Family-Centred Crime: Dowry, Domestic Violence and Child Maltreatment

Criminal justice systems were built primarily to respond to harm between strangers: an assault on a street, a theft from a shop. Harm inside a household ran into a separate legal tradition treating the family as a private domain the state should not enter except in extreme cases.

English common law's doctrine of marital unity, under which a wife's legal identity merged into her husband's, left a husband's physical discipline of his wife largely outside judicial reach into the nineteenth century, and equivalent doctrines of parental authority shielded the physical punishment of children.

This privacy doctrine did not disappear when formal legal disabilities were removed. It persisted as an informal norm: police reluctance to intervene in a domestic dispute, a presumption that a family should resolve its own conflicts, and a victim's own reluctance to invite state involvement into a household on which she may be economically dependent.

Sociologists studying family violence in the 1970s, notably work associated with Murray Straus and Richard Gelles in the United States, argued that the family was in fact one of the most violent institutions in society precisely because it operated outside the scrutiny that governs public life.

Three structural features compound the detection problem. First, the offender and the primary witness usually live under the same roof, so there is rarely an independent bystander to corroborate an account.

Second, the victim frequently depends on the offender for housing, income or immigration status, which makes reporting a decision with immediate practical costs, not only an emotional one. Third, family violence is often repeated rather than a single incident, so any one report captures only a fragment of a longer pattern, and a victim who withdraws an early complaint may be treated as unreliable rather than as coerced.

The response, described across the rest of this topic, has been to build tools designed for cases with no independent witness: presumption clauses that shift an evidentiary burden once foundational facts are proved, civil orders requiring a lower standard of proof than criminal conviction, and reporting duties that place the obligation on someone other than the victim.

Dowry, in the South Asian context, is a transfer of cash, goods or property from a bride's family to the groom or his family at or after marriage.

Where it began as a voluntary gift meant to give a daughter a measure of independent property, it hardened in many communities into a coercive precondition for marriage, with a groom's family setting an explicit demand and escalating it after the wedding. Because the payment flows from the bride's family while the pressure to pay lands on the bride, an unmet demand routinely becomes the pretext for harassment inside her marital home.

India's Dowry Prohibition Act, 1961 was the first legislative attempt to suppress the practice directly. Section 3 criminalises giving, taking or abetting the giving or taking of dowry, and Section 4 separately criminalises demanding dowry from the bride's or groom's family, each carrying a minimum custodial term.

The Act proved difficult to enforce on its own: both families to a dowry transaction are typically complicit, which removes the ordinary incentive for either side to report it, and the Act said little about the violence that followed a demand rather than the demand itself.

That gap was closed by two companion offences originally in the Indian Penal Code and now carried forward, largely unchanged, into the Bharatiya Nyaya Sanhita, 2023, in force since 1 July 2024. Section 85 criminalises cruelty by a husband or his relatives toward a wife, defined in Section 86 as wilful conduct likely to drive her to suicide or grave injury, or harassment to coerce an unlawful demand for property, applying whether or not the wife dies.

Section 80 creates the separate offence of dowry death: an unnatural death within seven years of marriage, preceded by dowry-related cruelty soon before death. The two offences address different points on one timeline, and one case can support both charges if the wife later dies.

Dowry is not confined to India. Dowry-adjacent marriage payments with associated coercion are documented in parts of South Asia more broadly, and similar marriage-related financial coercion surfaces in diaspora communities in the United Kingdom and elsewhere, which is one reason UK domestic-violence guidance now names dowry abuse as a recognised pattern rather than a wholly foreign concept.

Dowry-related violence follows a recognisable escalation pattern rather than arriving as a single event. A demand is typically raised at the time of marriage negotiation, renewed after the wedding when the groom's family judges the initial payment insufficient, and escalated through verbal harassment, then confinement or deprivation, and in the most severe cases culminates in a fatal assault or in the wife's death staged to resemble a suicide or an accident, commonly a kitchen fire.

