When the State Shows Up Late: Rural Pakistan After Disaster
This piece was developed through a collaborative writing process with AI assistance, based on my own ideas, research direction, and arguments.
In the summer of 2022, a third of Pakistan disappeared under water. Roughly 33 million people were affected. Nearly 8 million were displaced from their homes. Over 2.3 million houses were damaged or destroyed. More than 1,700 people died. And in the months and years since, the most damning detail isn’t any single number it’s how much of it was foreseeable, and how little was done before the water arrived.
This is the same sequencing problem this series has traced through education and economic policy, but here it stops being about opportunity or income. Here, it’s about who lives.
The Warning That Never Reached Anyone
Pakistan’s meteorological and climate authorities were aware, well in advance, that the 2022 monsoon season carried extraordinary risk. Some regions went on to receive over four times their normal seasonal rainfall. Researchers who reviewed the disaster afterward pointed to the same conclusion, plainly: adequate surveillance of water levels during monsoon season, paired with timely warnings, could have made mass evacuation possible before disaster struck and would have significantly reduced the death toll. The knowledge existed. The infrastructure to act on it, especially in rural areas, did not.
This isn’t a one-time failure from 2022. In August 2025, flash floods tore through Buner district in Khyber Pakhtunkhwa, killing hundreds. Villagers said afterward that no warning had ever been broadcast through the mosque loudspeakers the traditional method used to alert remote communities in an emergency. A provincial official’s response to the deaths was that people should not have built their homes along waterways in the first place. Three years after 2022, the same absence no reliable way to warn the people most at risk produced the same outcome, and the response to it was to place the responsibility on the villagers rather than on the missing warning system.
This is what weak structure actually looks like on the ground: not an abstract policy gap, but a village with no siren, no broadcast, no signal that the water was coming, until it already had.
When the Plan That Should Have Existed, Didn’t
Disaster response is supposed to run on pre-existing plans designated safe zones, evacuation routes, emergency shelters, stockpiled supplies, a clear chain of command. When a crisis of this scale hits without that structure already in place, the response becomes reactive instead of executed: officials making decisions in real time about where to send help, rather than activating a plan built and rehearsed long before the rain started. That distinction is the difference between an emergency response and an emergency improvisation, and rural Pakistan experienced the second one.
Get Noor-Ul-huda’s stories in your inbox
Join Medium for free to get updates from this writer.
Houses were swept away with people still inside them. Roads and bridges collapsed under the water, cutting off entire villages from any help that might have been coming. Over 1,600 kilometers of roads were damaged in Khyber Pakhtunkhwa alone within a single week of the disaster. Communication networks went down with the infrastructure, meaning some of the worst-hit communities couldn’t even report what was happening to them, let alone request help.
The Ten Percent Who Were Reached First
Rescue, when it came, did not come evenly. Balochistan and Sindh the most rural, least accessible provinces were also the hardest for relief teams to physically reach, which meant floodwater sat stagnant there far longer than anywhere else. By the time help began arriving in some areas, it had already been days. People survived those first days the only way they could: by themselves, on rooftops, on higher ground, without food, in cold and wet conditions, waiting.
When rescue teams and camps were finally established, they were not built at the scale the disaster demanded. Relief camps held a few hundred thousand people out of millions displaced, a fraction of those who needed shelter, food, and medical care. For everyone outside that fraction, it was often NGOs and ordinary citizens organizing their own relief efforts collecting food, water, and survival kits informally who reached people the formal response never did.
Then came the second disaster, layered on top of the first. Stagnant water in areas with no drainage infrastructure became breeding ground for disease. Malaria, cholera, and dengue spread through communities already stripped of shelter and food. Contaminated water caused waterborne illness in people with no access to clean alternatives. The same poor infrastructure that slowed the rescue also ensured the water had nowhere to go afterward so the crisis didn’t end when the rain stopped. It just changed shape.
What Was Lost Doesn’t End With the Flood
Nearly 1.7 million hectares of crops were destroyed. Over 800,000 livestock died. More than 30,000 schools and 2,000 health facilities were damaged, cutting off both a child’s education and a community’s access to care in the same breath. Farmland left waterlogged for weeks couldn’t simply dry out and return to use it needed to recover before it could be planted again, which meant no income and no food from land that used to provide both. Businesses that depended on those roads, those crops, those customers, simply stopped functioning. The floods are estimated to have pushed more than 8 million additional people into poverty, on top of everyone who was already there.
Where the Aid Actually Went
International pledges followed quickly, as they usually do this is the part of the sequence that photographs well. But a large share of displaced households, according to later assessments, received assistance late, and funding gaps repeatedly delayed recovery. By 2023, many people who had been displaced returned to their original communities, only to find the same absence of infrastructure waiting for them: damaged roads never rebuilt, schools never repaired, no early warning system installed where none had existed before.
This is the same money-trail question this series raised in Part 2, asked again in a harsher context: aid pledged in the name of a devastated population should be reaching that population, and should be leaving behind something sturdier than what existed before better drainage, real evacuation infrastructure, an actual warning system reaching every village, not just the ones near a city. The Buner flash floods in 2025 are the clearest evidence available that, in the places it mattered most, very little of that materialized. Three years after headlines called the 2022 floods “a monsoon on steroids,” a village still had no way to know the water was coming until it arrived.
The Same Sentence, a Third Time
Education, fuel prices, disaster response three completely different sectors, and the same sentence keeps fitting all of them: the response gets built for the audience watching from outside, and delivered, late and thin, to the people actually living through it. In a classroom, that costs a child their future. In an economy, it costs a family their next meal. Here, it costs people their lives, and it keeps costing them, disaster after disaster, because the plan that should exist before the water rises still doesn’t.
That repetition is the real subject of this series. Not one failure, but a pattern that survives every single time it’s exposed which is exactly what the final part has to reckon with: not just naming the pattern again, but asking what it would actually take to break it.
This is Part 3 of a series on policy sequencing in Pakistan. Part 4 brings the pattern together across all three pieces, and asks what it would take for governments to listen first, before the plan is already written.
