USDA NASS Quick Stats

USDA NASS Quick Stats is the federal estimate bank for American agriculture, and read through a Packaged Foods & Meats lens it doubles as a raw-materials ledger: cattle on hand, hogs kept, broilers placed, turkeys raised, head slaughtered, pounds produced, dollars received per unit - every figure the National Agricultural Statistics Service publishes, organized on one What/Where/When model. Pick the subject, pick the place, pick the period, and the matching estimates come back as flat rows.

Scale first, because it reframes what a meat company can know about its own supply base. The corpus spans six subject areas - Livestock and Animals, Crops and Plants, Economic and Prices, Demographics, Environmental, Research and Science - and runs to tens of millions of records; the animals-and-products slice alone occupies hundreds of megabytes of text before anyone has asked it a single question. Get a sample of this dataset cut to your species, states and years.

One estimate per row, identifiers attached. Two rows exactly as they ship:

SOURCE_DESC      : CENSUS
SECTOR_DESC      : ANIMALS & PRODUCTS
GROUP_DESC       : SPECIALTY
COMMODITY_DESC   : SPECIALTY ANIMAL TOTALS
STATISTICCAT_DESC: SALES
UNIT_DESC        : $
AGG_LEVEL_DESC   : COUNTY
STATE_NAME       : NEW MEXICO
COUNTY_NAME      : SAN JUAN
YEAR             : 2007
FREQ_DESC        : ANNUAL
VALUE            : (D)

SOURCE_DESC      : SURVEY
SECTOR_DESC      : ANIMALS & PRODUCTS
GROUP_DESC       : DAIRY
COMMODITY_DESC   : MILK
STATISTICCAT_DESC: FAT TEST
UNIT_DESC        : PCT
AGG_LEVEL_DESC   : STATE
STATE_NAME       : OTHER STATES
YEAR             : 2022
FREQ_DESC        : ANNUAL
VALUE            : 3.83

Read together, they demonstrate the two behaviors any protein analyst needs to plan around. The census row is a county-level sales figure withheld behind (D) because a single operation would dominate it - the dataset marking where disclosure rules bite rather than inventing a number. The survey row is a decimal-precision fat test showing how far the same schema stretches past headcounts. Numeric values land as strings with thousands separators, so ingestion handles markers like (D) and (Z) deliberately instead of tripping on them.

Fourteen documented columns define every estimate row, verified against the source's published file headers during the August 2026 research pass. The identifying columns compose a natural key - program, sector, group, commodity, class, statistic category, unit, geography, year, frequency - which is why a livestock series joins cleanly to your own commodity master without a crosswalk table.

Two further columns ride along only on subsets of records and therefore fold under additional fields on request: CV_%, published where an estimate carries a coefficient of variation, and LOAD_TIME, stamping when an estimate entered the database. Name them when you request a sample and they arrive switched on.

Geography - United States at six aggregation levels: national totals, states, counties, ZIP codes, watersheds and congressional districts. County is where protein work actually happens - feedstock concentration, processor catchment and supplier density all resolve at that grain.

Temporal - annual survey cycles plus the five-year Census of Agriculture, whose reference years reach back to the late 1990s in the historical files. Selected statistic categories publish on finer rhythms - weekly, monthly or point-in-time readings, carried in the FREQ_DESC column - so inventory and price series do not share one clock.

Granularity - one aggregated estimate per commodity x statistic x geography x period. There is no farm- or plant-level microdata anywhere; the agency publishes summaries by design, and that constraint is precisely what makes county-versus-county comparisons defensible.

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

• Raw-material availability tracking - cattle, hog, broiler and turkey inventory series flag herd expansion or liquidation quarters before they surface in processor guidance.
• Input-cost modeling - prices received for cattle, hogs, milk and eggs give the producer side of every packaged-protein bill of materials.
• County sourcing maps - county-resolved inventory shows where feedstock actually lives, feeding procurement footprints and supplier scorecards.
• Throughput context - slaughter and production categories set official baselines for how much protein moves through the system each period.
• Feed-side exposure - corn, soybean and hay acres and yields sit in the same schema, so the cost of gain never needs a second vendor.
• Disclosure-aware rural analysis - the (D) and (Z) markers show exactly which markets are too concentrated to publish, before a model treats silence as scarcity.

Investors and quant researchers build input-cost and supply-cycle factors for protein processors from inventory and prices-received series; see investors quants use cases. Market researchers and consultants size meat and poultry categories against official counts instead of panel extrapolation; see market researchers use cases. Data scientists and ML engineers train supply and demand models on county-resolved features that join on the natural key; see data scientists use cases. Competitive intelligence and product teams watch structural shifts in their own supply base decade over decade; see competitive intel product teams use cases. Journalists, academics and students cite the government's own numbers, down to the county.

• USDA AMS Market News - Livestock, Poultry & Grain - daily transactional reporting where Quick Stats supplies the settled statistical estimates; the noise beside the signal.
• USDA ERS Meat Price Spreads - the margin decomposition of the same pounds, farm gate to checkout scanner.
• USDA FoodData Central API - the composition of the finished products these animals become; the head-to-head sets out where each one wins.
• World Bank - Food Production Index & Food Commodity Indicators - global supply context wrapped around the US protein complex.

See the best Packaged Foods & Meats datasets ranking or browse the full Packaged Foods & Meats catalog.