Sawnwood and pulp trade data: the global wood-flow ledger, delivered as rows

It is the record of what the world's forests became, measured at the border and at the mill gate, kept annually since 1961. Sawnwood sits at the first transformation - roundwood squared into lumber - and wood pulp at the second-to-last stop before paper and paperboard, so the two series bracket the industrial heart of the sector. Around them the same grid carries roundwood, panels, charcoal, chips, pellets and further-processed goods: 104 product items, five measures deep (production, import quantity, export quantity, import value, export value), each observation one country-by-product-by-measure-by-year cell.

Two Datadory records describe this system. FAO Forest Product Statistics is the program-level record - the production-and-trade domains plus the pulp and paper capacity survey with its short-term forecasts. FAOSTAT Forestry - Forest Production and Trade is the production-and-trade cube itself, about 2.46 million rows spanning reference years 1961 through 2024. Collection runs through an annual questionnaire to official country correspondents, shared with Eurostat, UNECE and ITTO under a joint framework - which is why a European figure reconciles with its international counterpart instead of sitting beside it at a discrepancy.

Both records score 9/10 on the catalog rubric against a cross-catalog mean of 7.81 - among the 534 of 1,744 datasets that earn a nine - and they anchor the best forest-products datasets ranking.

They share one spine and differ in what rides along.

The Forest Product Statistics record is the wider instrument: thirteen verified fields per row, the capacity-survey series with forecasts alongside production and trade, and lookup tables resolving every area, item, element and flag code. Start there when the question touches planned pulp or paper capacity ahead of actuals, or when you want one entry that covers the whole program.

The Forestry - Forest Production and Trade record is the decoded grid: ten documented fields, fully keyed, with the M49 area code carried as a quoted string so joins onto customs, trade and macro tables behave without a reconciliation pass. Start there when the deliverable is a model input rather than a survey of the sector - apparent-consumption ratios, supply factors for a quant panel, country league tables.

The scored head-to-head against ground-truth inventory lives on our FAO Forest Product Statistics vs USDA Forest Service FIA DataMart comparison, and the European close-up from the same questionnaire continues on FOR_BASIC vs FAOSTAT Forestry. Either way the sample comes back cut to the products, economies and years you name.

One shape, repeated 2.46 million times - four verified observations exactly as they deliver:

Area     : Finland
Item     : Sawnwood, coniferous
Element  : Production
Year     : 1961    Unit : m3         Value : 7996000   Flag : A

Area     : United States of America
Item     : Wood, pulp and paper products (export/import)
Element  : Import value
Year     : 2000    Unit : 1000 USD   Value : 44754144  Flag : E

Area     : Albania
Item     : Wood pellets
Element  : Production
Year     : 2013    Unit : t          Value : 4900      Flag : X

Area     : Bhutan
Item     : Sawnwood, non-coniferous
Element  : Import value
Year     : 2022    Unit : 1000 USD   Value : 2         Flag : X

Read the four as a tour of what the ledger can hold. Finland 1961 is the depth claim made concrete: nearly eight million cubic metres of officially reported coniferous sawnwork output while most commercial databases were decades from existing. The United States row is the money half of the cube - the same keys that carry cubic metres also carry thousand-dollar trade values, so consumption math never leaves the table - and its flag E says the cell is an estimate, not a reported return. Albania marks the modern bioenergy economy arriving in a small market, flagged X because an external organization supplied the figure. And Bhutan is the honesty test: a two-thousand-dollar import is still a row, not rounding noise.

Every field above carries a verified definition with an example in the delivered dictionary. Request a sample cut to your products and economies and the dictionary travels with the rows.

Temporal: sixty-four consecutive annual vintages, 1961 through 2024. Few economic series run unbroken through that span; fewer still keep identical keys throughout, which is what makes a 1961 figure and a 2024 figure diffable without remapping anything. For sawnwood the window holds several complete housing cycles; for pulp it holds the entire postwar buildout of the fiber economy; for pellets it holds the whole category's life.

Geographic: 285 area codes - roughly 245 countries and territories beside about 40 regional aggregates including continents, subregions, the EU and the world total. Aggregates ship next to their members, which is convenient until someone sums the file naively and counts Europe twice, once as its members and once as Europe. Drop aggregate codes before any country ranking; the area lookup marking them ships with every delivery.

