Understanding Spend-Based Data Sources

For a detailed guide on EXIOBASE specifically, see our EXIOBASE guide. For the complete outline on principles and data modeling underlying these methodologies, see The Science Behind Spend-Based Emission Factors .

What are spend-based emission factors?

Spend-based emission factors estimate greenhouse gas emissions based on how much money was spent on a product or service, rather than on physical quantities like kilograms or kilowatt-hours. They are expressed in kgCO2e per unit of currency (e.g. kgCO2e per EUR) for most datasets (EXIOBASE also provides a breakdown into constituent gases) and are derived from economic models that map the relationship between monetary flows and environmental impacts across sectors of the economy.

Spend-based factors exist because most organizations already track what they spend, even when they do not have detailed physical data about what they purchased. This makes spend-based methods the most accessible way to estimate scope 3.1 (purchased goods and services) and scope 3.2 (capital goods) emissions, especially for organizations with hundreds or thousands of procurement line items where collecting activity data for each one would be impractical.

The trade-off is precision. Spend-based factors represent sector-level averages: they cannot distinguish between a high-efficiency supplier and a low-efficiency one, or between a product made from recycled materials and one made from virgin inputs. A price change (discount, inflation, currency fluctuation) changes the calculated footprint without any real-world emission change. This is why spend-based methods are best suited for screening and hotspot identification, not for tracking reduction progress over time.

How EEIO models work

Spend-based emission factors are derived from Environmentally Extended Input-Output (EEIO) models. Understanding the basics of how these models work helps you use the factors correctly.

An Input-Output model describes the flow of goods and services between all sectors of an economy. It captures how much each industry buys from and sells to every other industry, creating a complete map of economic interdependencies. For example, the automotive sector buys steel from the metals sector, electronics from the electronics sector, and energy from the utilities sector. These, in turn, buy goods and services from other sectors.

The “environmentally extended” part adds environmental data on top of this economic structure. By linking each sector’s monetary output to its greenhouse gas emissions, the model calculates the emissions intensity of every euro (or dollar) spent in each sector. Critically, this includes not just the direct emissions from the sector itself, but all the upstream emissions embedded in its supply chain. This full upstream coverage is built into the model by design.

In practice, this means that when you multiply your spend by an EEIO-derived emission factor, the result already includes emissions from raw material extraction through every intermediate processing step. You do not need to separately estimate supplier-level or sub-supplier-level emissions.

The databases we support are each built on a different EEIO model, described in detail in the Spend-based databases available via Climatiq section.

Each model uses different underlying economic data, sector classifications, and pricing conventions, which is why mixing factors from different EEIO models in the same calculation is not recommended.

Spend-based databases available via Climatiq

We offer spend-based emission factors from six sources. Each uses a different underlying model, covers different geographies, and has different requirements for input data.

EXIOBASE is the most widely used multi-regional EEIO database for spend-based emission factors. It covers 44 countries and 5 Rest of World regions with 183 product categories and 149 industries. Emission factors included in Climatiq are in basic prices (EUR). Like all multi-regional (MRIO) models, EXIOBASE estimates emissions from trade (imports and exports) using actual emissions data from the trading regions. This makes estimates more accurate than single-region models like USEEIO (see below). For detailed guidance, see our EXIOBASE guide.

CEDA is a multi-regional EEIO database provided by Watershed. CEDA is a multi-regional model, which covers the US and 147 other countries, with 400+ industries. This matters most when calculating the footprint of products consumed in the US that rely on imported components or materials: because CEDA models each exporting country individually, it captures the actual carbon intensity of production in those countries rather than relying on generic global averages for imports.

USEEIO, published by Cornerstone is the successor to the US EPA’s USEEIO dataset. The EPA dropped publication of the data, and it was taken up by the Cornerstone Initiative, which is supported by Watershed. Unlike CEDA, Cornerstone is a single-region input-output model focused on the US economy. Because it models only one region, it assumes US technology for all imports, which makes it less accurate for products with significant international supply chains. Cornerstone covers 400+ US sectors.

Although CEDA and Cornerstone both use core data from the US Bureau of Economic Analysis (BEA) and the EPA, their underlying algorithms for cleaning data, allocating environmental flows, and mapping sectors developed independently. This means you will see different emission values for the same region and industry across the two datasets. Watershed is actively working to harmonize the US modeling approach between the two, with a unified model expected in 2026. Until then, variations between the datasets are expected. EXIOBASE is built on a separate foundation, primarily Eurostat supply-use tables and national statistical office data, and is methodologically independent from CEDA and Cornerstone.

The UK Government provides UK-specific spend-based emission factors derived from the UK IO model. These are separate from the broader BEIS/DEFRA dataset, which is primarily activity-based. Spend-based factors use basic prices in GBP. While it only provides emission factors for the UK, it uses data from MRIOs like EXIOBASE to reflect actual emissions intensities of products made in other regions.

