U.S. Census Bureau Available APIs — Economic Census, Annual Survey of Manufactures, VIUS
Datadory delivers U.S. Census Bureau available APIs data covering the Economic Census (county-level establishment counts, payroll, receipts, capital expenditure and inventories for 2002 through 2022 across 130 documented variables), the Annual Survey of Manufactures for NAICS 31-33, and the 2021 Vehicle Inventory and Use Survey truck population - delivered as API feeds, files, or straight into your warehouse.
What is U.S. Census Bureau Available APIs — Economic Census, Annual Survey of Manufactures, VIUS?
The U.S. Census Bureau's API catalog is best read as three programs wearing one name. The Economic Census counts every American establishment every five years - 2002, 2007, 2012, 2017 and 2022 - and reports what each of them does: establishment and firm counts, sales and receipts, annual payroll, employment, capital expenditures, inventories, operating expenses and value added, down to individual counties, with 130 documented variables on its base endpoint alone. The Annual Survey of Manufactures fills the years between census waves for factory activity, publishing national sample estimates for all manufacturing establishments with paid employees across 2- through 6-digit NAICS from 2018 through 2021. And the Vehicle Inventory and Use Survey is the rarest of the three: a 2021 census of the US truck population producing national and state estimates of truck counts and miles broken out by body type, vehicle type, gross vehicle weight rating, trailer configuration and business use.
For anyone selling into or analyzing the auto parts economy, the two anchor industries sit ready to query: NAICS 3363 motor vehicle parts manufacturing and NAICS 4413 automotive parts, accessories and tire retailers, both cut by geography at whatever digit depth your question needs. The figures come from the U.S. Census Bureau. Get a sample of this dataset and see real industry-by-county rows before you build the pipeline yourself.
Sample rows from the Economic Census, ASM and VIUS
Each program returns flat tabular rows - a header line followed by one record per geography or characteristic. The shapes below follow the publisher's own documented example patterns:
Economic Census 2022, industry filtered to a single sector, nation
NAICS2022_LABEL EMP NAME GEO_ID
MOTOR VEHICLE PARTS MANUFACTURING (3363) ... United States 0100000USVIUS 2021, primary-characteristic tabulation, nation
NAME GEO_ID PRICHAR PRIVALUE VCOUNT VCOUNT_S
United States 0100000US <char> <value> <thousands> <flag>Annual Survey of Manufactures, 2021, all manufacturing
NAICS2017_LABEL EMP
MANUFACTURING (31-33) <employees>Every row answers one question cleanly: how many establishments, how many people on payroll, how many dollars of sales - or, for VIUS, how many thousand trucks of a given body type and weight class logged how many million miles. Because the response shape holds steady across geographies and industries, a script written against one county panel runs unchanged against the whole country. Request a sample and we will pull the exact NAICS-and-county slice your project needs.
What fields does the U.S. Census Bureau available APIs dataset include?
Field definitions below are verified against the publisher's documentation. The dictionary covers the fields that do most of the analytical work across the three programs; the long tail - sector-specific endpoints, additional ASM cost and inventory measures, and the remaining VIUS dimension codes - is kept as additional fields on request and ships with every sample.
Two columns earn special mention. NAICS2022 is the master filter: set it to 3363 and every downstream figure - establishments, payroll, receipts - narrows to motor vehicle parts manufacturing; set it to 44132 and you are reading the retailer rows. And VCOUNT arrives with a companion flag (VCOUNT_S) that grades the estimate's reliability, which matters because VIUS tabulates a sampled population rather than a full register.
What does coverage look like across geography, time and granularity?
Geography - the Economic Census reaches furthest: county level, plus state, region, metropolitan statistical area, offshore areas and the Island Areas. The manufacturing survey publishes national estimates only. The truck survey covers the nation and states, New Hampshire excepted in the published tabulations.
