Baseline survey design in Nepal

District statistics establish the context for a project baseline. Household surveys, an affected-person census and field research establish conditions among the people exposed to a project. Connect these sources through explicit populations, dates, administrative boundaries and measurement definitions.

The baseline assessment method explains how to select evidence for appraisal and mitigation. This guide addresses its application in Nepal: aligning project measures with national statistics, reviewing district sources and designing comparable household follow-up. Use the questionnaire and fieldwork guide for instrument development and the Maluku discussion for a complementary coastal livelihood context.

The searchable historical district index provides leads to local profiles. The dataset curation guide describes how to verify each edition, connect it to official demographic and mapping resources, and record its suitability for a study.

Interpret the historical district index

The retained index contains 77 district records. The counts below describe its references and recorded gaps; availability and analytical suitability require source-by-source verification.

The Eastern and Western Rukum records share one older Rukum reference. The Nawalpur and Parasi records also share a reference. Retain those relationships in the source register. Reusing an older combined-area total as the total for each successor district would duplicate population or resources.

The index records gaps for Bara, Rautahat, Siraha, Kathmandu, Dolakha, Makwanpur, Lalitpur, Mustang, Gulmi, Pyuthan and Rupandehi. Preserve that historical finding alongside the results of subsequent searches for official profiles or other suitable data.

The map supplies a geographic finding aid. Its boundary edition is separate from the reference dates of the indexed documents. Use the searchable table below to open the recorded endpoints, and inspect each document's coverage before comparing or allocating its values.

Dates embedded in filenames and upload paths also require interpretation. A directory dated 2019 does not establish the year of every table in its document. A Bikram Sambat year should be retained as reported, with a documented calendar conversion where needed; a blanket subtraction of 57 years does not determine an exact Gregorian date.

Use current national resources alongside the index

Nepal's National Statistics Office microdata catalog provides documentation for the Nepal Living Standards Survey IV, collected from July 2022 to June 2023. The catalog describes household and individual units, a frame based on the 2021 population and housing census, and 15 survey reporting domains. Its modules cover welfare, facilities, livelihoods, migration, and remittances. Use the design, weights, questionnaire, and dictionary together; a domain estimate is not a ward or project-corridor estimate (NSO, 2022/23; references).

The catalog also retains earlier surveys, including the Annual Household Survey 2015/16. Compare their concepts, coverage and design before constructing a time series. NLSS IV has no panel sample: comparisons between rounds describe population estimates, subject to comparability and sampling uncertainty (NSO, 2015/16, 2022/23).

The working 2021 census ward-table catalogue provides a current route to official aggregate tables. The NSO administrative code table identifies 7 provinces, 77 districts, and 753 local levels. Keep codes as strings, preserve their edition, and use the complete district–local-level–ward key: a ward number is not unique across Nepal. Inspect each downloaded table's units, reference date, and geographic coverage before joining it to geometry.

Use the official catalogue for microdata requests and documentation. NSO requires written permission for redistribution or sale of NLSS microdata; record the applicable permission with any retained or shared edition. Public discussion can cite the catalogue and published aggregates (catalogue conditions).

Build a source register before combining tables

Record one source entity for each document or dataset and separate entries for individual tables. Preserve the URL, title, producer, publication date, data-collection period, geography and boundary vintage, language, format, page or table number, units, denominator, revisions, access and reuse conditions, retrieval status, and file hash when downloaded.

Use explicit statuses such as indexed_not_retrieved, retrieved, link_failed, restricted, not_in_index and not_applicable. Distinguish an inaccessible URL, a missing table and an estimate outside the survey's reporting domains. Preserve each as a documented status, separate from a measured zero.

For PDF tables, retain a page reference and verify extraction against the rendered table. Check multirow headings, units, Nepali numerals, footnotes, and subtotals. Preserve ambiguous cells for review. A machine-readable output is useful only if its values still mean what the source table meant.

