Nepal was the first country in the world to introduce climate budget tagging, tracking which portions of government spending address climate risk, in 2012. More than eighteen countries have since adopted the practice. In its FY 2024/25 review, Nepal’s climate budget recorded the lowest spending utilisation of all seventeen Sustainable Development Goal categories, the gap between a genuine policy innovation and what it produces on the ground.
Nepal’s per-person greenhouse gas footprint is 1.6 metric tonnes, against a global average of 6.76. The country caused almost none of the warming now affecting its mountains and plains, and that injustice is real. But it obscures something closer examination reveals: Nepal has built its economic architecture on the same resource climate change is making unreliable, an architecture that has so far worked, generating revenue and household income while accumulating structural risk. The hydropower, the migration, the remittances substituting for a rural economy in decline: each appears, at its own scale, to be working. The question is what they collectively produce.
This piece examines four nodes of that system: the glaciers whose depletion Nepal’s legal framework was not designed to track; the hydropower strategy that subordinates upstream communities to export revenue; the migration and remittance flows that relieve pressure on highland investment; and a policy apparatus that measures climate risk more precisely than it manages it.
The glacier that appears on no balance sheet
The Hindu Kush Himalayan range holds 63,700 glaciers across 55,782 square kilometres, the largest cryosphere outside the poles. Nepal-specific surveys show close to a quarter of glacier area lost since the late 1970s, the rate doubling since 2000. Two ICIMOD assessments published in 2026 found up to twenty-seven metres of ice thickness lost since 1975, and snow persistence recently hit its lowest level in twenty-three years.
What those measurements describe is more precisely understood as depletion than as disaster. A glacier does not simply melt. It depreciates. The ice Nepal is losing is stored natural capital accumulated over millennia, the water-storage function that releases meltwater slowly through the dry season, regulating the rivers the highland economy and the energy strategy depend on. When that capital is drawn down, it generates no replacement capital elsewhere in the national accounts. The write-down appears nowhere until the systems it underwrote have failed, arriving as lost agricultural productivity, diminished hydropower reliability, or a sudden glacial lake outburst.
This invisibility is partly legal. The Water Resources Act 1992, the primary law governing Nepal’s water, predates the climate era by three decades and makes no reference to glacial retreat. It treats water as a fixed stock, written for a watershed that had not yet begun to retreat. Tajikistan, with comparable mountain geography, enacted a dedicated Law on the Protection of Glaciers in January 2024. Nepal, home to the largest mountain cryosphere outside the poles, has not. The ice enters Nepal’s legal architecture only as a disaster-relief expenditure, a post-failure accounting rather than an early-warning system.
The wager and the statute behind it
Nepal’s primary policy response to its resource endowment has been to monetise it. The roadmap targets 28,500 megawatts of installed capacity, with most of the conditional NDC commitment aimed at export, requiring tens of billions over a decade. Nepal became a net electricity exporter for the first time in 2024. The logic is simple: Nepal has the rivers; the region has the demand. This monetisation reveals two disconnects, one institutional, one physical.
The first disconnect is institutional. There is nothing wrong, in principle, with giving hydropower its own statute, regulator, and long licensing terms; most countries do this, since power generation needs investment certainty a general resource law rarely provides. Nepal’s Water Resources Act 1992 ranks water uses explicitly, drinking water first, hydroelectricity fourth, then carves hydropower out of that hierarchy entirely under a dedicated Electricity Act with its own regulator. What is harder to defend is what happened to the sectors ranked above it: no equivalent statute, no comparable regulator, no comparable financing. Agriculture still receives one to two percent of national energy consumption, and in the wet season of 2025 surplus generation exceeded India’s import ceiling and was spilled rather than redirected to the irrigation the law ranks first. The asymmetry is not that Nepal wrote a contradictory law. It is that having named four priorities, it built durable institutions for only the fourth.
The second disconnect is physical. The rivers being monetised are the same rivers being destabilised by the glacial retreat that determines their flow. Floods have repeatedly damaged the hydropower fleet, from thirty-four projects hit between 2015 and 2020 to thirty-two more in October 2025 alone. The single largest loss came in July 2016, when a glacial lake and landslide-dam failure swept away the Upper Bhotekoshi plant, with damage near four hundred and seven million dollars. A 2014 Asian Development Bank assessment projected annual climate-related losses reaching 2.2 percent of GDP by 2050; extreme years already approach that threshold. Nepal is capitalising on a resource simultaneously generating revenue and accumulating physical risk, and neither the royalty structure nor the water law prices the second into the first.
The migration economy and the political signal it sends
Most hill farmland is unirrigated and dependent on rainfall and snowmelt timing; shifts in water availability translate into crop failure, income loss, and a decision to leave. In the Karnali basin, thirty-eight percent of mountain farmland and twenty-one percent of hill farmland lies abandoned; thirty-two hill and mountain districts now record negative population growth. Structural poverty is consistently the primary driver, climate stress an accelerant of a threshold the economy had already set low. What leaves with a departing family is more than labour: mountain water has long been governed through local, community-enforced arrangements that depend on the people who maintain them, recognised in formal law only as licensing exemptions, not protected rights. When a village empties, that governance dissolves with it, and the unfarmed terrace erodes, raising sediment and risk in the river corridors feeding the hydropower infrastructure below.