Temporal - five Economic Census waves: 2002, 2007, 2012, 2017 and 2022. Manufacturing estimates run 2018 through 2021, with an earlier time-series covering 2002-2016 kept separately. The truck survey is a snapshot of 2021 - the most recent federal read of its kind.
Granularity - industry-by-geography aggregates at 2- through 6-digit NAICS depth; nothing at establishment or company level, by design. Truck-survey rows are characteristic-by-state tabulations, not microdata, so a row reads "this many thousand trucks of body type X in weight class Y" rather than one line per vehicle. A single industry-by-county extract can run from hundreds of rows to tens of thousands depending on how broad a NAICS net you cast.
How is the data delivered?
API, files, or your warehouse. Daily, weekly, or hourly.
You choose the pipe and the cadence; the records stay identical either way. Teams typically start with flat files to validate a hypothesis about their market, then promote the same fields to a scheduled feed or land them directly alongside their own sales and inventory tables. Because the variable names hold steady across census waves and geographies, a county panel built on the 2022 wave extends backward to 2002 without a rewrite - which is what makes twenty-year trend lines practical rather than aspirational.
What do teams do with Economic Census, ASM and VIUS data?
Six uses come up again and again, each owned by a different seat at the table:
- Motor-vehicle retail benchmarking - official five-year establishment counts by county and NAICS digit to size the dealer and parts-store universe against private estimates (Market Researchers & Consultants).
- Territory sizing before rep coverage plans - county establishment and payroll counts decide where a parts territory justifies headcount, drawn before the first sales call (Sales & Growth Teams).
- Dealer-network density snapshots per wave - framing share conversations with the official count of who operates in a county, wave after wave (Competitive Intelligence & Product Teams).
- County-by-year demand panels - establishment and payroll series joined across census waves plus the manufacturing survey as exogenous features for forecasting models (Data Scientists & ML Engineers).
- Local reporting on dealership economies - citable federal establishment and payroll counts, suppression flags handled honestly rather than papered over (Journalists, Academics & Students).
- County-to-industry lookups inside products - the stable keyed query shape drops straight into an app's backend for territory pickers and market-size widgets (Developers & Data-Product Builders).
Dig into the market researchers use cases or the data scientists use cases to see where this dataset slots into each workflow.
Frequently asked questions about the U.S. Census Bureau available APIs
The questions below cover content and fit - what the programs measure, how deep the history runs, why some cells arrive flagged. Anything about delivery mechanics, cadence, or getting rows into your stack belongs in a conversation: request a sample and we will walk your exact use case.
Notes and related pages
- Automotive Parts & Equipment industry hub - every dataset in the slice, scored and ranked, with the catalog-wide stats behind the rankings.
- Best automotive parts & equipment datasets - where these Census programs sit among quality-scored neighbors such as BLS dealership employment counts and NHTSA safety recalls.
- NHTSA Vehicle Safety API vs Census Bureau available APIs - real-time defect signal against slow-burn market structure; different jobs, compared honestly.
- U.S. Census Bureau source profile - what else this publisher ships, including the CBP wholesale-trade slice that complements these retail and manufacturing counts.
- BLS QCEW auto dealers data - quarterly employment and wages between census waves, the natural cadence companion to this five-year structure.
Field dictionary
Every field below is documented against real records. The full dictionary ships with the sample.