Keep a document's identifier separate from the geographic units it describes. One older profile may cover two present-day districts, and one district may have several files or editions. Record the relationship rather than choosing a filename by name similarity alone. A retained archive copy and a newer official table may both be useful, but they answer different historical questions.

Construct an indicator crosswalk

Use the original questionnaire wording and metadata from the relevant national survey round. A proposed crosswalk can classify relationships as exact, compatible_after_conversion, partial, or not_comparable.

Complete the crosswalk against the official instrument and dictionary for the selected round. Record the variable code, wording, universe, recall period and response categories. Identify project-specific questions separately, including tenure and resource-access measures with no equivalent national variable. Adaptations require cognitive testing, translation review and a field pilot; alignment with a national topic alone establishes neither equivalence nor validation.

Design a project baseline around impact pathways

For a road or transmission study, begin with the route, permanent footprint, temporary work areas, access restrictions and construction schedule. Identify settlements and resource-use areas connected to those activities. The following design considerations apply to a proposed study; project-specific evidence is presented in the Upper Trishuli-1 case.

List affected owners, occupants, tenants, users, businesses, and relevant employees through the appropriate census and inventory process. Use a designed household sample for broader area conditions where suitable. Keep households selected through an affected-person census separate from those selected through area sampling so that their weights and interpretation do not become confused.

Assess vulnerability through the interaction of livelihood dependence, resources, access and barriers. Examine how caste, ethnicity, gender, age and disability shape those conditions in the study context. Collect sensitive information for a defined analytical purpose, with culturally appropriate questions and disclosure controls. Include within-household differences in labor burdens and control over resources.

Analyze displacement as more than asset loss

Link households to parcels, productive activities, access points, and services. A narrow easement can have consequences different from acquiring the same area permanently; the restrictions and timing determine the pathway. Agricultural production, tenancy, employment, resource collection, and seasonal movement may require separate measures.

For a land-based household, a physical exposure measure can be the area subject to a defined restriction divided by total area used in the corresponding period. Its denominator must include the intended set of owned, leased, borrowed, or customary-use land. The ratio describes exposure, not entitlement or a full estimate of livelihood loss. Analyze productivity, crop cycles, access, costs, and substitute opportunities alongside it.

For remittance-dependent households, distinguish members resident in the household from absentees and transfer senders. A smaller resident roster can raise measured per-capita consumption without an improvement in the resources supporting the wider family. Keep household composition and transfers visible when explaining changes.

Compare consistent geography and survey rounds

Maintain stable geographic IDs, aliases, and valid-from/valid-to dates. A name match is insufficient when boundaries have changed. Aggregate to a common geography where a reliable correspondence exists. If an older total cannot be allocated credibly between successor areas, keep it at the older geography and mark the finer comparison unavailable.

Area-weighted redistribution assumes the measured population or quantity is distributed in proportion to area. That may be implausible in mountainous terrain or where settlements occupy a small part of a district. Population-weighted redistribution introduces its own assumptions and model uncertainty. Report these as estimates derived through a crosswalk, not as original official observations.

For household follow-up, preserve unit identifiers, season, question definitions, price basis, and membership history. Analyze attrition and changes in livelihood activity as well as monetary change. Triangulate unexpected results through interviews, local prices, service records, and verified spatial evidence.

Search the historical district index

This is the original 77-record finding aid, with its historical names, gaps, and compound reference cells retained. Search in English or Nepali. Links below are original public endpoints; they have not all been revalidated. The current census catalogue above and the curation workflow provide routes for finding a replacement or recording a reviewed archived edition.

The downloadable historical index contains source links and descriptive metadata, with no household records. Its row IDs are finding-aid identifiers, not official statistical geographic codes.

Resettlement costs and outcomes

The Upper Trishuli-1 cost analysis shows how source editions, household definitions and payment periods determine the meaning of a cost estimate. It connects the historical project record to baseline design and verification of livelihood restoration.