paid to the victim. Section 326B punishes the attempt, throwing or attempting to throw acid with the relevant intention, with five to seven years. Acid attacks had previously been prosecuted under the general provisions on grievous hurt, which neither named the method nor directed the fine to treatment. The offences sit at Section 124 of the Bharatiya Nyaya Sanhita.

Q: How do the fines imposed for rape reach the victim after the later amendments?

By statutory direction, in the categories where the direction applies. The pattern began with the acid attack provision of 2013, whose fine must be just and reasonable to meet the victim’s medical expenses and must be paid to the victim. The 2018 Act extended the same structure across its new categories: the fines under Sections 376(3), 376AB, 376DA and 376DB are directed to be just and reasonable and to be paid to the victim towards medical expenses and rehabilitation. The Sanhita retained the direction in Sections 65 and 70. The design converts the fine, traditionally a payment to the state, into a partial compensation mechanism inside the sentence itself, alongside the separate victim compensation schemes administered under the procedure code.

Q: How quickly must an appeal in a rape case be disposed of after the 2018 amendment?

The 2018 Act set six months as the period for disposal of appeals in the listed rape cases, by amending the appellate chapter of the Code of Criminal Procedure. The direction covers appeals arising from the rape provisions including the aggravated and child categories the wave created. It completed the wave’s set of clocks: two months for the investigation from the recording of information, and six months for the appeal after conviction or acquittal. Like the investigation deadline, the appeal direction states a norm the system’s capacity has to meet, and its presence in the statute gives an appellant or a complainant’s family a textual basis for pressing a court for an early hearing that did not exist before 2018.

Q: Which wave changed the definition of rape itself?

Only the 2013 wave. The 1983 Act added one clause to Section 375, covering consent vitiated by unsoundness of mind, intoxication or a stupefying substance, and otherwise left the definition’s narrow physical core untouched. The 2013 Act rewrote the section: it described the offence by seven descriptions of act, extending it to penetration with objects and other body parts and to oral acts, defined consent in Explanation 2 as an unequivocal voluntary agreement communicated by words, gestures or other communication, and provided that absence of physical resistance is not consent. The 2018 Act did not touch the definition at all. Under the Sanhita the definition, as rewritten in 2013, sits in Section 63.

Q: What did the Law Commission recommend on the age of consent in 2023?

In its 283rd Report of 2023, the Law Commission recommended retaining eighteen as the age of consent, against the argument for restoring sixteen, on the ground that lowering the line would weaken the protection against child sexual exploitation and trafficking that the higher line supports. But the Commission also recorded the cost the line imposes in prosecutions of adolescent relationships, and recommended guided judicial discretion in sentencing where a case involves a child between sixteen and eighteen and the facts show tacit approval, so that the mandatory sentencing structure would not apply with full force to a relationship the evidence does not show to be predatory. The recommendation had not been enacted as part of any amendment wave.