This timeline creates several distinct points, family intervention, a neighbour's report, a police complaint, a protection order, at which the escalation could in principle be interrupted, though each point carries the same underlying obstacles described above: dependency, stigma and the absence of an independent witness.

Because dowry deaths are frequently staged and occur without witnesses, prosecutors historically struggled to prove the causal link between dowry harassment and a death that looked accidental.

India's response was an evidentiary presumption, originally Section 113B of the Indian Evidence Act, 1872 and now carried forward as Section 118 of the Bharatiya Sakshya Adhiniyam, 2023. Once the prosecution proves the deceased was subjected to cruelty or harassment for a dowry demand soon before death, the court presumes the accused caused the dowry death, shifting the burden to the accused to rebut it rather than requiring the prosecution to prove causation from scratch.

Courts read the phrase soon before death narrowly: a long, cold gap between the last proven act of cruelty and the death defeats the presumption, and appellate courts have overturned convictions where the prosecution relied on it without first establishing that proximity. The presumption is a tool for a specific evidentiary gap, proving causation in a death that occurred in private, not a general licence to convict on suspicion of dowry harassment alone.

India's National Crime Records Bureau records several thousand dowry deaths annually, a figure that has stayed broadly stable despite awareness campaigns and legal reform, consistent with the attrition pattern discussed later in this topic: legal tools exist, but under-detection keeps the recorded and the actual scale of the problem some distance apart.

Domestic violence occurs across every income level and jurisdiction, with or without a dowry element, and criminology has developed two influential models for how it operates as a pattern rather than an isolated act.

Lenore Walker's cycle of violence, set out in her 1979 book The Battered Woman, describes a repeating three-phase sequence: a tension-building phase of escalating minor conflict and unpredictability, an acute battering incident in which the built-up tension discharges as an assault, and a reconciliation or honeymoon phase in which the abuser expresses remorse, promises change and temporarily restores affection, before tension begins building again.

Walker observed that with repetition the cycle tends to shorten and the violence within it tends to intensify, and that the honeymoon phase is precisely what makes the relationship psychologically difficult to leave, since it offers periodic, genuine-seeming evidence that the relationship can still work.

Evan Stark's coercive control, developed across decades of clinical work and published as a full model in his 2007 book Coercive Control: How Men Entrap Women in Personal Life, reframes domestic abuse around a different axis.

Stark argued that physical violence, even where present, is often not the primary harm; the primary harm is a sustained pattern of intimidation, isolation from friends and family, monitoring of movement and communication, and control over money and daily decisions, which together restrict a victim's liberty in a way closer to hostage-taking than to a series of discrete assaults.

A victim under coercive control can suffer profound harm with few, or even no, incidents that would independently qualify as a criminal assault, exactly what earlier assault-based legal frameworks failed to capture.

The two models are complementary rather than competing. Cycle-of-violence explains the emotional and psychological mechanism that keeps a victim inside an abusive relationship despite repeated harm, which matters for understanding why victims recant, delay reporting or return to an abuser after leaving.

Coercive control explains why a relationship can be severely harmful and dangerous even where physical violence is infrequent or entirely absent, which matters for legal reform, since a law built only to punish discrete assaults will miss this pattern altogether.

Stark's model directly shaped statutory reform. England and Wales created a standalone offence of controlling or coercive behaviour in an intimate or family relationship under Section 76 of the Serious Crime Act 2015, which came into force in December 2015 and does not require proof of any single violent act, only a repeated or continuous pattern that the offender knew or ought to have known would cause serious alarm or distress. This was one of the first criminal statutes anywhere to give coercive control, as opposed to assault, its own offence definition.

Because domestic violence victims often need immediate physical safety more urgently than a criminal conviction months or years later, most jurisdictions run a civil protection-order track alongside criminal prosecution rather than relying on either alone.

