In the early hours of July 6 this year, while most of Ukhiya in Cox’s Bazar of Bangladesh slept, the hillsides of Camps 7, 11 and 15 gave way. Eight people died beneath the mud—among them a seven-year-old boy, a married couple with their four-year-old son, and a woman buried alongside three children aged thirteen, five and three. Two days later, another slope collapsed onto a madrasa during classes, killing at least five children while they were attending lessons. The Bangladesh Meteorological Department, with the weary precision of an institution that has issued this forecast many times before, predicted more rain.
The temptation, in reporting such events, is to reach for the vocabulary of natural disaster: the monsoon “struck”, the hillside “gave way”, nature “claimed” its victims. But political ecology teaches us to distrust that grammar. Slopes do not simply give way. They give way because forests were felled, because settlements were carved into laterite hills never meant to bear them, because a million people were compressed into twenty-six square kilometres and then told, for nine years, that their situation was temporary. The monsoon is meteorology. The landslide is policy.
Original Sin of 2017
I was working in northern Rakhine with a French organisation when the violence erupted in Maungdaw, the violence that, in the last months of 2017, drove more than seven hundred thousand Rohingya across the Naf. There was no time for planning, and no one seriously argues there could have been. When I moved to Cox’s Bazar with the same organisation to support the influx response, I felt what everyone felt: shelter first. Bulldozers and machetes cleared thousands of hectares of forest, much of it within the Teknaf and Ukhiya reserves, in a matter of weeks. Elephants lost their corridors; the soil lost its anchoring; and the hills of Cox’s Bazar, stripped bare and terraced by hand, became the world’s largest refugee settlement almost overnight.
That improvisation was forgivable. What has followed is less so. The emergency logic of 2017—cut, flatten, shelter—has never been superseded by any other logic. Nine monsoons later, the camps remain what they were in the first winter: bamboo frames and tarpaulin skins pinned to slopes that a first-year geologist could identify as unstable. Recent risk assessments have found that roughly 15 per cent of the settlement area lies within zones of high or very high landslide risk. This is not a secret buried in academic journals. It is known, mapped, published and lived with. Into the ninth year, things are getting even worse.
To understand where Kutupalong is heading, one needs only walk through Nayapara. The registered camps at Nayapara and Kutupalong date back not to 2017 but to the exodus of 1991–92. Children born there in the Clinton years are now raising their own children within the same fences. Three decades of “temporary” arrangements have produced a third generation of statelessness and, tellingly, no relocation plan worth the name. Bhasan Char, the silt island in the Bay of Bengal, absorbed a fraction of the population amid controversy and remains what it always was: a relocation of the problem, not a solution to it.
The deeper pathology is that temporariness has become a political instrument. Permanent structures are prohibited in the camps as a matter of policy—concrete implies acceptance, acceptance implies integration, and integration is politically unspeakable in Dhaka as long as repatriation remains the official horizon. Yet repatriation to Rakhine, where fighting has intensified and fresh arrivals now gather at the border, is a horizon that recedes with every passing year. The result is a settlement engineered to be fragile. The bamboo that snaps under mudflow and the tarpaulin that feeds the flames are not accidents of poverty; they are the material expression of a political refusal to admit that these camps will exist next year, and the year after, and the decade after that.
Fire and Water, in Rotation
The monsoon is only half the calendar of catastrophe. Come the dry season, the same tarpaulin and bamboo that dissolve under rain become tinder. The great fire of March 2021 killed at least fifteen people and left tens of thousands shelterless in a single afternoon; smaller fires have recurred almost annually since, some accidental, some suspected to be acts of arson linked to the armed groups that flourish wherever the state’s presence thins. A camp that floods in July and burns in February is not experiencing bad luck twice a year. It is experiencing the same vulnerability through two different elements.
And here is where disaster risk reduction, as practised in Cox’s Bazar, has calcified into ritual. Slope stabilisation with sandbags and vetiver grass, early-warning loudspeakers, evacuation drills, and the annual pre-monsoon relocation of a few thousand households from “high-risk” plots to marginally less lethal ones—all of it is necessary, none of it sufficient, and all of it is repeated each year as though repetition were progress. DRR has become the management of an unacceptable baseline rather than an effort to change it. When the same agencies issue the same warnings before the same season and count the same categories of the dead afterwards, we are no longer reducing risk. We are administering it.
The Kivu Premonition
Anyone who has worked in the eastern Democratic Republic of Congo will recognise the trajectory. The camps of North and South Kivu were also born of emergency, also declared temporary, also starved of durable infrastructure lest permanence be conceded, and they are still there, thirty years on, their populations churned by each new cycle of violence, their temporariness fossilised into a way of life. The Kivus teach an uncomfortable lesson: camps do not wither because we refuse to invest in them. They simply persist, badly. The refusal to build for the long term does not shorten displacement; it only ensures that the long term, when it inevitably arrives, is lived in squalor.
Cox’s Bazar is following that script with unsettling fidelity. The settlement is made to last forever—its boundaries, its dependency, its legal limbo while nothing within it does: not the shelters, not the drainage, and not the futures of the children who died at their lessons on July 8.
The rains will stop in October. The fires will start soon after. And unless the political question, belonging, rights, and a future somewhere, is finally addressed as the root cause it is, the hills of Ukhiya will keep doing what stripped, overloaded, abandoned hills do. We will call it a natural disaster. It will be nothing of the sort.
