Oecd Stan Industry Accounts Data: Dataset Structure and Field Coverage

STAN stands for the OECD's STAN Database for Structural Analysis, the organisation's harmonised set of industry accounts covering member and accession countries. On Datadory it arrives inside the record OECD.Stat — Retail, Industry & Trade Indicators, a quality-score-8 catalog entry whose SDMX REST API lists roughly 1,546 published dataflows. The STAN dataflow itself carries the identifier DSD_STAN@DF_STAN_2025 — the 2025 edition — and resolves along six dimensions: FREQ, REF_AREA, ACTIVITY, MEASURE, PRICE_BASE and UNIT_MEASURE.

The grain is annual, country by economic activity. REF_AREA takes ISO 3-letter codes (DEU, GBR, AUT), and ACTIVITY draws on the CL_ACTIVITY_ISIC4 codelist of ISIC Rev.4 activities, so the same query skeleton that returns German wearing-apparel rows under C14 returns any other manufacturing division you point it at. For semiconductor work that matters because the searcher's goal here is cross-country output, value added and employment benchmarks for electronics manufacturing — a job STAN handles with one consistent measure vocabulary across every reporting economy, rather than the patched-together national publications you would otherwise reconcile by hand.

Eight workhorse measures from the CL_STAN_MEASURE codelist do most of the analytical work in STAN, and the 2025 edition publishes all eight for every activity and reference area in the dataflow. The table lines them up with the job each one feeds in an electronics-manufacturing benchmark:

Four mechanical traps account for most bad pulls. UNIT_MULTIPLIER is applied to OBS_VALUE (6 means millions), so unscaled charts mislead immediately. OBS_STATUS E marks estimates, which deserve different treatment than final prints in any revision-sensitive model. Mixing XDC national-currency series across countries without converting to the USD unit option produces sums that mean nothing. And classification boundaries differ: STAN's ISIC Rev.4 activities, Comtrade's HS codes and Census NAICS codes do not map one-to-one, so a join across the three needs a concordance step rather than a naive merge on code strings.

Geographic coverage follows the OECD's own perimeter: member countries plus accession partners and aggregates, keyed through REF_AREA. That makes STAN the natural place to line up the US, Japan, Germany, Korea and Taiwan-class manufacturing economies on one measure definition — though for economy-level export shares beyond the membership, the World Bank indicator's 200-plus-economy span is wider.