Physical or monetary method: choose per emission source, not per company

Negotiate a 10% discount on your steel and, in a footprint built on spend, your emissions fall by 10%. You bought the same tonne, from the same supplier, made the same way.

That arithmetic is the entire subject. Behind every carbon footprint sits one multiplication repeated a few hundred times, activity data times emission factor, and the only real question is what kind of activity data you feed it. Treating that as a company-level decision is the classic first-year mistake. It is a line-by-line decision, and the footprints worth reading mix both methods deliberately.

What each method actually does

The physical method starts from a quantity: liters of diesel, kilowatt-hours of electricity, tonne-kilometers of freight, kilograms of aluminum. Each quantity is multiplied by a physical emission factor, the kilograms of CO2-equivalent per unit, drawn from a recognized database.

The monetary method starts from spend: euros of IT services, euros of office furniture, euros of subcontracted machining. Each amount is multiplied by a monetary factor, the kilograms of CO2-equivalent per euro spent in that category, derived from national economic statistics.

So the physical method describes what you actually consumed, and the monetary method describes what an average euro in that category consumed across the whole economy. Every trade-off between them follows from that one difference.

Where each is the honest choice

Physical, wherever a quantity exists. Energy, fleet fuel, refrigerants, freight, waste, water, and any purchased material you buy by weight or volume. The data is in invoices and statements, the factors are well established, and the result responds to real changes: switch a route from road to rail and the number moves.

Monetary, where only spend exists. Services, IT, marketing, insurance, maintenance contracts, and the long tail of purchases too diverse to weigh. The accounting export is the only complete data source you have for these, and a monetary estimate covers them all at once, which is worth a great deal in year one.

The weakness deserves stating plainly, beyond the discount example. Buy the same steel from a low-carbon producer and a monetary line does not move. Inflation pushes it the wrong way while nothing physical changes. A footprint built entirely on monetary factors is a spend analysis wearing a carbon costume, and it will report progress in the years your procurement team negotiates well and regression in the years it does not.

Good to know: the two methods also fail differently. Physical errors are usually small and traceable, a missed site, a unit mix-up. Monetary errors are structural: the factor is an economy-wide average that may describe your specific supplier badly, in either direction. Treat monetary lines as estimates with wide error bars, and say so in the report rather than letting a precise-looking total imply otherwise.

The databases behind the factors

A factor is only as good as its source, and the sources worth naming are few.

• Base Empreinte (ADEME) is the French public reference, born from the merger of Base Carbone and Base IMPACTS. It carries both physical and monetary factors, adapted to the French energy mix.
• Ecoinvent is the international life cycle inventory database, the usual choice when your supply chain crosses borders and French averages stop being representative.
• Agribalyse (ADEME) covers food and agricultural products, essential for anyone with a canteen or an agri-food supply chain.

Whichever tool or consultant you use, ask which database each factor came from and which year it was updated. Factors get revised as energy mixes change, so a footprint computed on stale factors drifts without anyone touching the data.

A ratchet, not a religion

The policy fits in one sentence: physical data wherever a quantity exists, monetary factors for the rest, and every year convert your largest monetary lines to physical or supplier-specific data.

That last clause is the one that does the work. Rank your monetary lines by size, take the top two or three, and ask what quantity data would replace them: tonnages from your main material suppliers, real server counts from your IT provider, actual tonne-kilometers from your forwarder. The companion article on getting carbon data out of your departments covers where those documents live.

So plan the conversion before you publish, and name next year's two lines in this year's report. The direction of travel is one-way: each line moved from monetary to physical makes the footprint more sensitive to real decisions, which is the entire point of measuring. A number that cannot move when you improve is not measuring your improvement.

FAQ

What is the difference between the physical and monetary method in carbon accounting?

The physical method multiplies quantities consumed (liters, kWh, tonnes, kilometers) by an emission factor per unit. The monetary method multiplies spend by an average emission factor per euro for that category. Physical data reflects what you actually consumed; a monetary factor reflects the economy-wide average for that type of spending.

Why not build the whole footprint with monetary factors?

Because a monetary line only responds to spend, not to what you buy. Choosing a low-carbon supplier changes nothing in the result, a discount lowers reported emissions without lowering real ones, and inflation raises them. Monetary factors are legitimate for the long tail of diverse purchases, but the largest lines deserve physical or supplier data.

Which emission factor databases should a French company use?

ADEME's Base Empreinte is the French public reference, formed from the merger of Base Carbone and Base IMPACTS. Ecoinvent is the usual complement for international supply chains, and Agribalyse covers food and agriculture. Always record which database each factor came from and the year it was updated.