A new study from the University of the Western Cape (UWC) has found that the question “why doesn’t she just leave?” misunderstands the lived reality of women trapped in abusive relationships.

Researchers say leaving is often not a simple choice but a calculation between competing dangers, where safety itself may hinge on staying.

The research, published in the Social Sciences journal, was conducted by UWC’s Centre for Interdisciplinary Studies of Children, Family and Society. It draws on the experiences of 262 mothers, grandmothers, adolescents and practitioners from four South African communities: Cape Town, Saldanha Bay, Fisantekraal and a university student cohort.

Participants from diverse backgrounds described strikingly similar experiences, pointing to patterns that cut across race, class and geography.

The team, led by Prof Nicolette Roman and co-authored by Dr Chante Johannes and Shenaaz Wareley, developed a framework they call the “distorted process of care” – a way of understanding how relationships that should offer love, protection and support can instead become instruments of control, danger and entrapment.

When love becomes a trap

The first of five interconnected themes identified by the researchers is what they describe as “distorted care”. In many of the relationships studied, the pattern began with genuine affection before gradually shifting towards control and abuse.

Emotional attachment, the researchers found, often blinded women to early warning signs. Jealousy and controlling behaviour were misread as proof of devotion rather than recognised as danger signals.

As rigid gender roles, intergenerational violence, and alcohol or drug abuse took hold, the caring foundation of a relationship was progressively eroded.

“Care under constraint” is the second theme. For many women, staying was not a sign of passivity but a strategy for survival. A significant number depended on their abusive partner for basic necessities – food, shelter, school fees.

For mothers especially, the calculus was agonising: leaving could mean losing the family’s only income, forcing a choice between escaping violence and keeping children fed and housed. The wellbeing of children was consistently prioritised over a woman’s own safety.

Fear, shame and the erosion of hope

The third theme – “silence of care” – speaks to the isolation that abuse breeds. Fear of judgement, deep shame, and repeated disappointments from the very institutions meant to help left many women withdrawn and silent. Over time, a significant number stopped seeking help altogether, having lost faith that support would arrive.

Some women, however, did eventually reach a point the researchers call “reclaiming care”. In these cases, women chose their own safety, drew strength from faith or community, and actively worked to prevent their children from inheriting the cycle of violence. The researchers stress this is not a linear progression – it is hard-won, non-linear, and deeply personal.

A system that fails those it should protect

The fifth and final theme – “structural failure” – places responsibility squarely on the families, communities and institutions that surround women in crisis. Police, social services and government departments, the study found, frequently retraumatised women rather than protecting them.

Crime, gangsterism and poverty in already-marginalised communities further narrowed the options available to women seeking help. Families themselves were often too financially stretched to offer refuge.

The danger of leaving was not abstract for many participants. Women who attempted to separate from their partners reported being followed, assaulted or fearing for their lives. In this context, the researchers argue, the decision to stay was often rational – a calculated assessment of which option carried less immediate risk.

Traditional gender roles reinforced the cycle. Women bore the bulk of childcare and household responsibilities while abusive partners remained largely absent from family life. Children growing up in violent homes were found to be at greater risk of the cycle continuing into the next generation.

Alcohol and substance abuse by partners, the study noted, intensified both the frequency and severity of violence and the climate of fear.

Implications for policy

The researchers are unambiguous about what their findings mean for how gender-based violence is addressed in South Africa, where GBV remains a national crisis affecting women across every community and class. They argue that current approaches, which tend to focus on individual women and crisis responses, are structurally limited.

“For policy, the implications of this framework are uncomfortable but necessary,” the authors state. “GBV cannot be addressed through individualised, crisis-response models alone. Interventions that focus solely on the individual, such as counselling, shelters and legal protection, without addressing the material and institutional conditions that make remaining in abusive situations appear rational, are likely to remain insufficient.”

The researchers are calling for a fundamental shift in how gender-based violence is understood and tackled in South Africa. Rather than placing the burden on individual survivors to leave, the focus must shift to creating the conditions that make safety a real possibility for everyone.

ALSO READ: From victim to survivor: Woman’s fight against domestic violence in South Africa