Femicide: What It Is, Why It Happens, and What Prevention Actually Looks Like

Femicide is the intentional killing of women and girls because they are women. The distinction matters: it is not a synonym for any homicide that happens to have a female victim. It names a category of killing in which gender is the reason, the motive or the enabling condition — and naming it that way changes what a country counts, what it investigates, and what it tries to prevent.

The term was brought into modern use by the sociologist Diana Russell in the 1970s, and it has since been adopted by the United Nations, the World Health Organization and a growing number of national legal systems. Several Latin American countries have written femicide into criminal law as a distinct offence. Most countries have not, and a large number do not disaggregate homicide statistics by the victim's relationship to the perpetrator at all — which means the scale of the problem is systematically undercounted in exactly the places least equipped to address it.

What separates femicide from homicide

Three patterns distinguish femicide from homicide generally, and all three are visible in the data wherever the data exists.

The first is who kills. Men who are murdered are most often killed by strangers or acquaintances, frequently in public, frequently in the context of other crime. Women who are murdered are most often killed by someone they know intimately — a current or former partner, a father, a brother, a son. The home, statistically, is the most dangerous place for a woman, and that single fact inverts almost every assumption built into conventional public safety policy.

The second is motive. Femicide is frequently preceded by a woman exercising autonomy: leaving a relationship, refusing a marriage, reporting abuse, seeking a divorce, earning her own income, or declining a sexual advance. The killing is an assertion of control at the point where control is being lost. This is why the period immediately after a woman leaves an abusive partner is consistently identified as the most dangerous, and why advice to "just leave" is more complicated than it sounds.

The third is escalation. Femicide is rarely a first act. It is usually the end of a documented sequence — threats, strangulation, stalking, property destruction, threats to children, escalating frequency and severity. Non-fatal strangulation in particular is one of the strongest known predictors of a subsequent killing, and it is routinely treated as a minor assault by systems that do not know this.

The main forms femicide takes

• Intimate partner femicide — killing by a current or former husband, partner or boyfriend. This is the single largest category worldwide.
• Family-related femicide, including so-called "honour" killings carried out by relatives over a woman's perceived transgression of family or community norms.
• Dowry-related femicide, concentrated in South Asia, where a woman is killed over dowry demands.
• Femicide connected to sexual violence, where killing follows or accompanies rape.
• Femicide of women in public life — journalists, human rights defenders, activists and politicians targeted for their work.
• Femicide connected to armed conflict and displacement, where the collapse of protective systems increases both the frequency and the impunity.

What the global data shows

United Nations estimates consistently find that around 137 women are killed by a member of their own family every day worldwide, and that roughly six in ten female homicide victims are killed by an intimate partner or relative. For male victims the equivalent proportion is a small fraction of that. These figures are best understood as a floor rather than a measure: they rest on the countries that record the relationship between victim and perpetrator, and many do not.

Africa, and the specific problem of measurement

Across much of sub-Saharan Africa the picture is obscured less by the absence of violence than by the absence of infrastructure to record it. Civil registration is incomplete. Post-mortem examination is rare outside major urban centres. Police recording practices vary enormously between regions. The result is that femicide in the region is discussed largely through survey data on non-lethal violence and through case reporting by journalists and civil society organisations, rather than through national statistics.

What survey data does establish is the size of the pool from which femicide emerges. Across the continent, the proportion of women reporting lifetime experience of physical or sexual violence from a partner is among the highest measured anywhere. In Cameroon, where our own programmes operate, Demographic and Health Survey data indicates that close to one in two women has experienced physical or sexual violence. Femicide is the visible tip of that distribution; the distribution itself is what determines how large the tip will be.

Why prevention is realistic

The escalation that precedes most femicides is the reason prevention is achievable. Lethal violence against women is not random, and it is not usually unprecedented. In case review after case review, the warning signs were present and often known — to neighbours, to relatives, to health workers, to police. What was missing was not information but a system configured to act on it.

