There is a comfortable way to talk about violence against women in India and it usually begins with a demand for a new law. The instinct is misplaced. On the statute book, the Indian woman is among the better-protected anywhere.
The Dowry Prohibition Act has stood since 1961. Section 304B of the Penal Code, now Section 80 of the Bharatiya Nyaya Sanhita, makes dowry death a distinct offence carrying a minimum of seven years. Section 498A, now Sections 85 and 86, criminalises cruelty. Section 113B of the Evidence Act goes further than most penal provisions ever do for a complainant: once the foundational facts are shown, it presumes the accused caused the death and shifts the onus onto him. A prosecutor in a dowry-death trial begins, in other words, with the law's thumb already on her side of the scale.
And still the conviction rate sits between roughly 11 and 17 per cent (NCRB, Crime in India). Most cases never finish trial at all. That gap, between a framework this protective and outcomes this poor, is the only thing worth talking about; and it is not a drafting problem. You cannot amend your way out of it, because the defect is not in the text. It is in everything that happens after the text is invoked.
Consider what that machinery actually consists of. It is the officer who decides whether a first information report (FIR) is registered or whether the woman is sent home to make peace. It is the investigating officer who does or does not secure the scene, order the post-mortem, seize the relevant material while it still exists. It is the witness who turns hostile, not from caprice but under a pressure the system does nothing to relieve. It is the public prosecutor carrying an impossible docket and the trial that runs long enough for memory and resolve to decay. Each of these is a human being formed in the same society that produced the offence, holding the same assumptions about family, honour and a wife's place that the legislation was written to override.
The reason Arnesh Kumar v. State of Bihar recorded chargesheeting above 90 per cent alongside the lowest conviction rate of any comparable head is not that the cases are false. It is that the system can arrest and charge and then cannot, or will not, prove. An acquittal is not an acquittal of the law. It is an indictment of its enforcement.
This is also the honest answer to the misuse argument, which has dominated this debate for two decades and distorted it throughout. Yes, a measurable fraction of Section 498A complaints are closed as false; the NCRB records it and Sushil Kumar Sharma v. Union of India and Arnesh Kumar responded with procedural safeguards, correctly. But the figure trotted out to prove mass fabrication is the acquittal rate. And that is a category error a lawyer should not indulge. A case that ends in acquittal, or is closed for want of evidence, is not a case found to be false. Conflating the two converts the State's own investigative failure into the complainant's dishonesty.
The disciplined response to genuine misuse is better procedure, which we have. It is not the dilution of protection for the far larger number of women the system never reaches in the first place. Indira Jaising put the asymmetry more memorably than I can. Speaking at the launch of her book The Constitution Is My Home, she observed that when it comes to violence, women are more honoured by the courts in their death than in their life: a dowry death draws long lectures from the bench, but the woman who is still alive and asks the court to protect her is met with the suspicion that she is misusing the law. The concern for the dead and the condemnation of the living are, as she has written elsewhere, the same system viewed from two ends.
The constitutional objection to all this is not subtle and it has been available for years. A regime of protective law that is systematically under-enforced for one class of victims does not sit easily with Article 14 and the matrimonial home is where its failure is most complete. The clearest illustration is the one provision the legislature had every opportunity to fix and chose to preserve. When parliament enacted the Bharatiya Nyaya Sanhita in 2023 and described it as the decolonisation of our criminal law, it carried forward, as Exception 2 to Section 63, the marital rape exception, a doctrine traceable to a single unsourced sentence of Hale's from 1736, which England itself discarded in R v R in 1991. The challenge to it, after the split verdict in RIT Foundation, remains pending before the Supreme Court.
The contradiction has since sharpened to the point of absurdity. In March 2026, the Madhya Pradesh High Court found itself compelled to quash charges of unnatural offences against a husband, reasoning that once the 2013 amendment folded such acts into the expanded definition of rape and the rape provision exempts husbands, no offence survived. A stranger who commits the act is a rapist. A husband is, in the eyes of the criminal law, nothing at all. The same body, the same act, the same harm; the only variable is a marriage certificate, which the law treats not as a bond of trust but as a grant of immunity. A court reached that conclusion not because it wished to, but because the statute left it no room. That is the signature of a problem the judiciary cannot cure by interpretation. It requires the legislature to act and the legislature has twice declined - in 2013 after the Justice Verma Committee recommended removal and again in 2023.
None of this is severable from money and this is the part the rights discourse too often omits. A woman's formal entitlements are worth precisely as much as her capacity to act on them and that capacity is economic. The right to refuse a marriage, to leave a violent one, to insist on her share, all of it presupposes that she can support herself outside the marital household. Most cannot and the law has helped ensure it. Vineeta Sharma v. Rakesh Sharma affirmed the daughter's coparcenary right by birth, yet the lived practice is that she is induced to sign it away at marriage, the wedding treated as a settlement of her inheritance. She enters adulthood equal on paper and dependent in fact.
The labour data compounds the point and should be read with care, because it is routinely misread. The recent, much-trumpeted climb in women's workforce participation - to roughly 41.7 per cent in 2023-24 (PLFS 2023-24) - conceals a decade in which it had first slumped into the low twenties. And the rebound, when it came, was no advance at all. It was overwhelmingly rural women pushed into self-employment and unpaid family labour - a retreat into agriculture rather than entry into secure work - with more than 9 in 10 women workers stranded in the informal economy, beyond the reach of every protection we keep legislating.
Our maternity mandate of 26 weeks is among the most generous anywhere and touches under a tenth of working women, while its employer-borne cost hands firms a quiet incentive not to hire women of child-bearing age at all. The unpaid labour that holds households upright - valued conservatively at 3.1 per cent of GDP on the Time Use Survey and far higher in truth - stays uncounted, uncompensated and dismissed as no work at all. A jurisprudence of equality that ignores this material base is decorative.
Put the three together - the enforcement that fails, the Constitution that is honoured in form and breached in practice, the economic dependence that makes exit impossible - and the picture is not of a country short of laws. It is of a country whose institutions - and the people who run them - have not yet done the harder work the laws assume. The remedies are not mysterious. Registration that is monitored rather than discouraged. Investigation and evidence preservation done competently and early. Witness protection that means something. Trials that conclude before the complainant's life has moved on without her. Consequences for officials who shield the influential. Gender-sensitivity training for everyone in the chain, built on a plain handbook drafted by lawyers, judges, clinicians and social workers and put in the hands of the constable who first meets the woman at the station. And, overdue by any measure, the removal of Exception 2 and accession to the Optional Protocol to CEDAW, which India signed the parent Convention to in 1993 and has kept at arm's length since.
The woman who was promised the protection of the law and discovered at the police station, in the courtroom, in her own home, that the promise was notional, is not asking for a new statute. She is asking that the ones we have be made to work. That is a less stirring demand than law reform and a more difficult one, because it implicates not the drafter but the enforcer, which is to say all of us who operate this system. It is also the only demand that would change anything.
Purva Singh is an international law and policy advisor and the founder of Bhagwani Consultancy.
A longer, referenced version of this argument is available on SSRN.