OpenIO-Canada is an open-source Canadian EEIO model based on Statistics Canada’s supply-use tables. It provides spend-based emission factors for the Canadian economy, covering Canadian provinces individually. While it only provides emission factors for Canada, it uses data from EXIOBASE to reflect actual emissions intensities of products made in other regions. Emission factors are expressed in purchaser prices (CAD), including taxes, retail margins, and distribution costs. The current version available through our platform is v2.11.

Market Economics Limited (New Zealand) is an independent New Zealand-based consultancy that provides spend-based emission factors derived from its Consumption Emissions Modeling work. The dataset covers the New Zealand economy with factors expressed in NZD. It is the only New Zealand-specific spend-based source available through our platform. It is a single-region model (like USEEIO). It assumes that goods and services purchased from outside of New Zealand generate the same quantities of emissions per dollar of expenditure as equivalent goods and services produced in New Zealand.

Which one should you use?

For global or multi-country scope 3.1 assessments, EXIOBASE is the most common choice. It is the most widely used MRIO database for spend-based reporting. A European manufacturer spending across suppliers in 15 countries, for instance, can estimate scope 3.1 across all regions using a single consistent model without stitching together multiple national datasets.

For US-centric reporting where products have international supply chains, CEDA models each exporting country individually. This matters if you are a US company importing components from countries like China, Taiwan, or South Korea and want emission factors that reflect the carbon intensity of production in those countries rather than a generic global average.

For US-specific reporting with primarily domestic supply chains, Cornerstone provides high sectoral resolution for the US economy. A US food manufacturer sourcing mostly from domestic farms and processors, for example, benefits from Cornerstone’s detailed sector-level factors and its lineage from the EPA methodology.

For UK-specific reporting, BEIS/DEFRA spend-based factors are often required or recommended. A UK retailer reporting under the Streamlined Energy and Carbon Reporting (SECR) framework would typically use BEIS for its purchased goods and services.

For Canadian-specific reporting, OpenIO-Canada provides province-level spend-based factors built from Statistics Canada’s supply-use tables. A Canadian manufacturer reporting under federal or provincial GHG requirements would use OpenIO-Canada for its domestic purchased goods and services.

For New Zealand-specific reporting, Market Economics Limited is the only NZ-specific spend-based source available through our platform. A New Zealand-based organization estimating scope 3.1 for domestically procured goods and services would use this dataset.

Do not mix factors from different EEIO models within the same scope 3.1 assessment. Their methodologies, sector classifications, and pricing conventions differ, which produces inconsistent results. Pick one source and apply it consistently. This is especially important for CEDA and Cornerstone: despite sharing some underlying data sources, their independent methodologies mean they produce different values for the same industry and region.

Key concepts

Basic price vs. purchaser price

This is the single most common source of error in spend-based calculations.

Basic price is the cost at the factory gate: what the producer receives for the product, without taxes, trade margins, or transport costs. Purchaser price is the total amount you actually pay, including trade margins (wholesale and retail markups), transport margins (shipping and logistics), and tax margins (non-deductible VAT, duties).

EXIOBASE emission factors are built on basic prices. If you apply an EXIOBASE factor directly to what you paid (purchaser price), you will overestimate your emissions because your price includes margins that are not part of the product’s production footprint.

Before applying the factor, you need to strip out the relevant margins. Which margins to remove depends on how you purchased the product. If you bought through a retailer or distributor, you need to remove trade, transport, and tax margins. If you bought directly from the manufacturer, trade margins do not apply and should not be removed.

Inflation adjustment

EEIO models are built on economic data from a specific base year. EXIOBASE v3.11+, for example, uses a 2023 base year. Where possible, we recommend using the most recent version of the dataset available, as newer versions are built on more recent economic and environmental data and will produce more accurate results. Older versions like EXIOBASE v3.8.2 (base year around 2017) remain valid for GHG accounting, but unless you are specifically quantifying emissions for a 2017 reporting period, the underlying carbon intensity data is increasingly outdated and does not capture structural changes in the economy since then. If your spend data is from 2025 and your emission factor is based on an earlier year, you need to deflate the expenditure to the model’s base year before applying the factor. Without this adjustment, inflation alone increases your calculated footprint without any real change in emissions. Our Mapping Agent, Procurement and Basic Estimate endpoints handle inflation adjustment automatically for spend-based calculations when inflation adjustment is enabled in the request. See the API reference for detailed documentation.

Sector classification

Each EEIO model uses its own sector classification scheme. EXIOBASE has 200 product categories; EPA/USEEIO uses NAICS-based sectors; CEDA uses US BEA codes which are based on NAICS codes, the UK Government uses CPA codes.

