Harmonised Nepal Customs Trade Records, FY 2071/72–2082/83 (2014–2026)
Description
This dataset is a harmonised, machine-readable release of Nepal’s merchandise trade statistics. It is compiled from the twelve annual Foreign Trade Statistics workbooks published by the Department of Customs, Government of Nepal, for fiscal years 2071/72 to 2082/83 in the Bikram Sambat calendar (mid-July 2014 to mid-July 2026). The main file, nepal_fts_records.csv.gz, contains 427,333 records at the eight-digit national tariff line. They are given by flow (import or export), by commodity and by commodity × partner country. Each record has the reporting unit, quantity, customs value (NPR ‘000) and, where published, import revenue, and is traced to its source workbook and sheet. Four derived layers are added: (1) a concordance linking HS 2012 and HS 2017 codes to an HS 2022 basis, using UNSD correlation tables, with link type and weights (hs_concordance_hs2022.csv); (2) ISO 3166-1 alpha-3 partner codes (partner_lookup.csv); (3) standardised units (unit_lookup.csv); (4) seven row-level quality flags (F1–F7) that mark suspect records without changing any published value. Commodity totals for FY 2071/72, which has no commodity table in the source, are provided separately (nepal_fts_commodity_2071_72_derived.csv). Source workbooks are identified by SHA-256 checksums in source_manifest.csv. Validation: all 43 annual totals agree with the Department of Customs’ published figures to within 0.07%. Partner-level and commodity-level tables agree for all 60,174 tariff lines published in both. The Python scripts in /code reproduce the release from the source workbooks, and /validation holds the reconciliation and coverage outputs. README.md gives the data dictionary, flag definitions and reproduction steps. Important for use: partner and commodity rows are stored in one file and each set sums to the annual total, so filter on the “level” column before summing. The data can be reused to estimate electrical and electronic equipment placed on the market, to model e-waste generation with the UNU-KEY framework, to study battery and photovoltaic trade, and to monitor transboundary waste movements.
Files
Steps to reproduce
Download the 12 Annual Foreign Trade Statistics workbooks (FY 2071/72–2082/83) from customs.gov.np and check them against source_manifest.csv. Obtain the UNSD HS2022–HS2017 and HS2022–HS2012 correlation workbooks (data-raw folder of the ‘concordance’ R package, github.com/insongkim/concordance). Then run: python extract.py <workbook_dir> <stage_dir>; python harmonise_v2.py <stage_dir> <unsd_tables_dir> <release_dir>; python validate.py <stage_dir> <release_dir>. Tested with Python 3, pandas 3.0.5, openpyxl and pycountry 26.2.16.