Granularity: one annual observation per area, product and measure - no subnational split anywhere in the panel. County-level physical detail for the United States belongs to the plot record, not this one.

Two thin spots are printed rather than buried. Estimated and imputed cells concentrate in older vintages and several aggregate series, so decide flag tolerance before modeling. And units mix by design - cubic metres, tonnes and thousand US dollars in one table, set by the measure - so the guardrail is the typed unit column rather than vigilance.

Three derivations turn the grid into analysis, and all three are reshapes Datadory can pre-cut into your extract.

Apparent consumption. Production plus imports minus exports, per country, product and year, approximates what an economy actually absorbed regardless of how weak its own reporting is. It is the standard first move in fiber-security work, and the five-measure design means it needs no second source - see the workflow sketched on our market sizing page.

Unit-value proxies. Export value divided by export quantity gives a realized price direction for any partner's shipments; tracked across years it separates volume-driven growth from price-driven growth in a way quantity-only feeds cannot.

Chain reads. Because roundwood, sawnwood, pulp, paper and pellets sit in one item classification, conversion-chain questions - did pulp capacity pull roundwood, did sawmill residues feed pulp - become group-bys across item codes rather than a stitching exercise across publishers. The vocabulary lives in our primers on sawnwood, wood pulp and roundwood.

Filter discipline matters more than cleverness here: select on item and element codes, then verify labels match the product family intended - the 104-item list includes export/import aggregates and discontinued series that will happily contaminate a sum.

Nothing else in the catalog matches its combination of width and depth; the comparators trade one axis for another.

Eurostat FOR_BASIC is the European close-up from the same joint questionnaire: 144,153 observations across 32 European geographies, 1988 through 2024, with species groupings and extra-EU splits the global grid flattens away.

Luke in Finland reaches back to 1949 across eight price regions - seventy-five years of national texture below the country-year grain.

Statistics Canada goes monthly where FAO is annual, with shipments and stocks by species group since January 2014 - May 2026 printed 3,869.8 thousand cubic metres of Canadian production against 4,969.1 in January 2014.

ITTO's Global Timber Index measures enterprise sentiment against the 50% line across ten tropical markets - forward-looking where this ledger is definitive. FIA's plot record measures standing trees, not flows. Route by the job: structural baselines want this grid, monthly confirmation wants Ottawa, sentiment wants the index, biology wants the plots.

Ranked by how directly a sixty-four-year flow panel answers the day job:

1. Timberland investors and REIT analysts. Country-level sawnwood and pulp flows frame the harvest and processing demand behind timberland valuations - the demand-side input no filing reports directly. Workflows: investors and quants x forest products.
2. Fiber-security and procurement planners. Apparent-consumption trends flag regions where demand outruns domestic supply before spot markets do; sizing patterns sit on demand forecasting.
3. Trade and market-entry teams. Quantities beside values read both tonnage and realized price direction of a partner's shipments - the difference between entering a growing market and a cheapening one.
4. Quant researchers. Sixty-four annual years of stable-coded panels make usable supply-side factors for timber, panel and paper exposures; the pattern is written up on quant backtesting.
5. Data scientists and ML engineers. A fully keyed, schema-stable panel whose codes survive releases intact trains on cleanly; workflows live on data scientists x forest products.
6. Academics, journalists and policy researchers. An intergovernmental compiler with per-row estimation flags survives peer review and comment letters; the citation pattern is on citation-grade research.

Start with the forest products data guide - the pillar mapping all twenty pooled datasets in the slice, from statistical ledgers to satellite layers - and browse the forest-products data hub for coverage chips and quality scores across the full record set. Product pages carry the detail summarized here: FAO Forest Product Statistics and FAOSTAT Forestry - Forest Production and Trade, each with captured sample rows and the verified field dictionary. The land-change counterpoint is covered on tree cover loss data by country.

Every record here ships from Datadory as typed rows with dictionaries attached, codes resolved to labels and provenance flags kept on the row - delivered daily, weekly, or hourly, your call. Name the products, economies, measures and years you need and get a sample; real rows come back first.