Scope breakdown in additional indicators (EXIOBASE v3.11+ and Enterprise CEDA)

Some spend-based emission factors, including EXIOBASE v3.11+ and Enterprise CEDA, provide a breakdown of the total emission factor into scope 1, scope 2, and scope 3 components via additional indicators. It is important to understand that these scopes refer to the supplying industry’s scopes, not yours.

• Scope 1: Direct emissions from the supplying industry (e.g. your supplier burning fossil fuels in their production process)
• Scope 2: Indirect energy emissions from the supplying industry (e.g. your supplier’s electricity consumption)
• Scope 3: All other indirect emissions from the supplying industry (i.e. your supplier’s own procurement and upstream supply chain)

This breakdown is useful for understanding where emissions sit within your supplier’s value chain, which can inform supplier engagement and reduction strategies. If you know a supplier’s actual scope 1 or scope 2 emissions, you can substitute them for the modeled values, and the breakdown also helps you sense-check the plausibility of supplier reduction plans. However, when reporting your own scope 3.1, 3.2, or 3.15 emissions, use the total that already represents the full upstream impact of your purchase.

When (and when not) to use spend-based factors

When spend-based is the right approach

• Activity data is unavailable. You have procurement data (invoices, ERP exports) but not physical quantities for what was purchased. This is the most common scenario for scope 3.1.
• You need a fast screening. You want to identify which procurement categories drive the most emissions before investing in detailed data collection.
• You have thousands of line items. Mapping each procurement line to an activity-based factor manually is impractical. Spend-based methods scale to large procurement datasets.
• Your framework accepts it. The GHG Protocol provides a spend-based calculation method for multiple scope 3 categories, including purchased goods and services (3.1), capital goods (3.2), upstream transportation (3.4), waste (3.5), business travel (3.6), employee commuting (3.7), and others. PCAF and most corporate reporting frameworks also accept spend-based methods. Spend-based is typically the least specific method available and should be used as a starting point where more detailed activity data is not yet available.

For a full comparison of activity-based and spend-based approaches, see the Activity-Based vs. Spend-Based section of our database overview guide.

When not to use spend-based factors

• Scope 1 and scope 2. These scopes require activity-based data (fuel consumption in liters, electricity in kWh). If you only have expenditure data for fuel or electricity, convert it to physical units using average prices before applying activity-based factors.
• Product carbon footprints. PCFs require process-level, material-level data. Use ecoinvent, Carbon Minds, or sustamize instead.
• Supplier-specific comparisons. Spend-based factors cannot distinguish between individual suppliers. If you need to compare suppliers on an emissions basis, use activity-based factors or request primary data from the supplier.
• Tracking emission reductions over time. Spend-based factors reflect sector averages and cannot detect real operational improvements. If your supplier switches to renewable energy, your spend-based result will not change. Use activity-based methods where you need to demonstrate actual progress.
• Drawing conclusions from price-driven changes. A negotiated discount, a currency fluctuation, or a period of deflation will reduce your calculated footprint without any real-world emission change. The model sees less money and returns fewer emissions. Be aware of this limitation when interpreting results.

Technical mistakes to avoid

Applying purchaser price without converting to basic price. If the source requires basic prices (EXIOBASE, BEIS), using your total purchase price without stripping out the relevant margins leads to overestimation. Which margins to remove depends on how you purchased: if you bought through a retailer or distributor, remove trade, transport, and tax margins. If you bought directly from the manufacturer, trade margins do not apply.

Ignoring inflation. Applying an emission factor built on older economic data to current-year spend without deflating the expenditure overstates your footprint. Always use the most recent dataset version available. If you are using an older version, adjust for inflation. Our Mapping Agent and Procurement endpoint handle inflation adjustment automatically.

Mixing EEIO models. Using EXIOBASE for some categories and CEDA or Cornerstone for others within the same scope 3.1 assessment creates methodological inconsistencies. Different models use different sector boundaries, pricing conventions, and underlying economic data. Pick one and apply it consistently.

Double-counting with activity-based factors. Fuel, electricity, and transport costs should be calculated with activity-based factors under scope 1, 2, or other scope 3 categories (e.g. scope 3.3 FERA). Including them in a spend-based scope 3.1 calculation counts them twice.

Selecting factors based on the lowest value. Searching across multiple databases and picking whichever factor produces the smallest number for each category is not methodologically defensible. Choose your source based on geographic fit, methodological relevance, and data quality, not on which factor produces the smallest result.

Browse spend-based emission factors in the Data Explorer. For the science behind the methodology, see The Science Behind Spend-Based Emission Factors . For a step-by-step technical walkthrough of spend-based calculations, see Calculating Scope 3.1 Emissions. For questions, contact us .

Related resources