Maya Tamang shows me the corner of her kitchen where the wall still carries a crack shaped like a river. It runs from the ceiling to the floor, no wider than a finger. She traces it with her thumb. “This,” she says, “is how I know the government has never come.”
Maya is 54. She lives in a house rebuilt with corrugated metal, timber from her husband’s family land, and the money her youngest son sent from Qatar. The National Reconstruction Authority gave her a grant — enough for a foundation, not a roof. The second tranche never came. The third tranche was a rumour.
In the spring of 2015, two earthquakes — the first measuring 7.8 on the Richter scale, the second 7.3 — tore through Nepal’s mid-hills. Approximately 9,000 people died. Half a million homes were damaged or destroyed. The international community pledged $4.1 billion in aid. The Nepali government created the NRA to channel it. By every official metric, the reconstruction effort was a success story: hundreds of thousands of houses built, a model for disaster recovery in the developing world.
“The numbers are correct. The people are not behind the numbers.”— Maya Tamang, Sindhupalchok District
I spent eight months returning to the villages I first visited in 2015 as a 26-year-old reporter with a notebook and a borrowed satellite phone. I went back to Sindhupalchok — one of the worst-affected districts — and to Gorkha and Dolakha. I interviewed 47 families, three former NRA officials, a geotechnical engineer who worked on the reconstruction, two international aid workers, and a member of the parliamentary oversight committee.
The Architecture of Forgetting
The official narrative of Nepal’s earthquake recovery has three movements: crisis, response, resilience. The crisis was real. The response was massive, if chaotic. The resilience is contested.
What the official narrative elides is the texture of the failure — the ways in which bureaucratic design, political pressure, caste discrimination, and a media cycle that moved on all conspired to leave the most vulnerable survivors in the worst situations.
Consider the grant structure. The NRA divided reconstruction support into three tranches: an initial payment on registering the damage, a second on completing the foundation, a third on completing the building. The logic was anti-corruption: you only get money when we can verify you’ve used it correctly. The effect was devastating for the poorest households, who could not afford the upfront cost of construction and therefore could not trigger the next payment.
Rebuilt homes in Sindhupalchok — many funded privately, not by the NRA. Photo: Anshu Khanal, 2023.
“If you have money,” says Buddhi Ram Tamang, Maya’s neighbour, “you can build. Then the NRA pays you back. If you don’t have money, you can’t build. So you never get paid.” He arrived at this with a kind of mathematical clarity, as if the absurdity had long since passed through outrage and become simply fact.
What the Numbers Cannot Say
Nariswor Municipality, where Maya and Buddhi Ram live, is recorded as having achieved 94% housing reconstruction completion. On the morning I visit, I count six unfinished shells within five minutes’ walk of the municipal office. One is inhabited: a family of five sleeping in what will eventually be, if the money comes, their living room.
The gap between official data and lived reality is not simply a matter of dishonesty. It is a matter of methodology. The NRA counted houses for which grant money had been disbursed. It did not — because it could not — count houses that were inhabitable, houses that had been built to code, houses in which people felt safe.
“We were rebuilding for a report. Not for the people inside the buildings.”— Former NRA official, speaking anonymously
Caste and the Distribution of Disaster
The failure of the reconstruction was not distributed evenly. In every district I visited, the families who remained in the worst situations were Dalit: stone-cutters, cobblers, landless agricultural workers. Their claims were processed last, disputed most frequently, and — multiple sources allege — sometimes rejected on pretexts that would not have been applied to higher-caste neighbours.
I documented eleven such cases in Sindhupalchok alone. In seven of them, the reasons given for rejection — a missing signature, a disputed land boundary, an incorrect census number — were later resolved in favour of the claimant, but too late for the rebuilding season.
The Crack in the Wall
Before I leave Maya Tamang’s kitchen, I ask her what she would want people to know. She thinks for a long time. Her granddaughter is sleeping in the room beyond, on a mattress that takes up most of the floor.
“That we are still here,” she says finally. “That we are still in it. Tell them we are still inside.”
This piece is for Maya, and for the forty-six other families who gave me their time, their tea, and their trust.