Comparing resettlement budgets in Nepal

Nepal's disclosed resettlement plans provide practical examples of how land acquisition, affected populations and restoration commitments shape a budget. This comparison extends the Upper Trishuli-1 case through six further project plans. It asks which costs can be reproduced, which obligations can be matched, and what additional evidence is needed to assess sufficiency.

All figures are historical nominal Nepalese rupees. Source editions date from 2013 to 2020; their totals describe planned provisions. Each project remains a separate observation, including the two Nagdhunga projects. Parallel document hosts and later editions remain linked source records.

Select the comparison around the loss

Project and source edition Disclosed budget, NPR million Recorded population What the comparison contributes
Upper Trishuli-1, final LALRP, December 2018 177.28 154 affected families; 149 in the livelihood cohort Additional compensation and livelihood commitments, alongside earlier land acquisition
Kanchanpur–Kamala, Board-approval RP, June 2018 81.98 479 affected households Road right-of-way impacts, including partial structures, kiosks and trees; this edition requires no private-land acquisition
Tanahu Hydropower, updated RIPP, November 2018 2,189.16 547 affected households A same-year hydropower estimate, with an explicit comparison to the 2012 plan
Kabeli-A, SAP, July 2013 29.58 13 households losing agricultural land Land classes and valuation rates; wider social measures extend beyond the land-loss cohort
Mugling–Pokhara Phase 1, draft RP, July 2019 1,494.76 810 affected households Land and structure quantities for Abukhaireni–Pokhara and the Madi and Seti bridges
Nagdhunga–Naubise–Mugling, RAP, April 2020 295.78 78 affected households Private losses, public assets and assistance under World Bank operation P170409
Nagdhunga Tunnel, final RAP, November 2014 2,556.77 190 estimated affected households A land-dominated estimate preceding detailed-design confirmation

Exact amounts, source links and page locators appear in the evidence file and bibliography below. The populations retain their source definitions. A programme allocation includes shared delivery costs and collective measures as well as individual benefits.

The three 2018 plans illustrate the limits of matching by country and year. Dividing each total by its recorded population gives approximately NPR 0.171 million for Kanchanpur–Kamala, NPR 1.151 million for Upper Trishuli-1, and NPR 4.002 million for Tanahu. Private-land acquisition, loss severity, local markets and accounting scope differ. These observations support selection of component comparators; a national compensation benchmark would require a consistently defined dataset.

Kabeli-A's SAP budgets NPR 11,076,641 for 7.678 hectares of private land, with rates by land class. It describes no physical population displacement and places specified related costs in biological mitigation and environmental management budgets. Its full SAP total covers wider social measures, so the 13-household land-loss cohort is unsuitable as a denominator for that total. Tables 14-1 and 14-4, printed pp. 116–119; PDF pp. 144–147.

Tanahu's disclosed total falls from NPR 2,553,603,572 in 2012 to NPR 2,189,158,961 in 2018, a nominal decline of 14.3%. Removing the earlier transmission and rural-electrification provisions reduces the calculated decline to 3.6%, consistent with the plan's rounded account of approximately 4%. Further interpretation needs a component and footprint crosswalk before examining quantities or prices. Budget discussion and Annex 9, printed pp. 87–90 and 163; PDF pp. 99–102 and 175.

Reconcile the source tables

Upper Trishuli-1's detailed budget reconciles to its summary and rounded contingency. The other plans contain useful examples of why an audit must also examine quantities, rates and transfers between tables.

Source-table check Reproduced difference, NPR Evidence and follow-up
Kanchanpur–Kamala: implementation detail versus summary 3,220,000 Table 29's listed amounts sum to 5,920,000; its subtotal and Table 30 carry 2,700,000. Obtain the governing implementation worksheet.
Mugling–Pokhara: Madi Bridge land row 387,500 Table 40 gives 241.18 m² × NPR 15,500/m² = 3,738,290, against a printed amount of 4,125,790. Confirm which entry governs.
Mugling–Pokhara: district land subtotals versus grand total 372,000 District subtotals sum to 269,560,333.50; the carried grand total is 269,188,333.50. Reconcile the same land component.
Nagdhunga–Naubise–Mugling: assistance detail versus summary 120,000 Table 8-7 gives 57 and 45 entitlement-category cases at NPR 40,000, totalling 4,080,000; Table 8-13 carries 4,200,000. Category counts may overlap.
Nagdhunga Tunnel: implementation detail versus summary 1,327,500 Table 9.5 sums to 1,475,000; Table 9.7 carries 147,500. Confirm the intended staffing and activity provision.

