Nepal Budget Execution: Actual vs Allocated Capital Spending (FCGO Tracker)

A national budget has two numbers that rarely match: the amount allocated (विनियोजन) when the budget is tabled in Jestha, and the amount actually spent by the close of the fiscal year on 31 Asar. The gap between them is the heart of budget execution. Nepal's Financial Comptroller General Office (FCGO / महालेखा नियन्त्रक कार्यालय), the government's central accounting and treasury authority, publishes the year-end actuals in its Consolidated Financial Statements and ongoing budgetary reports, so allocation can be measured against real spending.

Nepal's chronic problem is not on the recurrent (current / चालु) side. Salaries, pensions, grants and interest are paid almost in full, recurrent execution typically runs in the high-80s to mid-90s percent of target. The failure is concentrated in the capital (पुँजीगत) budget, the development outlays meant to build roads, hospitals, irrigation and power. Over the past decade, actual capital spending averaged just 64.1 percent of the allocated capital budget, even as capital was only about 19 percent of total federal spending and recurrent dominated at roughly 66 percent.

The result is a widening gap between budget headlines and reality. The World Bank notes that over the past decade Nepal's budgets averaged 33.7 percent of GDP in allocation but only 26.8 percent of GDP in actual spending, a gap of about 6.9 percentage points. Money is appropriated for development that the state then proves unable to spend.

The capital-expenditure absorption rate is simply actual capital spending divided by the capital allocation for that year. It is the single clearest measure of execution, and in Nepal it is persistently low. At the federal level, only around 62 percent of capital allocations were executed on average in recent years, and the share of capital in the total budget has been shrinking, from 27.1 percent in FY 2020/21 (FY21) to 20.9 percent in FY 2023/24 (FY24).

The decline is also visible as a share of the economy. Federal capital expenditure fell from about 5.3 percent of GDP to 3.4 percent of GDP between FY21 and FY24, and total government capital spending across all three tiers of government dropped from 11.4 percent of GDP to 7.8 percent. The World Bank estimates Nepal needs sustained public investment of 10 to 15 percent of GDP to close its infrastructure gap, so the country is investing roughly half of what it requires.

Two warning signs recur. First, capital spending now lags debt servicing: the state increasingly spends more on repaying and servicing loans than on building new assets. Second, low absorption is not a revenue problem, the money is allocated and largely available, it is the spending machinery that fails.

The clearest fully closed year on record is FY 2081/82 (2024/25). Against a total budget of Rs 1,860.3 billion, FCGO reported total actual expenditure of about Rs 1,508 billion, roughly 81.9 percent of the annual target. The capital budget told the familiar story: of the Rs 352.35 billion allocated, only about Rs 222.68 billion was actually spent, an absorption rate of 63.2 percent. Recurrent spending reached Rs 980.38 billion (85.95 percent of target) and financing/financial-management spending Rs 320.04 billion (87.87 percent). Revenue collection was about Rs 1,178 billion, 83.06 percent of the target.

The pattern repeats across years. In FY 2080/81 (2023/24), against a total budget of Rs 1,751.31 billion, total expenditure was about Rs 1,408 billion and revenue about Rs 1,030 billion; capital spending that year was about 16 percent lower than the following year's Rs 222.68 billion, again leaving the development budget far from fully used. For the years still in progress, FY 2082/83 (2025/26) and FY 2083/84 (2026/27), only the allocations (Rs 407.89 billion and Rs 431.1 billion of capital respectively) are durable facts; final FCGO actuals are not yet published and should not be assumed.

Read alongside the budget pages on this site, the message is consistent: allocations keep rising (from Rs 1,751 billion in FY 2080/81 to a record Rs 2,124 billion in FY 2083/84), but actual spending, especially capital, trails far behind. Budgets grew about 12.3 percent a year before the pandemic and only about 4.1 percent a year in the five years after, yet even the slower headline growth has outpaced the state's capacity to deliver.

Because so little is spent early, agencies scramble to exhaust funds before the fiscal year closes on 31 Asar (mid-July). The World Bank finds that over 40 percent of capital spending occurs in the last month of the fiscal year. Independent reporting on the closing weeks captures the surge in real time: capital spending that averaged only about Rs 7.49 billion per month over the first 11 months can jump to several times that in Asar alone.

The macro footprint is striking. In Asar 2082 BS alone, the government spent about Rs 233 billion, equal to 15.4 percent of the entire fiscal year's expenditure, in a single month. This eleventh-hour dash is so entrenched it has its own name in Nepali: 'asare bikas' (असारे विकास), literally 'development done in the month of Asar'.

The rush is not just an accounting quirk, it degrades quality. Asar coincides with the onset of the monsoon, so roads laid, drains dug and structures rushed to completion in the rain are often poorly built and, as commentators note, can be 'simply washed away'. Front-loading work into the wettest, most compressed window is a recipe for waste, cost overruns and assets that do not last.

The bottleneck is structural, not a one-off. On the budgeting and planning side, too many projects are approved with poor preparation, outdated cost data and weak monitoring; Nepal's 'project bank' exists largely on paper, without clear selection criteria. Projects then stall on readiness: tree-cutting clearances average about two years, and land acquisition two to three years, slowed by fragmented records, outdated valuation methods and compensation disputes.

Cash management compounds the problem. Quarterly cash rationing releases funds in tranches that push spending toward year-end. Procurement is slow and risk-averse, World Bank-financed procurement in Nepal took about 231 days, the longest in South Asia against a regional average of 192, driven by lowest-bid mandates, strict penalties and paper-based processes. Inflexible mid-year reallocation rules, frequent staff turnover and low donor-fund disbursement round out the picture. At the current pace and funding, the World Bank estimates Nepal's National Pride Projects would take some 41 years to finish.

The consequences ripple outward: unfinished infrastructure, recurring 'asare bikas' waste, a public capital stock that has fallen as a share of GDP, and budgets whose ambition is undercut every year by the inability to spend.

The World Bank's 2025 capital-expenditure analysis and its public expenditure review converge on a handful of fixes. Prioritise only investment-ready projects rather than approving a long wish-list; raise project readiness through digital land records, standardised valuations and streamlined environmental and utility clearances; and make cash management credible so fund releases are smooth rather than back-loaded into Asar.

Two further reforms target the machinery directly: simplify and delegate mid-year budget-reallocation rules so unspent funds can move to projects that can absorb them, and overhaul procurement, clearer evaluation, screening of abnormally low bids, digital processes and stronger agency capacity. Nepal's own FY 2083/84 budget echoes these themes with promises of 'Mission Mode' project delivery, a development-projects sunset law and a hybrid-annuity model pipeline; whether those pledges lift the absorption rate is exactly what FCGO year-end actuals will reveal.

The benchmark to watch is simple. If a future year's capital absorption rate climbs durably above the ~64 percent decade average, and if the Asar share of capital spending falls below 40 percent, Nepal will finally be closing the gap between what it budgets and what it builds.