BEA GDP-by-industry (iTable)

BEA GDP-by-industry is the industry dimension of the United States national accounts, published through the Bureau of Economic Analysis' Interactive Data application (iTable) and its underlying GDPByIndustry family of tables. Where a shipment survey counts what left the factory door in a census year, the national accounts measure what each industry adds: value created after subtracting what was consumed to create it, alongside gross output, intermediate inputs, KLEMS productivity components and employment statistics, all classified on a NAICS basis.

For anyone working in electrical components and equipment, the relevant line is Electrical equipment, appliances, and components - NAICS 335 territory - sitting inside sector 31-33 manufacturing with every other industry beside it rather than in a siloed extract. That arrangement is the quiet advantage: the electrical line arrives pre-joined in spirit to its customers, suppliers and substitutes, because they are literally adjacent rows. A companion UnderlyingGDPbyIndustry dataset supplies finer detail beneath the headline measures, and dozens of tables cover all NAICS industries with hundreds of rows across industries and years apiece.

Within Datadory's catalog of 1,744 datasets across 159 viable industries, the electrical components slice holds 22 catalogued records averaging 7.23 on our quality rubric against a 7.81 catalog-wide mark; this one rates 9/10, carried by field documentation, access reliability and the sheer length of its history. Get a sample of this dataset scoped to the industries and years you care about.

One documented observation, shaped as delivered:

# gdp-by-industry observation - electrical equipment line, delivered shape
industries         : Electrical equipment, appliances, and components
industry_code      : 335                          frequency : A
year               : 2023                         table_id : 1
value_added        : 74000 (millions of current dollars)
gross_output       : 218000                       intermediate_inputs : 144000

# same industry, chained-dollar view - real growth without price noise
year               : 2015                         measure : chained dollars
value_added        : 71000                        gross_output : 205000

Read what the shape already says. industries resolves to exactly one line - no sector aggregate burying the electrical story under appliances and components it does not mean. industry_code is the join key that keeps the same NAICS line identified across years, across tables and across companion BEA products. table_id states which production account the row belongs to, and frequency keeps mixed-cadence pulls honest by declaring A or Q outright.

The three measures travel together for a reason. value_added is the contribution-to-GDP figure a market-sizing model is actually about; gross_output is sales-scale, larger because it double-counts what the industry bought and resold embodied in product; and intermediate_inputs is the difference between them - the wire, magnet steel and semiconductors consumed along the way. A row where gross output jumps but value added does not tells you margins compressed; that inference costs nothing extra because all three numbers arrive on the same row. Figures shown are illustrative of magnitude and shape, not quoted vintages.

Eight fields carry each observation, and they reduce to three jobs. Identity: industries, the NAICS-basis label; industry_code, the numeric code behind the label; table_id, naming which GDP-by-industry table the row came from; frequency, the A-or-Q declaration that stops blended cadences; and year. Measurement: value_added in millions of dollars, chained and/or current depending on the table; gross_output; and intermediate_inputs.

Employment statistics and the full KLEMS decomposition arrive through sibling tables keyed identically, which is why they sit under additional fields on request rather than in the core eight: the dictionary stays immediately testable against the industries you know, and the wider panels fold into the same delivery once scoped.

One declaration belongs here: the definitions above were compiled from the documented GDPByIndustry contract during the August 2026 research pass rather than transcribed off live result screens, and the catalog record ships no verified sample extract. Treat the eight names as documentation-faithful - stated plainly because inferred is a fact about provenance, not an apology.

• Geography - United States, national totals by industry. Subnational questions have their own answer: state, county and metro detail lives in a separate BEA Regional product, not in these tables. Trying to extract a regional split from national accounts is the classic wrong-dataset move.
• Temporal - annual series spanning multiple decades back to 1997 and beyond on detailed tables, with quarterly estimates layered on for aggregate measures in recent years. Exact start year varies by table and vintage, and national-accounts estimates get revised as source data improve - pin the vintage when caching, because last year's 2023 may not equal this year's 2023.
• Granularity - one observation per industry-year (industry-quarter on the Q side), at NAICS detail down to roughly three-to-four-digit equivalents. That is the native unit of macro work, and it joins cleanly against anything else keyed on industry and period.

This granularity also explains why the record pairs naturally with the Economic Census data tables: Census answers structure at six-digit detail within every geography in census vintages, while this record answers movement annually between those vintages.

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

You choose the channel and the cadence; the eight-field dictionary travels unchanged through all three. Rows land flattened and typed - labels as strings, codes as strings, measures as numbers - so an industry-year table joins against your own NAICS-keyed series on the first attempt instead of after a parsing project. Cadence changes are a setting, not a migration, and a sample scoped to your industries and years proves the shape before anything ships.

• Market sizing with a spine - Baseline the US electrical equipment industry on official value added, then extend the sizing annually instead of waiting for the next census vintage to age out.
• Cycle and trend work - Two decades-plus of annual history with chained and current views makes real growth rates, price gaps and cyclical turns fall straight out of the series.
• Productivity attribution - The KLEMS decomposition says why the industry grew - labor deepening, capital services, materials intensity - not merely that it grew.
• Supplier-intensity analysis - Gross output beside intermediate inputs exposes what electrical equipment makers consume, the demand-side structure shipment values never show.
• Cross-industry benchmarking - Every NAICS line arrives in the identical shape, so electrical equipment ranks against machinery, automotive or renewables peers in one pass.
• Macro features for models - Typed industry-year rows with stable codes serve as exogenous regressors for forecasting pipelines and quant screens.

Data scientists and ML engineers treat the series as ready-made macro features keyed by stable industry code (data scientists use cases). Investors and quants read quarterly value added with decades of history as a cycle signal for industrial exposures (investors quants use cases). Market researchers and consultants anchor client sizings to federal accounts. Journalists and academics cite numbers nobody disputes. Sales teams draw territories along gross-output rankings. Developers ship economic-context widgets off one stable dictionary (developers builders use cases).

The national-accounts view reads differently depending on its neighbours. The Economic Census data tables supply the structural counterpart - establishments, payroll and shipments at six-digit NAICS 335 detail inside every county and metro - while this record supplies the annual flow between census vintages; the trade-off gets its own page at BEA GDP-by-industry vs Economic Census data tables. The NEMA electroindustry statistics & market programs adds the industry association's sentiment layer over the same territory, OpenEI energy data covers the power systems electrical equipment feeds into, and the SAM.gov contract awards record shows who actually wins public procurement. On the distribution side, the Digi-Key Electronics product catalog carries SKU-level reality where these tables stop at industry totals. Joined on NAICS code, the stack answers the question neither half can answer alone: whether the structure the Census counts agrees with the motion BEA measures. The full sweep of the slice lives at best electrical-components-equipment datasets, the electrical components equipment data hub, and the source behind both records at BEA, profiled.