Sources: Kanchanpur–Kamala Tables 29–30, printed pp. 32–33 / PDF pp. 43–44; Mugling–Pokhara Table 40, printed p. 43 / PDF p. 55; Nagdhunga–Naubise–Mugling Tables 8-7 and 8-13, printed pp. 62 and 65 / PDF pp. 63 and 66; Nagdhunga Tunnel Tables 9.5 and 9.7, printed pp. 50–51 / PDF pp. 57–58.

The two Mugling land checks concern the same component. Their differences should be investigated together before any adjustment. The assistance categories in the Nagdhunga road plan likewise need a recipient crosswalk before their counts can describe unique households.

Kanchanpur–Kamala also has a contingency inconsistency: the prose states 10%, while the summary uses 15%. The disclosed total reproduces the latter rate to two decimal places; its printed contingency is NPR 0.49 lower. If the detailed implementation provision governed and 15% were retained, the conditional total would be NPR 85,680,056.10. For the tunnel, carrying the detailed implementation provision with the existing 5% contingency would give NPR 2,558,162,104. These are review scenarios; the evidence file preserves the disclosed amounts.

Match components and financing boundaries

Provision Comparison unit Evidence needed
Land NPR per m² or hectare, by productive class Valuation date, tenure, access, productive potential, area and transaction costs
Structures NPR per m² or ft², by specification Materials, quality, floor area, labor, transport and partial-loss treatment
Transition support NPR per eligible case per month Interruption period, eligibility, household needs and extension conditions
Livelihood measures Defined intervention and recipient cohort Inputs, market feasibility, unique participation and follow-up outcomes
Implementation and monitoring Staffing period or specified activity Workload, access, verification, reporting and responsibility

The two highway plans quote frame-structure rates of NPR 2,340/ft² for Mugling–Pokhara and NPR 2,520/ft² for Nagdhunga–Naubise–Mugling, a difference of 7.7%. Their descriptions differ: the former specifies frame structures with concrete; the latter specifies an RCC frame and RCC roof. Engineering specifications and price periods need checking before treating them as matched rates. Frame structures also account for different shares of the recorded private floor area: 50.7% and 69.1%, respectively. Mugling–Pokhara Table 41, PDF p. 56; Nagdhunga–Naubise–Mugling Table 8-3, PDF p. 62.

The Nagdhunga road plan includes NPR 63,892,458 for public structures and utilities, 23.8% of its base budget. Section 8.4 also identifies civil-works provisions. Link both references to the obligation and funding source before consolidation. This establishes whether the civil-works budget delivers the RAP commitment or finances additional work.

The tunnel's land provision accounts for 99.1% of its base budget, while its socioeconomic survey covers 64 of the 190 estimated affected households, or 33.7%. Its cost ledger and household-coverage record answer separate review questions. Confirm the affected register and losses through detailed design, then reconcile eligibility, quantities and restoration measures. Tunnel RAP, printed pp. 8 and 48–51; PDF pp. 18 and 55–58.

The Upper Trishuli-1 case develops this connection between survey coverage, transition duration, replacement-land productivity and delivery. The Nepal baseline discussion and spatial baseline workflow explain how household records and resource-use evidence support those tests. Compare prices across years only after selecting a documented price basis and an index suited to the component.

Reproduce the checks

The cells below use the published evidence file. Calculations run in this notebook; source amounts remain fixed to the cited editions. Kabeli-A is omitted from the programme-allocation table because its recorded land-loss cohort has narrower scope than the SAP budget.

display(Inputs.table(reconciliationChecks(evidence.projects), {
  columns: ["project", "check", "differenceNpr"], header: {differenceNpr: "Difference, NPR"},
  format: {differenceNpr: value => value.toLocaleString("en-US")}, rows: 5
}));

Sources

Project-plan dates identify the reviewed editions. PDF locators count from the first page. Review each operation against its governing policy and financing agreements; these historical ADB plans cite the 2009 Safeguard Policy Statement.

Signed-in readers with the relevant archive permission can consult retained source editions below.