India's Protection of Women from Domestic Violence Act, 2005 is a civil statute: it creates no new criminal offence but lets an aggrieved woman, or a Protection Officer appointed under the Act, apply to a magistrate for a protection order, a residence order preserving her right to stay in the shared household, and monetary relief, on the ordinary civil standard of proof, markedly lower than proof beyond reasonable doubt.

The United States built a comparable dual track earlier. The Violence Against Women Act, 1994 created the first comprehensive federal criminal framework addressing domestic violence, funded investigation and prosecution, and required every state to give full faith and credit to a protection order issued in any other state, closing a gap in which an abuser could simply follow a victim across a state line.

Individual states separately provide civil restraining orders obtainable on an expedited basis, often the same day for an emergency order, apart from any criminal charge.

The United Kingdom layers a criminal offence, coercive or controlling behaviour under the Serious Crime Act 2015, alongside civil non-molestation and occupation orders available under family law, and consolidated much of this framework under the Domestic Abuse Act 2021, which for the first time gave domestic abuse a statutory definition covering physical, emotional, coercive and economic abuse.

South Africa reached the civil-first model even earlier. Its Domestic Violence Act, 1998 defines domestic violence broadly, physical, sexual, emotional and economic abuse, intimidation, harassment and other controlling behaviour, and lets a magistrate grant an interim protection order the same day, before the respondent is even served, with a final order and reliefs such as removal of a firearm following once the respondent has had a chance to answer.

As in India, the order itself is civil, but breaching it is a separate criminal offence carrying up to five years' imprisonment, which gives the civil remedy real teeth without requiring a criminal trial on the underlying conduct first.

The practical logic across these jurisdictions is the same: a civil order can be obtained quickly, on a lower evidentiary bar, and gives a victim breathing room, while a criminal prosecution takes longer, requires a higher standard of proof, but carries a custodial sentence that a civil order cannot.

Running both tracks means a victim is not forced to choose between rapid protection and eventual accountability, though it also means the systems can move at very different speeds for the same underlying facts, a mismatch that legal aid and victim-support services in each jurisdiction are specifically designed to help navigate.

Child maltreatment is conventionally divided into four categories, and criminology and child-protection practice treat the distinction as important because each has a different detection pathway and a different evidentiary problem.

Physical abuse is non-accidental injury, historically under-recognised until Henry Kempe and colleagues named it battered-child syndrome in a landmark 1962 paper in the Journal of the American Medical Association, which described a clinical pattern of repeated fractures and injuries inconsistent with a caregiver's explanation. The paper's clinical framing, presenting the pattern as a medical syndrome a doctor could recognise, is credited with prompting the first legislative reporting duties in the United States within a year of publication.

Sexual abuse of a child, unlike physical abuse, typically leaves no medical evidence and depends on disclosure, often delayed by years and complicated by the child's relationship with the offender, who is a family member or trusted adult in most documented cases rather than a stranger.

India's Protection of Children from Sexual Offences Act, 2012 created a dedicated, child-centred criminal framework here, including in-camera testimony and a duty under Section 19 requiring anyone with knowledge or apprehension of an offence to report it to the police or the Special Juvenile Police Unit, with a penalty for failing to do so.

Brazil takes the same reporting-duty approach for child maltreatment generally: Article 13 of its Child and Adolescent Statute, Lei 8.069 of 1990, requires suspected abuse to be reported to the local Conselho Tutelar, a specialised child-protection council rather than the police directly.

Emotional abuse, a sustained pattern of belittlement, threats or rejection that damages a child's development, and neglect, the failure to provide adequate food, supervision, medical care or education, are both harder to prove than physical or sexual abuse: they leave no single incident a court can point to, only a cumulative pattern documented over time, typically by teachers, health visitors or social workers who see the child repeatedly rather than one examining clinician.

Across all four categories, children raised in a household where domestic violence between adults is also occurring show measurably elevated rates of direct maltreatment and of long-term psychological harm even where they are not themselves physically struck, which is part of why family-violence categories are studied as a connected system rather than as unrelated offence types.