FAOSTAT Crops and Livestock Products (QCL)

FAOSTAT Crops and Livestock Products — domain QCL in the FAO's catalog — is the longest-running production ledger in world agriculture, and Datadory hands it over as typed, analysis-ready rows. Every record pins one measured element to one commodity in one country in one year: 312 commodity items, 244 countries and territories and roughly 4.2 million observations stacked across 64 reference years, 1961 through 2024. Wheat, maize and rice sit alongside grapes, almonds in shell, tobacco leaf and natural rubber; the livestock side carries meat, milk and eggs with their own element vocabulary. Twenty measured elements run through the domain, from the cropping trio of area harvested, production and yield to livestock-specific counts such as Milk Animals, Laying and Producing Animals/Slaughtered.

Why does a farm-machinery catalog lead with a crop-production dataset? Because equipment demand is derived demand — hectares and tonnes are what put tractors in fields — and this is the one series that measures those inputs identically in every country on earth. Commodities carry both FAO item codes and UN CPC codes, areas carry M49 codes, and every value ships with a provenance flag separating official figures from estimates. Get a sample of this dataset and we return rows shaped exactly like the dictionary below, cut to the countries and commodities you name.

Three real observations for Italian almonds in shell, exactly as the domain records them — one row per element, long format, flag attached:

Area    : Italy                 Item : Almonds, in shell      Year : 2022
Element : Area harvested        Unit : ha                     Value: 53890.000000
Flag    : A  (official figure)

Area    : Italy                 Item : Almonds, in shell      Year : 2022
Element : Yield                 Unit : kg/ha                  Value: 1384.100000
Flag    : A  (official figure)

Area    : Italy                 Item : Almonds, in shell      Year : 2022
Element : Production            Unit : t                      Value: 74590.000000
Flag    : A  (official figure)

Livestock commodities follow the identical shape with their own element vocabulary — a dairy herd row reads Milk Animals, counted in Head, where an orchard row reads Area harvested in ha:

Area    : <country>             Item : <livestock product>    Year : <reference year>
Element : Milk Animals | Producing Animals/Slaughtered | Laying | Stocks
Unit    : Head | 1000 Head      Value: <filled in your sample>
Flag    : A | E | I | X | M

Two things worth noticing before you model anything. The long format means one commodity-year is several rows — join on area, item and year to pivot elements side by side. And the flag travels with the value everywhere, so an official figure and an imputed estimate never have to be treated as equals in the same regression unless you choose to.

Fourteen columns carry every observation, and all fourteen map cleanly to the domain's published layout — nothing here is inferred. Three of them exist purely to make joins painless: the FAO item code paired with its UN CPC equivalent on the commodity side, and the M49 code alongside the FAO area code on the geography side. The table below is the working dictionary; request a sample and the same columns arrive populated for your selection.

Geography — 244 countries and territories worldwide. Not the biggest producers plus a long tail of gaps: every reporting economy carries the same commodity list and the same element vocabulary, which is precisely what makes a cross-country yield comparison a query rather than a reconciliation project.

Temporal — every year from 1961 through 2024, sixty-four reference years deep. Few agricultural series survive that long without a methodology break; this one keeps its codes stable across the whole span, so a 1961 observation and a 2024 observation land in the same table without remapping.

Granularity — country × commodity × element × year. There is no sub-national split and no farm-level detail here; when you need regional machinery counts inside Europe, Eurostat's farm structure survey picks up where this stops.

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

Name the countries, commodities and elements when you request the sample — the full 4.2-million-row span, or a slice cut to the supply chain you actually model. The sample lands first either way; the ongoing feed arrives on whatever cadence your models need, already typed and flagged rather than reformatted by hand.

A production ledger this wide earns its keep in five specific jobs:

• Equipment market sizing — harvested hectares by crop and country are the demand base under tractor and harvester sales. Pair the output series with the FAOSTAT Agricultural Machinery domain's fleet counts and the World Bank's tractor panel, and a mechanisation gap becomes arithmetic rather than anecdote.
• Cross-country benchmarking — 312 commodities measured one way across 244 areas is the citation-grade base for any "who grows what, how efficiently" study, with yields directly comparable between, say, Italian almonds and Spanish olives.
• Agri-food supply and price modeling — sixty-four years of production history is enough runway to test whether supply signals led prices, and enough depth to separate trend from weather noise.
• Downstream input demand estimation — tyres, agrochemicals and implements all ride on planted and harvested area; this series quantifies the base for each crop in each market.
• Adjacent-industry screens — grapes feed the distillers and vintners work, tobacco leaf and natural rubber feed adjacent category research, and the same rows serve all of them without a new extraction.

Market Researchers & Consultants get scope, geography and years available in one place — the three qualifying questions every client asks answer themselves from the coverage alone. Investors & Quant Researchers get a sixty-four-year panel with provenance flags intact, long enough to test supply stories against price history honestly. Data Scientists & ML Engineers get clean numeric keys — item, CPC, M49 — that join to other domains without an entity-resolution project in front of the modeling. Journalists, Academics & Students get the original measurement rather than someone else's chart of it. Start from the agricultural farm machinery data hub, then read vs Eurostat Agriculture & Farm Structure for where a global production ledger beats a European structural survey.

Provenance note — compiled by the Food and Agriculture Organization of the United Nations, whose FAOSTAT catalog spans production, trade, inputs and food balance domains sharing one code system. QCL is its flagship production domain, and the codes below reappear across its siblings.

Methodology note — every observation carries one of five provenance flags: A official figure, E estimated, I imputed by a receiving agency, X from an external organization, M missing because the data cannot exist. Filter on the flag before you average; a mean that mixes officials with imputations is a number nobody signed.

Completeness note — Datadory scores this record 10/10, one of the few perfect marks in a catalog of 1,744 datasets where the average sits at 7.81. Field definitions are fully verified against the domain's own documentation, and the sample rows above are drawn from the live record rather than reconstructed.

Where to go next — the FAOSTAT Agricultural Machinery domain pairs fleet stocks with this output ledger; Eurostat Agriculture & Farm Structure adds sub-national European detail; and the best agricultural farm machinery datasets ranking shows where the marketplace listings fit around the official statistics.