Nepal’s remittances reached 28.2 percent of GDP in FY 2024/25, among the highest in the world, and careful research finds they measurably improve household food security. At the household scale, the strategy works, which is why this is a governance challenge rather than a simple failure: the positive signal relieves pressure on investing in highland agriculture. The Foreign Employment Act 2007 does not extend to migration toward India, the largest single destination for climate-relevant departure; under the 1950 Treaty of Peace and Friendship Nepalis move there without a labour permit, so a significant share of Nepal’s GDP moves through a channel the law cannot see.
The IMF has modelled the compound shock: when climate stress and remittance contraction arrive together, food imports surge, outmigration accelerates further, and household assets fall by more than three and a half percent in a single cycle. The Gulf states absorbing most of Nepal’s migrants have no stable foundation either, cycling through conflict, oil-price swings, and labour-nationalisation drives like Saudi Arabia’s Nitaqat quotas, each tightening demand for Nepali migrant work for reasons that have nothing to do with Nepal’s glaciers. Nepal has not diversified away from a fragile income source so much as moved it offshore, into a labour market less stable than the rivers it replaced.
The architecture that measures what it cannot govern
Nepal’s climate governance framework is substantive on paper: a National Climate Change Policy, a National Adaptation Plan to 2050, a third NDC submitted in 2025, and the climate budget tagging system it introduced before any other country. Three indicators reveal the gap between what these measures and what it governs.
The first is the utilisation rate already noted: Nepal tags its climate spending more precisely than most comparable economies yet converts less of it into expenditure than the category’s own importance implies, a product not of insufficient allocation but institutional discontinuity. Nepal’s political landscape has seen considerable instability since 1990, with more than thirty governments and no administration completing a full term; the consequence is that no highland water or agricultural policy has survived long enough to build an implementation record worth continuing.
The second is the financing behind Nepal’s own commitments. The National Adaptation Plan costs an estimated $47.4bn through 2050; Nepal’s own contribution is $1.5bn, the rest contingent on finance that Climate Action Tracker finds roughly ninety-six percent of Nepal’s NDC targets depend on, contingent rather than binding.
The third is the legal gap for displacement. The Disaster Risk Reduction and Management Act 2017, Nepal’s primary climate-crisis statute, was designed for sudden-onset events, not the decade-long agricultural decline that empties a village or the aquifer that falls out of reach over twenty years. Nepal’s own National Adaptation Plan acknowledges ghost villages as a real outcome, yet offers no response mechanism, and no bill defining a climate-displaced person has ever been tabled. The gap has been judicially identified: in December 2018, Nepal’s Supreme Court found the Environment Protection Act 1997 inadequate for the country’s climate commitments and ordered a dedicated Climate Change Act, with provisions for compensating those harmed by environmental degradation. That order is now eight years old. The Law Commission has prepared a draft. The current parliament, with a political durability no recent government has had, holds both the judicial mandate and the conditions to act on it.
The pattern across all three is the same: a governance architecture built for measurement and international reporting, at a level of sophistication that exceeds many wealthier economies, with the implementation layer, the statutes, the institutions to sustain policy across political cycles, left as the distance between commitment and condition.
What the moment now requires
Nepal’s claim on international climate finance is legitimate, but a set of institutional and legislative deficits within domestic reach sits between that claim and what Nepal can absorb, regardless of when external finance arrives. The March 2026 election, following the interim government’s tenure, is the first chance in a generation to test whether a parliament can complete a legislative cycle long enough to matter.
What connects every gap this piece has examined is not a shortage of diagnosis: a Supreme Court order has already named the need for dedicated climate legislation, a National Adaptation Plan has already named ghost villages as a real outcome, a water law has already named drinking water, not export revenue, as the country’s first priority. Each diagnosis has produced an instrument built to the scale of the political cycle that wrote it, not the scale of a challenge that compounds across decades. Average government tenure since 1990 has been little more than a year; a glacier does not retreat on that timeline, nor a village empty on it. This is a local version of what Mark Carney, then governor of the Bank of England, called the tragedy of the horizon: a crisis whose worst costs arrive beyond the time horizon of the institutions meant to prevent them. Carney was describing financial markets. Nepal’s version is harder to miss: a government measured in months, regulating a glacier on a fifty-year clock.
A genuine reorientation is less a checklist than a change in method: legislation built to survive the government that drafts it, slow-onset decline treated as its own legal category, and value extracted from a resource returned to the places it comes from. Tajikistan moved from no glacier law to a dedicated one inside a single legislative term, proof this is achievable for a country with no greater capacity than Nepal’s. What Nepal has lacked is not diagnosis, but the durability to carry one law from the page it was written on to the decade it was written for.
The mountains are not waiting. The terraces above the Karnali are not recovering on their own. The arithmetic of abandonment does not pause for an election cycle.
Nepal can keep refining the instrument that measures this problem, or it can build the one that would govern it. The difference will not show up in next year’s budget utilisation report. It will show up in whether the highland villages this piece has described still have anyone living in them.
