Share & more
“I’m hiding” was what one woman told me when I asked what she does most days inside the world’s longest-running refugee complex.
She is Somali, in her thirties, a single mother. She has lived in Dadaab’s Dagahaley camp for 19 years: hiding from the men in the water queue; from the humanitarian rosters that fix her in place; from the night, when the walk to the latrine becomes a risk she calculates.
For many women, surviving Dadaab is about being invisible.
Dadaab is 35 years old this year. It was established in a remote corner of northeast Kenya to shelter Somalis fleeing civil war, and then drought. The Kenyan government – despite its hospitality – has always insisted it was temporary and steadfastly refuses to permit anything resembling permanence.
That means no durable housing; no street lighting; no locks for peoples front doors or land titles; no infrastructure that would concede the complex is a city – even though it is home to more than 400,000 people.
I interviewed 29 women across Dadaab’s three camps of Dagahaley, Ifo, and Hagadera in March 2025. I deliberately oversampled what Dadaab’s administration classifies as “minorities” – South Sudanese, Ethiopian, Burundian, Rwandan, Congolese – and one Kenyan Somali from the host community born inside the camp.
These nationalities make up roughly four percent of Dadaab's population. They are also the women most exposed to the camp's sexual and gender-based violence – and the ones its official picture renders least visible.
Urban scholarship has long celebrated what displaced people build: the markets, the schools, the kinship networks, the improvised infrastructures that bloom in the gaps of failed planning.
There is truth in that. But it obscures something else. In Dadaab, the women I worked with were not always building. Often, they were subtracting, withdrawing, erasing: Visibility itself had become dangerous.
Survival strategies
I documented four modalities – one specific to minority women, three shared by all.
The first is religious passing. Dadaab is overwhelmingly Muslim and non-Muslims are widely treated as unclean. Christian women in particular are assumed to be prostitutes. The hostility is not only male-generated. I had expected solidarity among women across religious lines and found very little.
The exclusion is also economic. Muslim traders and customers tend not to buy from non-Muslims, so a Christian woman who tries to make a living by trading finds no market for what she sells. Her livelihood narrows to almost nothing. So she passes. She adopts the dress, the speech, the public bearing of the Somali Muslim majority. She refuses her own language in front of strangers and teaches her children to do the same.
Religion and ethnicity – the canonical markers of refugee identity in international protection law – are precisely the categories these women have taught themselves to shed in order to get by.
The remaining three modalities are not solely about minority status. They are also about being a woman in a camp built without regard for women’s safety, where sexual and gender-based violence is extensively documented, and where a woman with a partner often faces violence inside the shelter as well as outside it. One major study found that close to 67% of women had faced intimate partner violence.
The second is temporal – what I call the withdrawn night. After dark, women’s geography contracts to the compound. The walk to the latrine, 50 metres away, becomes a calculation, because the path is unlit and the latrine has no lock.
The collapse of women’s mobility is hidden in plain sight because the camp’s official map – roads, markets, distribution points – is drawn for people who can move through it freely, and women cannot.
Several women told me they reduce evening food and water intake to avoid the trip. Others use a bucket inside the shelter and empty it before dawn. The hours between sunset and sunrise are not lived. They are endured behind doors that do not lock because the state forbids the permanence a lock would imply.
The third is delegated visibility. Women tend to send children – their own, their neighbours’ – to perform the public tasks that adult women cannot always safely do: fetching water, collecting rations, carrying messages. Children become the proxies through which women remain present in camp life while keeping their own bodies out of view.
The cost is borne by daughters who miss school and sons who learn early that women’s labour is meant to be unseen.
The fourth is spatial contraction. The compound becomes a prison-refuge. Several women in my sample had not left their block in years – the basic administrative unit of roughly 100-150 households organised around shared latrines and water points. One had not left her shelter, except for the latrine, in 14 months.
Men in Dadaab describe the camp as a prison too – but they are not the ones rationing their water to avoid a nighttime walk, or confined to a single room for over a year.
The collapse of women’s mobility is hidden in plain sight because the camp’s official map – roads, markets, distribution points – is drawn for people who can move through it freely, and women cannot.