| Field | Type | Definition | Example |
|---|---|---|---|
NAICS2022 | string | Economic Census: 2022 NAICS code used to filter industries - 3363 motor vehicle parts manufacturing, 44132 automotive parts retailers. | 3363 |
ESTAB | integer | Number of establishments in the industry-geography cell. | 914 |
EMP | integer | Paid employees during the reference period (Economic Census and ASM). | |
PAYANN | number | Annual payroll in $1,000 (Economic Census and ASM). | |
RCPTOT | number | Sales, value of shipments, or revenue in $1,000 (Economic Census and ASM). | |
FIRM | integer | Number of firms - companies that may operate several establishments. | |
VALADD | number | Value added in $1,000: sales less the cost of inputs (Economic Census and ASM). | |
CEXTOT | number | Total capital expenditures for buildings, structures, machinery and equipment, new and used, in $1,000. | |
GEO_ID | string | Geographic identifier code shared across all Census statistical products - join key for any county panel. | 0100000US |
NAICS2017 | string | ASM: 2017 NAICS code, sectors 31-33, filterable to 6-digit detail such as 3363. | 31-33 |
VCOUNT | number | VIUS: number of vehicles in thousands within the tabulated characteristic cell. | |
VMILE | number | VIUS: vehicle miles in millions accumulated by that cell of the fleet. | |
AMILE | number | VIUS: average miles per vehicle in thousands. | |
VEHTYPE | enum | VIUS: vehicle type dimension code. | |
GVWR | enum | VIUS: gross vehicle weight rating class dimension. | |
BTYPE | enum | VIUS: body type dimension code - the field that separates box trucks from dump trucks from pickups. | |
PRICHAR / PRIVALUE | string | VIUS: primary characteristic code and value pair defining what each tabulation row counts. |
Coverage summary
| Dimension | Coverage |
|---|---|
| Geography | United States: Economic Census to county level (plus state, region, CBSA, offshore areas, Island Areas); ASM national; VIUS national and state (excluding New Hampshire) |
| Time | Economic Census waves 2002, 2007, 2012, 2017, 2022; ASM 2018-2021 with a separate 2002-2016 series; VIUS 2021 |
| Granularity | Industry x geography aggregates at 2- to 6-digit NAICS; VIUS characteristic-by-state tabulations rather than vehicle microdata |
Questions buyers ask
Can this dataset pin down the auto parts market at county level?
Yes. The Economic Census covers NAICS 3363 motor vehicle parts manufacturing and NAICS 4413 automotive parts retailers down to individual counties, with establishment counts, employment, payroll, sales and inventories on each row. The same variables roll up to state, metro area and national totals, so a bottom-up county view and a top-down market view come from one consistent schema.
How far back does the history go?
Five census waves: 2002, 2007, 2012, 2017 and 2022, giving roughly two decades of structural history in one queryable shape. Between waves, the Annual Survey of Manufactures contributes national estimates for 2018-2021, and a separate earlier series reaches back across 2002-2016. The truck survey stands alone as a 2021 snapshot - the only recent federal profile of its kind.
What does the 2021 Vehicle Inventory and Use Survey add?
A census-flavored profile of the US truck population as it actually operated in 2021: vehicle counts and miles by body type, vehicle type, gross vehicle weight rating class, trailer configuration and business use, published in six dataset variants covering all vehicles, in-use vehicles and business-use vehicles. For parts planning it answers which weight classes and body styles dominate the running fleet - the demand base behind wear parts.
Why do some returned values show letter flags instead of numbers?
Confidentiality and reliability rules, not missing data. Flags such as D, S, X and r mark values withheld, unavailable, or too imprecise to publish where too few establishments would otherwise be identifiable - standard practice across federal statistics. State and national aggregates stay complete, so treat flagged county cells as signals to aggregate around rather than as zeros in a sum.
Does this include company names or contact lists?
No - and that is the point of the design. These are aggregate statistics: counts, dollars and hours summed over industries and geographies, never establishment identities, so no privacy review gates your analysis. If your workflow needs named accounts to call, pair this dataset with a commercial contact source and use these figures as the sizing layer that tells you which territories justify the outreach.
How current can the figures be?
Structural data moves slowly by nature, and the newest Economic Census wave is 2022 - published after a multi-year collection cycle. That lag buys completeness no quarterly survey matches: every establishment, not a sample. Teams needing fresher signal typically interleave higher-frequency labor or retail indicators between waves and re-anchor to the census baseline each cycle.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.