Four things consistently reduce risk, and all four are within reach of local organisations rather than requiring national reform first.

1. Risk assessment at first contact. Structured questions about strangulation, threats to kill, weapon access, stalking and recent separation identify high-risk cases far better than professional intuition does.
2. A functioning referral pathway. A woman who discloses needs health care, a safety plan, legal advice and somewhere to go — and she needs them to connect to each other rather than sending her back to the start each time.
3. Safe, confidential reporting. Where disclosure carries social cost, confidentiality is not a nicety; it determines whether anyone discloses at all.
4. Working with men and boys. Prevention that addresses only women manages risk after it exists. Norms work with men, boys, traditional authorities and faith leaders addresses where the risk comes from.

What this means for our work

Women's Light Foundation does not operate a femicide observatory, and we would not claim the epidemiological reach to produce national mortality estimates. What we do is work at the stage before — safe spaces, survivor-centred case management, referral pathways between health facilities, legal aid providers and social services, and community norms work with men, boys and traditional authorities in the Northwest and Southwest Regions of Cameroon.

That work is shaped by the pattern described above. Because separation is the period of greatest danger, safety planning matters more than encouragement to leave. Because the home is where the risk concentrates, a referral pathway that ends at a woman's front door has not finished its job. Because escalation is visible, the people most likely to see it first are neighbours and relatives — which makes community-level awareness a protective measure rather than an awareness-raising exercise.

And because economic dependence is one of the most reliable predictors of whether a woman can leave at all, violence prevention disconnected from economic livelihoods work is prevention with a hole in it. A woman with no independent income and children to feed is making a different calculation from the one most safety advice assumes.

The reason the word matters

There is a reasonable objection to the term: murder is murder, and a separate word for the killing of women risks fragmenting a category that the law already covers. The reply is practical rather than ideological. What is not named is not counted. What is not counted does not appear in budgets, in police training curricula, in health service protocols or in national plans. Countries that began recording femicide as a distinct category discovered patterns that aggregate homicide statistics had concealed for decades — most importantly, how many of the deaths had been preceded by a report that no one acted on.

Naming it is the first step to measuring it. Measuring it is the first step to preventing it. That sequence is unglamorous and it is the whole argument.

Frequently asked questions

What is femicide?

Femicide is the intentional killing of women and girls because of their gender. It is distinguished from homicide generally by motive and context: the victim is killed because she is a woman, most often by an intimate partner or family member, and frequently after asserting some form of independence such as leaving a relationship or reporting abuse.

What is the difference between femicide and homicide?

Homicide describes any killing of one person by another. Femicide describes a gendered subset in which the victim's gender is the reason or the enabling condition. The practical difference shows in the data: men are most often killed by strangers in public, while women are most often killed by someone they know, in the home.

What are the main types of femicide?

The largest category is intimate partner femicide. Others include family-related killings such as so-called "honour" killings, dowry-related femicide, killings connected to sexual violence, the targeting of women in public life such as journalists and activists, and femicide occurring in armed conflict or displacement.

How many women are victims of femicide each year?

United Nations estimates indicate around 137 women are killed by a family member every day worldwide, with roughly six in ten female homicide victims killed by an intimate partner or relative. These figures should be read as a minimum, because many countries do not record the relationship between victim and perpetrator, and some do not recognise femicide as a distinct category at all.

Is femicide preventable?

Largely, yes. Most femicides are preceded by an escalating and often documented pattern of abuse — threats, stalking, non-fatal strangulation, threats to children. Because those warning signs are usually visible before the killing, structured risk assessment, working referral pathways, confidential reporting and prevention work with men and boys all measurably reduce risk.

Why is the period after leaving an abusive partner considered dangerous?

Because femicide is frequently an assertion of control at the point control is being lost. Separation is one of the most consistently identified risk factors in case reviews worldwide, which is why safety planning around a departure matters more than simply advising a woman to leave.