What Are EPA EEIO Emission Factors and How Are They Used?
If you have looked into the engine room of spend-based carbon accounting, you have met the acronym USEEIO: the US Environmentally-Extended Input-Output model. It sounds like something only a government economist could love, but its output, a table of emission factors per dollar for about a thousand US commodities, lets a small company estimate its supply chain emissions without surveying a single supplier. ERG, one of its maintainers, calls it one of the world's most widely used datasets for Scope 3 accounting.
The factors are published as the Supply Chain Greenhouse Gas Emission Factors for US Industries and Commodities. EPA released v1.2 in 2023 and v1.3 in July 2024 (2022 emissions data, IPCC AR5 global warming potentials, 2022 dollars, 1,016 commodities). The USEEIO model's new home is the Cornerstone Sustainability Data Initiative, founded by Stanford's Sustainable Solutions Lab, Watershed and ERG, which published v1.4 on 14 October 2025 (2023 emissions data, AR6 global warming potentials, 2024 dollars, the same 1,016 commodities). This guide covers how the factors are built, which of the three columns to use, how to apply them, and where they fall short.
The anatomy of an EEIO factor
An EEIO (Environmentally-Extended Input-Output) factor links spending to emissions. It answers the question: for every dollar spent on this commodity, how many kilograms of CO₂e were emitted to produce it?
To build the factors, the model combines two datasets:
- Economic input-output tables: the Bureau of Economic Analysis (BEA) Make and Use tables track how money flows between several hundred US industries (how much the construction sector buys from the steel sector, for example). v1.3 uses the 2017 benchmark tables.
- Environmental data: this is the "extended" part. A sector attribution model assigns the emissions in EPA's US Greenhouse Gas Inventory (carbon dioxide, methane, nitrous oxide and fluorinated gases) to the industries that emit them.
Dividing an industry's direct emissions by its economic output gives its direct intensity. For a hypothetical industry that emits 1 million tonnes of CO₂e and has $2 billion of output, that is 0.5 kg CO₂e per dollar. The supply chain factor goes further: the input-output model traces every input the industry buys, and the inputs to those inputs, and adds their emissions. The result is direct plus upstream emissions per dollar.
In v1.3 the combined factors range from 0.029 to 3.924 kg CO₂e per dollar, with a median of 0.173. Cement manufacturing has the highest. Professional services sit at the low end: Offices of Lawyers is 0.041 and Architectural Services 0.103.
Without margins, margins and with margins (cradle to gate vs cradle to shelf)
Each commodity in the dataset has three factors. Picking the right one avoids double counting or leaving emissions out.
1. Supply chain factor without margins (cradle to gate)
Emissions from raw material extraction through manufacturing, up to the factory gate. EPA's documentation says organizations buying intermediate products at the factory gate will likely find this factor the most appropriate.
2. Margins (gate to shelf)
Emissions from moving and selling the product after it leaves the factory and before it reaches you:
- Transportation (EPA assumes the same mix of modes for every commodity)
- Wholesale trade
- Retail trade
Most services carry no margin factor: for Architectural Services it is zero.
3. Supply chain factor with margins (cradle to shelf)
The sum of the two. EPA's documentation says end users of products will likely find this factor the most appropriate. For a company buying office supplies, laptops or furniture from a retailer or distributor, this is the one to use. Office furniture (NAICS 337214), for example, is 0.184 without margins, 0.056 for margins and 0.240 with margins.
All three are expressed per dollar of purchaser price. EPA also warns against adding the without-margins or margins result to the with-margins result, since that counts the same emissions twice.
Why these factors are widely used
- Coverage: v1.3 gives a factor for 1,016 six-digit NAICS commodities, from "Fluid Power Cylinder and Actuator Manufacturing" to "Drycleaning and Laundry Services (except Coin-Operated)". Those 1,016 codes share 353 distinct sets of factors, because several NAICS codes map to the same USEEIO sector: Architectural Services and Engineering Services both carry 0.103, for example.
- Transparency: the methods, data and code are public. A reviewer can trace any factor back to its NAICS code, version and price year.
- A stated price year: each version is expressed in one year's dollars (2022 for v1.3, 2024 for v1.4). Prices generally rise over time, so a dollar spent in 2024 typically buys less than a dollar did in 2022, and applying a 2022-dollar factor to 2024 spend without adjustment tends to overstate emissions. EPA's documentation calls adjusting the dollar year "optional but recommended". Price change also explains much of the movement between versions: the v1.4 documentation reports a correlation of -0.9 between each commodity's 2022-2024 price change and its factor change.
Their main limitation is that they are averages. A spend-based factor cannot tell one supplier from another in the same sector, it moves with price rather than with actual emissions, and EPA's model assumes imported goods are made the way US goods are made (with international transport left out of the margins). Our engineering estimate for spend-based screening is roughly ±25-40% uncertainty at the category level; the GHG Protocol publishes no figure. The GHG Protocol's Scope 3 Standard describes EEIO data as particularly useful for screening, and recommends supplier data for the activities that matter most.
How to use these factors correctly
Applying these takes a consistent method you can explain to a reviewer.
- Step 1: Map your ledger. Export a year of spend and assign each account or vendor line to the closest NAICS commodity. A cloud hosting bill maps to Data Processing, Hosting, and Related Services (NAICS 518210, 0.093 kg CO₂e per 2022 dollar); legal fees map to Offices of Lawyers (541110, 0.041). Leave out electricity (Scope 2; the dataset has no electricity factor), fuel you burn (Scope 1), and lines with no purchased good or service behind them, such as payroll, taxes and loan repayments. Report travel spend under Category 6 rather than Category 1.
- Step 2: Adjust the price year. Restate spend in the factor's dollar year (2022 for v1.3) with a price index. EPA's documentation describes commodity-specific price indices; an economy-wide index such as the BEA GDP deflator is a simpler alternative.
- Step 3: Apply the factor. Multiply the dollars by the factor. $10,000 of commercial printing (NAICS 323111, 0.236 with margins) gives 2,360 kg, or 2.36 tCO₂e.
- Step 4: Document the version. The Scope 3 Standard requires you to report the types and sources of data you used, including emission factors and GWP values. State the dataset, version, price year and GWP set, for example "EPA Supply Chain GHG Emission Factors v1.3, 2022 USD, AR5". Versions differ: between v1.2 and v1.3, 796 of the 1,016 with-margins factors fell by more than 5% and 139 rose by more than 5%. Solid Waste Landfill went from 10.989 to 0.988 kg per dollar because v1.3 gave every waste sub-sector the same factor; v1.4 restores the waste detail.
Common mistake: classifying by the vendor's industry instead of by what you bought. If you buy laptops from an office-supply store, use Electronic Computer Manufacturing (NAICS 334111, 0.058 with margins), not Office Supplies and Stationery Stores (453210, 0.111). The product's with-margins factor already includes the retail step, and the store's own code describes the retail service, not the laptop.
EPA's supply chain factors let a 50-person company build a Scope 3 screen from public, documented factors without surveying any supplier. Because they are sector averages, treat the result as a screening estimate: use it to find the categories and suppliers that matter, then replace the largest lines with supplier data over time.
Upload your accounting export at aisustainablefuture.com/carbon-draft and get a cited, screening-grade carbon report in minutes. Scope 1 and 2 from $20, Scope 3 from $49. For US purchases, Carbon Draft uses EPA v1.3 factors adjusted to each purchase's year with the BEA GDP deflator, and every factor in the report is cited to its source document and table or section. The methodology page lists the sources for other countries.
