Meat & Seafood Preparations Trade Profiles Data (HS16)

The Observatory of Economic Complexity publishes a trade profile for every node of the Harmonized System, and chapter 16 is the one stocked by delicatessens, grocers and specialty food counters: preparations of meat, fish, crustaceans and molluscs. Five headings define it. HS 1601 covers sausages, HS 1602 other prepared or preserved meat, HS 1603 animal extracts, HS 1604 processed fish and HS 1605 processed crustaceans.

The figures underneath come from BACI, CEPII's harmonized revision of UN Comtrade, published at six-digit product depth by year and bilateral pair. In 2023 the chapter carried roughly $63.8 billion of world trade: other prepared meat led at $20.15 billion, processed fish reached $19.11 billion, processed crustaceans $17.08 billion, sausages $7.23 billion and animal extracts $232 million. Each of those totals decomposes into exporter-importer pairs rather than stopping at the world aggregate, which is what turns a market-size sentence into a sourcing decision. Get a sample of this dataset cut to your headings and partners.

Two cuts of the same record set, exactly as a sample arrives. First, one observation behind a single heading:

# one observation: HS 1601, calendar year 2023, world trade value
Year            : 2023
HS4 Official ID : 1601
HS4 Official    : Sausages
Trade Value     : 7232821955          # current US dollars

Second, the headline layer for the whole chapter in 2023:

YEAR  HS4   HS4_OFFICIAL             TRADE_VALUE_USD
2023  1601  Sausages                     7232821955
2023  1602  Other Prepared Meat         20148073019
2023  1603  Animal Extracts               232301523
2023  1604  Processed Fish              19114692525
2023  1605  Processed Crustaceans       17081888490

Read the anatomy rather than the digits. Year pins the observation, HS4 Official ID selects one of the five headings, and Trade Value states what moved in current dollars. The two partner columns - exporter and importer - are where the interesting work happens: the world total above is the sum of bilateral pairs, so the same query that returns $20.15 billion for other prepared meat returns each supplier-destination edge underneath it. Quantity arrives only where the underlying reporting supplies it, and Unit Value derives from the two rather than being asserted separately.

Nine fields define every row, all verified during the August 2026 review. Together they carry the full coordinate system - year, product heading at HS2 chapter and HS4 heading levels, both trading partners - plus the three measures: trade value, quantity where reported, and derived unit value. Definitions follow the classification's own terminology, so a join against any other Harmonized System-keyed dataset holds without a translation layer.

• Geography: worldwide bilateral flows between all reporting economies, with both partners available at continent and country levels - so the same extract answers "who supplies the world" and "who supplies one region" without a re-pull.
• Temporal: 1995 through 2024 on the HS92 BACI cube - thirty annual observations confirmed during the August 2026 review. Later Harmonized System revisions cover shorter windows, down to 2022-2024 on the newest revision, which is why long-run analysis rides the 1995 series while recent-revision questions get finer product detail.
• Granularity: annual, disaggregated to HS6 product depth and bilateral exporter-importer pairs, rolling up cleanly to the five HS4 headings and the chapter itself.

Set against the wider catalog - 1,744 datasets averaging 7.81 - this record scores 7/10, carried by the combination of bilateral resolution and a thirty-year continuous history: few shelves let you watch the prepared-protein trade reorganize itself decade by decade at supplier-destination precision.

API, files, or your warehouse. Daily, weekly, or hourly.

Pick the channel your team already works in: flat-file drops sized to the headings, partner economies and date ranges you actually named, or a direct pipe into Snowflake, BigQuery or Redshift. Extracts arrive in the nine-field scheme above unchanged, with quantity handling settled during the sample pass and unit-value derivation applied only if you ask for it. Cadence gets chosen after the sample validates, not before.

• Specialty retail and e-commerce category teams read supplier concentration before negotiating: the bilateral edges behind each heading show which exporting economies dominate sausages versus processed crustaceans, and how fast those shares move. Sourcing workflows continue on our supply-chain mapping page.
• Consultants and market-sizing teams cite heading-level dollar values directly - $20.15 billion of other prepared meat moving in 2023 beats an anonymous "multi-billion-dollar category"; fuller workflows sit on our market sizing page.
• Demand planners and forecasters feed thirty years of bilateral flows into models that need trade to lead domestic availability; patterns continue on demand forecasting.
• ML engineers train price and sourcing models on typed rows whose schema never drifts; feature work continues on ML model training.
• Journalists and academics quote dated, attributed dollar figures on the global prepared-protein trade instead of hedging with "industry estimates"; writing workflows continue on citation-grade research.

For contrast inside the same catalog: FAOSTAT Food and Agriculture Statistics covers production and food balance on the farm side, USDA FAS Global Agricultural Trade System drills into US agricultural export detail, and the FDA Recalls and Safety Alerts listings watch the enforcement side. None gives you every exporter-importer pair in chapter 16 at once - that lane belongs to this profile.

E-commerce and specialty retail operators get supplier-market visibility per heading before assortment and sourcing decisions - see e-commerce operators in other specialty retail. Market researchers and consultants get citable bilateral figures for category sizing - see market researchers in other specialty retail. Data scientists and ML engineers get thirty years of typed trade rows that load in one read - see data scientists in other specialty retail. Developers and data-product builders get a fixed nine-column schema that ships inside dashboards unchanged - see developers and builders in other specialty retail. Journalists, academics and students get the quotable record - see journalists and academics in other specialty retail. Teams closer to the packaged-foods aisle can widen the aperture on the packaged foods & meats data hub.