data.gov federal open data catalog

It is the United States government as one addressable table: 552,271 dataset records as of August 2026, drawn from federal agencies, states, counties and city portals, each described in the same DCAT-US metadata vocabulary. A query for electrical equipment returns the spread in one screen - the EPA's Supply Chain Greenhouse Gas Emission Factors v1.3 by NAICS-6, covering 1,016 six-digit commodities with electrical equipment manufacturing among them; NREL's SMART-DS synthetic electrical network models; wind turbine gearbox condition-monitoring benchmarking data; a translated patent titled Electrical Equipment Housing; Los Angeles building-and-safety inspections modified August 17, 2026. Hundreds of records answer that single query, and the catalog answers every other query besides.

That breadth is the point and the trade-off. This is the discovery layer beneath every single-agency collection: where the OpenEI energy data record curates the Department of Energy's shelf, this one reaches every other cabinet department plus state capitols and city halls. It scores 8/10 against a catalog-wide 7.81 average across all 1,744 datasets we hold, and ranks eighth of ten on the best electrical components & equipment datasets list precisely because it tells you what else exists.

Get a sample of this dataset - name the topic, agency or geography, and matching records come back shaped exactly like the dictionary below.

Five records pulled from an electrical-equipment query during our August 2026 research pass, shown as they arrive:

title      : Supply Chain Greenhouse Gas Emission Factors v1.3 by NAICS-6
publisher  : U.S. EPA Office of Research and Development (ORD)
media_type : text/csv                        modified : 2024-07-05
note       : GHG emission factors for 1,016 NAICS-6 commodities,
             electrical equipment manufacturing included (2022 data)

title      : Building and Safety Inspections
publisher  : data.lacity.org
media_type : application/json, text/csv, KML modified : 2026-08-17

title      : SMART-DS Synthetic Electrical Network Data OpenDSS Models
             for SFO, GSO, and AU
publisher  : National Renewable Energy Laboratory (NREL)

title      : Wind Turbine Gearbox Condition Monitoring Vibration Analysis
             Benchmarking Data
publisher  : National Renewable Energy Laboratory

title      : Patent AT-E400988-T1: [Translated] ELECTRICAL EQUIPMENT HOUSING
publisher  : National Center for Biotechnology Information (NCBI)
media_type : text/html                        checked  : 2026-07-31T21:19:39

Read what the variety argues. The first row is supply-chain carbon accounting ready to join onto a vendor master - emission factors keyed by six-digit NAICS, electrical equipment manufacturing explicitly covered, built on 2022 data. The second is municipal enforcement tempo: building-and-safety inspections for a city of four million, in tabular and geospatial formats, its modification stamp only days old at our research pass. The third and fourth are national-laboratory engineering assets - synthetic grid models and rotating-machinery vibration benchmarks - that never surface through commercial channels. The fifth proves the index reaches even intellectual property: a foreign patent, translated, classified under electrical equipment housing.

Every row carries its publisher, its file formats and its two timestamps, so judging whether a record is alive takes reading the row, not asking anyone.

Fourteen verified columns, defined in the table below. Three do the structural work. identifier - usually a persistent DOI - is the join key to citation graphs, repositories and any other DOI-keyed set. The paired timestamps, modified and last_harvested_date, put freshness on the row itself: one date belongs to the publishing program, the other to the catalog's verification pass, and disagreement between them is itself a signal about how actively a program maintains its data. And distribution lists file-level resources with explicit media types, so a format-inventory question - CSV here, KML there, nothing but HTML for that one - resolves by reading, not by opening.

Then there is the administrative plumbing few commercial aggregators bother to carry: bureauCode and programCode tying records to budget programs, a structured contactPoint per dataset, and a popularity score that turns the whole corpus into a demand map. Those columns are why analysts treat this record as reference infrastructure rather than another listing.

Three chips summarize the footprint:

• Geography: the United States at every level - federal agencies, states, counties and cities in one index - with some records carrying global or geospatial extents. One query spans all of it; nobody maintains a wider net.
• Time frame: continuously growing, and honestly labeled - each record publishes its own modification date and verification stamp, so staleness is visible per row rather than hidden behind a blanket promise. Historical depth varies by publishing program.
• Granularity: one row per dataset record, linked to file-level resources whose media types ride in the row. Underlying granularity varies by agency - census tables, inspection events, engineering models - which is exactly why the metadata layer earns its keep as the map before the terrain.

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

Pick the channel your pipeline already speaks and set the cadence to match the decision being fed. Every delivery ships the field dictionary unchanged, sample rows for validation, and records flattened so publisher, formats and both timestamps sit in one row - a freshness audit becomes a sort, not a survey. Get a sample and the anatomy above comes back populated with the topics you name.

Ranked by how directly one well-formed index settles the day job:

1. Discovery and make-or-buy decisions. Before commissioning custom collection, teams enumerate what government already publishes on a topic - the catalog answers does this exist in minutes, with named publishers attached.
2. ESG and supply-chain carbon accounting. NAICS-keyed emission factors join onto vendor masters by industry code, turning scope-3 estimation into a lookup with a documented vintage (market researchers use cases).
3. Policy, grant and market-entry research. Bureau and program codes plus publisher provenance map who funds what at which level of government - the substrate for positioning any regulated product (competitive intel product teams use cases).
4. Data-product prototyping. Uniform metadata over half a million records makes this the cheapest place to test a retrieval or classification idea before committing to deeper feeds (developers builders use cases).
5. Municipal and grid analytics. City inspection records and national-laboratory grid models sit one query apart - unusual adjacency, routinely useful for siting and permitting work (data scientists use cases).

• Market Researchers & Consultants - mapping the public-data perimeter around a market before buying anything private (data for market researchers).
• Competitive Intelligence & Product Teams - regulatory, environmental and municipal signals on one index (data for competitive intel and product teams).
• Developers & Builders - uniform metadata at scale for retrieval prototypes and data products (data for developers and builders).
• Data Scientists & ML Engineers - DOI-keyed records with tags and timestamps for corpora and entity resolution (data for data scientists).
• Investors & Quants - publication tempo and program activity as slow-cycle indicators (data for investors and quants).
• Sales Growth Teams - account intelligence grounded in which agencies and programs buy what a component maker sells (best data for electrical components equipment teams).

Three things, stated up front.

First, this is a catalog of descriptions before it is a pile of data. Each row describes a dataset held by its publishing program; the underlying files vary in depth and age by agency. We cut samples that resolve that distance - named records first, then the referenced content where you want it.

Second, relevance is a vocabulary conversation. Half a million records reward a precise brief: the topics, agencies and geographies you care about decide whether a cut returns fifty rows or five thousand.

Third, the electrical slice is discovered, not pre-assembled. Nothing here arrives pre-filtered, which is the same property that makes it the widest net in the industry pool - and the reason it pairs so well with a curated set.

Notes worth reading next:

• OpenEI energy data - the curated Department of Energy shelf; the head-to-head comparison settles breadth versus depth.
• EU Open Data Portal - Europe's union-level counterpart index.
• BEA GDP-by-industry (iTable) - the macro layer beneath component demand.
• SAM.gov contract awards - federal buying, award by award.
• Electrical components equipment data hub - the full pooled industry view.