Double counting in Scope 3 emissions: why the GHG Protocol allows it, and when it is a real problem

Double counting in Scope 3 means the same physical tonne of carbon appears in more than one company's emissions report. It is not an error and the GHG Protocol does not try to prevent it. Your supplier burns gas making a component; that is their Scope 1 and your Scope 3 category 1. Your customer then counts the finished product in their own Scope 3. One tonne, three reports, all of them correct. The standard treats this as a feature, because the point of Scope 3 is to give every company in a value chain a reason to act on emissions it influences but does not directly control.

This trips people up because Scopes 1, 2 and 3 are mutually exclusive within a single company. Nothing you report can sit in two of your own scopes. That neat property does not extend across company boundaries, and expecting it to is where the confusion starts.

Why the GHG Protocol allows it deliberately

Consider a third-party carrier moving your finished goods. The diesel is the carrier's Scope 1. It is also your Scope 3 category 9, downstream transportation. If only one of you were permitted to report it, which one? If the carrier alone counts it, you have no reported incentive to consolidate shipments, switch modes, or choose a lower-emission carrier, even though your decisions determine most of that fuel burn. If you alone count it, the carrier has no reported incentive to buy more efficient trucks.

The GHG Protocol's answer is that both count it, because both influence it. Overlapping inventories are the mechanism by which pressure travels along a supply chain. A large buyer asking suppliers for carbon data works precisely because those suppliers' direct emissions are the buyer's Scope 3. Remove the overlap and the entire supplier-engagement model collapses.

Where double counting does become a real problem

Three situations turn an accepted overlap into a genuine distortion, and it is worth knowing them because the first two get argued about in public and the third will cost you money.

Adding Scope 3 across companies

You cannot sum the Scope 3 totals of many companies to get a regional, sector or national figure. Do that and you will count some emissions five or six times. This is why corporate inventories and national greenhouse gas inventories are built on different principles: national accounting is territorial and each tonne is counted once, corporate Scope 3 accounting is influence-based and deliberately overlaps. Analysts who blend the two produce inflated numbers. If someone hands you a market-sizing figure built from aggregated Scope 3 disclosures, treat it as unreliable.

Overlap inside your own Scope 3

This is the one that affects your report directly. The 15 categories are not perfectly disjoint in practice, and spend-based estimation makes overlap easy to create. Freight you paid for might appear in category 4 as upstream transportation, and appear again inside category 1 if the input-output factor you applied to purchased goods already embeds the supplier's delivery emissions. Input-output factors cover an entire supply chain by construction, so layering an activity-based freight calculation on top of a spend-based goods estimate double counts the freight leg. Leased assets are the other classic: a facility can be counted as Scope 1 under operational control and again in category 8 if the boundary was never settled.

Claiming the same reduction twice

Overlapping inventories are fine; overlapping claims are not. If you and your supplier both fund the same efficiency project and both claim the full reduction against your targets, the world gets one reduction and two claims. The same problem applies to carbon credits retired by more than one party and to renewable energy attributes claimed by both the generator and the buyer. The GHG Protocol is explicit that companies should acknowledge potential double counting when making claims about Scope 3 reductions. Report the reduction, state who funded it, and do not present a shared achievement as exclusively yours.

How to avoid double counting inside your own inventory

1. Settle the organizational boundary before anything else. Decide operational, financial or equity control, apply it to every entity, and check that nothing counted as Scope 1 also appears in a leased-asset category. Our guide to operational vs financial control covers the choice.
2. Know what each factor already includes. A spend-based input-output factor for purchased goods typically bundles the supplier's own inputs, energy and inbound freight. An activity-based supplier factor usually covers a narrower gate-to-gate boundary. Mixing methods across categories is fine; not knowing which boundary each factor uses is not.
3. Pick one method per activity. If you calculate outbound freight from tonne-kilometers in category 9, make sure the same shipments are not also inside a spend-based estimate elsewhere. Write down which category owns which activity.
4. Handle upstream fuel and energy carefully. Category 3 covers the extraction, production and transport of the fuels and electricity you consume, and specifically excludes what is already in your Scope 1 and 2. It is a common place to accidentally recount combustion.
5. Document exclusions and overlaps in the inventory management plan. An assurance reviewer is not looking for a perfectly disjoint set of categories, which does not exist. They are looking for evidence that you thought about it and applied a consistent rule.

The reason spend-based screening is still the right starting point, despite the overlap risk, is that it shows you the shape of the footprint in days rather than quarters. You then graduate the two or three categories that dominate to activity data, and at that moment you remove the spend estimate they replace. Doing it in that order is what keeps the arithmetic clean. Our spend-based vs activity-based comparison covers the tradeoff, and the Scope 3 materiality assessment guide covers picking which categories deserve the upgrade.

Does double counting make Scope 3 numbers meaningless?

No, but it does define what the number is for. A Scope 3 figure is a measure of the emissions associated with your value chain and the leverage you hold over them. It is not a share of a global carbon budget and was never designed to be one. Used as intended, it is decision-useful: it tells you that purchased goods, cloud services or freight dominates your footprint, and it tells you which suppliers to talk to first. Used as an accounting identity that should sum neatly across the economy, it fails, because that is not the job it was built for.

A practical example most companies share: the emissions behind your cloud and SaaS spend are your provider's Scope 1 and 2 and your Scope 3 category 1, at the same time. Both parties reporting it is what makes provider carbon dashboards worth publishing in the first place.

Practically, the way to stay out of trouble is to be explicit. Say which method produced each category, say where boundaries overlap, and never claim a reduction someone else also claims. Our carbon accounting software keeps the method and source attached to every line, so when you replace a spend estimate with supplier data the old figure is retired rather than quietly added alongside it. For the standard itself, see the GHG Protocol Corporate Standard explained, and for the full category list, the 15 Scope 3 categories.