As monsoon rains close Bandarban's tourist trails and bury families from Chittagong to the Rohingya camps of Cox's Bazar, the real question is this: Why do the same hills fail every year?
Bandarban shut its tourist spots this week, and the closure barely registered as news. Heavy rain, landslide risk, a precaution, the kind of notice that disappears from the feed by evening.
Somewhere in the hills above Cox's Bazar, at least eight people, five of them children, were being pulled from mud and debris after hillsides collapsed onto Rohingya refugee shelters in the dark.
In Chittagong, a road-construction landslide killed a worker; children have died before him in nearby colonies, and will likely die again before the season ends.
None of this is new. It has simply been raining, and once again, the hills have moved.
The instinctive question is meteorological: Why is it raining so much? It's the wrong question, or at least an incomplete one. A more honest question is why our landscapes and our institutions have become so unable to withstand rain that has always fallen on these hills, in these months, since long before anyone kept records.
Heavy rainfall is not new to south-eastern Bangladesh. What is comparatively new is the number of people living where a hillside can no longer hold, and the number of hillsides that can no longer hold at all.
Disaster researchers have argued for decades that a flood or a landslide becomes catastrophic not simply because water falls from the sky, but because societies have already, quietly, arranged themselves in ways that make certain people unable to absorb the shock.
A hazard is a natural event. A disaster is what happens when that hazard meets a population that has been placed, by poverty, by policy, by the absence of any better option, in its path.
The rain over Chittagong this week is a hazard. The fact that more than a thousand families still live on or around the district's 25 identified high-risk hills is not.
Consider the maths of repetition. On a single day in June 2007, landslides killed 127 people across Chittagong. In the two decades since, at least 250 more have died in the same hills, according to environmental groups tracking the toll.
A committee formed after the 2007 disaster identified 28 causes and proposed 36 recommendations for reducing the risk. Nearly 20 years on, environmentalists say almost none of them have been meaningfully implemented.
This is not an unlucky region. It is a landscape that has been told, in writing, what would keep it safer, and has largely gone on as before.
Why would a society do this?
Not through malice, and rarely through ignorance. Families living beneath unstable slopes are usually well aware of the danger; many describe moving to safer ground during heavy rain and returning once it passes, because the land they occupy is the only land available to them at a price, or an absence of price, they can survive.
The choice to live on a hillside is not evidence of poor judgement. It is evidence of a housing market, a labour market, and a land-tenure system that have already made the decision for them, long before the monsoon arrives.
This is the quieter, slower-moving disaster that precedes the visible one: Forest cleared for construction and commercial cultivation, hills cut for building material and settlement, tourism infrastructure expanded onto slopes without regard for what held them together.
Each of these is a small, individually defensible economic act. Together, over enough years, they are how a hillside stops being able to hold itself, let alone a family sleeping on it.
The situation in the Cox's Bazar camps sharpens this argument rather than complicating it.
The Rohingya families killed in this week's landslides did not choose their hillsides in any meaningful sense; they were placed there, onto some of the most deforested and densely settled ground in the country, by a set of decisions about where displaced people without citizenship are permitted to exist.
Early-warning systems and evacuation drills, and Bangladesh does have functioning ones, developed with volunteers, meteorological forecasting, and community networks, can tell a family that a hill is about to fail.
They cannot tell that family where safer land might be, because in many cases, no such land has been made available to them at all. Warning without an alternative is not protection. It is simply advance notice of the same outcome.
None of this is unique. It is the same logic that governs why flood-exposed neighbourhoods are so often the poorest, why industrial risk clusters near communities with the least power to refuse it, and why the aftermath of a disaster so rarely disturbs the arrangements that produced it.
Reconstruction has a way of rebuilding the world exactly as it was, only more so. What changes, briefly, is that the arrangement becomes visible, the hillside, the camp, the illegal water and electricity connections that let unauthorized settlements persist on land everyone in authority already knew was dangerous.
None of this argues for indifference to environmental sustainability or for endless tree-planting campaigns that treat forests as ornamental rather than structural.
It argues for something closer to home: Watershed and slope governance that outlasts a single monsoon news cycle, an actual enforcement mechanism for hill-cutting regulations that currently exist mostly on paper, resettlement that offers viable land and livelihood rather than eviction notices, and a role for the people who have lived on these slopes the longest in deciding how they are managed.
Disaster governance that begins after the rain has already fallen has already failed.
Every monsoon, southeastern Bangladesh produces the same forecasts, the same closures, the same small number of names in the news.
What it reveals, if we let it, is not a freak of weather but a mirror, one that shows precisely how a society has chosen, and continues to choose, who gets to live on solid ground.
Nur Nishat Anjum is a researcher at the Department of Anthropology of the University of Dhaka.