The last three modalities fall on all women – but they fall harder on a minority woman, who is passing as well. She makes the same calculation about the unlit walk to the latrine, but with less to draw on: less community solidarity; less ability to buy from Muslim traders (even the household items that might make staying inside at night feel safer); and the knowledge that if she is assaulted, a Christian woman is assumed to have invited it.
The gendered threat and the religious one do not sit side by side. They compound.
This is not only the refugees’ doing. When I reported, still shocked, how the camp’s “minorities” were treated, a male UN refugee agency (UNHCR) officer in Dadaab answered: “What do you want? They are the majority.”
That is the logic of the camp, delivered by the institution built to protect them.
It is not incidental that he was a man. Dadaab’s residents are mostly women; its staff are mostly men – an imbalance UNHCR officers themselves name as a problem. A majority-male apparatus manages a majority-female population, and when the treatment of minority women is put to it plainly, it shrugs.
An architecture of erasure
What women in Dadaab call hiding is the rational response to an architecture that does not protect them. This is not the women’s failure to integrate. It is the camp’s success at producing exactly the population it was designed to produce: visible enough to count, invisible enough to ignore.
It’s what Cameroonian political theorist Achille Mbembe might call a necropolitics of social death – a place where the right to appear has been administratively withdrawn.
One woman in Dagahaley has lived in the camp for all the 35 years of its existence. She arrived as a child in 1991. She is in her late forties now. Her three children were born inside its fences.
When I asked what she wanted from the next 10 years, her answer was not for herself. “I lost hope for myself,” she said. “But at least my children should be able to escape the refugee reality.”
This is the bargain Dadaab teaches women to make. Not survival. Not return. The transfer of the future to someone younger, in the hope that someone, someday, might be permitted to leave.
In March 2025, the Kenyan government and UNHCR launched the Shirika Plan, a multi-year framework to convert Kenya’s two camps – Dadaab and Kakuma – into municipalities under the administration of their respective counties – Garissa and Turkana. It aims in theory to grant refugees the right to work, freedom of movement within the two counties, and integrated access to schools and health systems.
Protracted displacement is not solved by reclassifying camps as cities. It is solved by building infrastructure that makes invisibility unnecessary.
One year on, donor reviews call it a historic shift from encampment to inclusion. They are not wrong about that ambition. Shirika names housing, water, and energy among its pillars – but naming is not building. Its enormous projected cost rests on World Bank and private financing not yet secured, and the infrastructure it describes is pitched to make refugees bankable, not to make women safe.
None of its pillars name a lock, a light, or a latrine inside the compound. Integration that reclassifies the camp without redesigning it is paperwork – and the architecture of assault survives the change of name.
Protracted displacement is not solved by reclassifying camps as cities. It is solved by building infrastructure that makes invisibility unnecessary: lockable doors, lit pathways, latrines within compounds, water inside the household, police who record the assaults the system refuses to register, women on the committees that decide where the next pump, light, and shelter go.
These are not luxuries. They are the prerequisites for any integration that does not simply relocate the violence.
And the pattern is not unique to East Africa. In four internally displaced persons (IDP) camps on the edge of Abuja, Nigeria, where I have been conducting ongoing fieldwork, displaced women fetch water from contaminated puddles within sight of government-built houses meant to relocate earlier IDPs.
The invisibility there is not chosen the way it is in Dadaab – it is imposed, the quiet erasure of women the state has decided not to count. Different mechanism, same result: A woman’s world shrinks to what she can reach on foot before dark.
Read across these places, women’s suffering stops looking like a series of local tragedies and starts looking like what it is – a structural pattern, produced by infrastructure, wherever the people who design cities decide some lives need not be planned for.
Cities do not fail women by accident, they fail women by design. Building otherwise would mean beginning with where the design ends: with the door, with the latrine, with the light.
Fabienne Hoelzel’s chapter ‘Invisible by Necessity: Tactical Disappearance and the Embodied Politics of Survival in Dadaab’ is forthcoming in ‘Improvised Urbanism’, ed. Gihan Karunaratne (Routledge, 2026).
