Decarbonisation and MACC Concept Card Template

Decarbonisation and MACC Concept Card: [Your name, company or case, and date]

An elective-level concept card for the economics of cutting greenhouse gas emissions: the scope 1, 2 and 3 baseline that every plan starts from, the marginal abatement cost curve (MACC) that ranks the measures by cost per tonne, the net present value (NPV) of each measure with and without an internal carbon price, and the choices that make a MACC honest or misleading (sequence, interactions, the scope 2 method and the quality of scope 3 data). It builds on mba/mba-stakeholder-esg-card, which teaches how the inventory is built, and on mba/mba-npv-irr-card, and does not re-teach either. It is for MBA students in a sustainability, energy or corporate finance elective and for managers who must defend a transition plan in front of a board or investors. Worked example A builds a nine-measure MACC for Northfield Outdoor Co. from the inventory in the stakeholder and ESG card, Worked example B is a half-solved MACC for Tallow Creek Medical Center, and Exercise C asks for a least-cost target plan and break-even carbon prices. Source basis: the GHG Protocol Corporate Standard (revised edition, 2004, updated by the Scope 2 Guidance, 2015) and the Corporate Value Chain (Scope 3) Standard (2011); IFRS S2 Climate-related Disclosures (ISSB, June 2023) with the December 2025 Amendments to Greenhouse Gas Emissions Disclosures; Weitzman (1974), Kesicki and Ekins (2012) and Gillingham and Stock (2018), written up in original words. Every company, measure, price and emission factor is fictional or illustrative, and every number was computed in Python. Replace every [bracketed] value and every row marked Example. This is a learning aid, not investment, engineering, legal, tax or reporting advice.

Document control

Contents

• When to use it
• The idea in plain language
• Formulas and notation
• Standards and reporting (dated facts)
• Assumptions and limits
• How companies use the MACC and an internal carbon price
• The build step by step
• Worked example A (fully solved): Northfield Outdoor Co.
• Worked example B (half solved): Tallow Creek Medical Center
• Exercise C (unsolved): a least-cost target and break-even prices
• Connections to other cards
• Common mistakes and misconceptions
• Self-test questions
• My notes and next steps
• Sources and further reading
• Answer key

When to use it

Use the tools on this card when:

• a board, an investor or a lender asks for a credible plan behind an emissions target, and the plan must show what each tonne costs and which measures pay for themselves;
• a sustainability team has a long list of measures (efficiency, fuel switching, renewable electricity, supplier programmes, product changes) and capital is limited, so the measures must be ranked;
• finance must decide whether to use an internal carbon price, at what level, and how it changes the hurdle a low-carbon project must clear;
• a company must disclose under IFRS S2 or a similar regime how it uses carbon prices in decisions and how its transition plan is resourced;
• a case asks whether a company's net-zero claim is backed by real abatement or mainly by contractual instruments.

Do not use a MACC to decide whether to act at all on climate; that is a strategy and governance question (see mba/mba-sustainability-strategy-guide and mba/mba-corporate-governance-card). Do not use it as the only test for a large, lumpy or irreversible investment either: a heat pump for one building fits a MACC, a new plant design needs a full NPV with scenarios and real options (mba/mba-options-real-options-card).

The idea in plain language

Every decarbonisation plan starts from a baseline: the company's emissions in a reference year, split into scope 1 (direct emissions from sources it owns or controls), scope 2 (emissions from the electricity, heat and steam it buys) and scope 3 (all other emissions in its value chain, in the 15 categories of the Scope 3 Standard). The baseline tells you where the tonnes are. For a retailer or a brand, most of them are in scope 3, in the goods it buys; for a hospital or a factory, scopes 1 and 2 matter more.

A measure is anything that reduces those emissions: an LED retrofit, an electric van, a renewable electricity contract, a programme that helps suppliers cut their own energy use. Each measure has a cost (capital spent now, operating costs or savings each year) and an abatement (tonnes of CO2e avoided each year). Dividing the annualised net cost by the annual abatement gives the cost per tonne: the marginal abatement cost of that measure. Some measures have a negative cost per tonne because the energy they save is worth more than they cost; they pay for themselves even without any value on carbon.

The MACC sorts the measures from cheapest to most expensive per tonne and draws each one as a bar. The width of a bar is the tonnes it abates, the height is its cost per tonne. Reading from left to right, you see how many tonnes the company can cut at negative cost, how many more it can cut below any given price, and which last tonnes are very expensive. The area of each bar is its annual cost.

The internal carbon price links the MACC to investment decisions. If the company values each tonne at a price $P$, a measure with a cost per tonne below $P$ has a positive NPV once the avoided tonnes are counted. The break-even carbon price of a measure is exactly its cost per tonne. So a carbon price of USD 100 per tonne says: do every measure to the left of the point where the curve crosses 100.

Economists have argued about prices against quantities for half a century. Weitzman (1974) showed that a price instrument (a tax or internal price) and a quantity instrument (a cap or a target) give different results when costs are uncertain; companies often use both, a target for the total and a price to choose the measures. Kesicki and Ekins (2012) list the weaknesses of MACCs: they ignore interactions between measures unless built carefully, they hide uncertainty in single numbers, they leave out non-financial barriers and they are static. Gillingham and Stock (2018) add that static cost per tonne can mislead for technologies whose costs fall as they spread, where an early, expensive tonne buys learning that lowers the cost of later tonnes. A good MACC is therefore a starting point for a discussion, not an answer.

Formulas and notation

Capital recovery factor and equivalent annual cost (EAC) of the capital spent on measure $i$:

Marginal abatement cost of measure $i$, with annual change in operating cost $\Delta O_i$ (negative when the measure saves money) and annual abatement $A_i$:

NPV of measure $i$ at an internal carbon price $P$ per tonne, with annuity factor $AF(r, n)$:

Break-even carbon price (set $NPV_i = 0$ and use $AF = 1/CRF$):

Abatement from an activity change, with activity $Q$ and emission factor $f$ before and after:

Excel: with the capital in B2, life in C2, annual change in operating cost in D2 (USD thousand) and abatement in E2 (tCO2e), the equivalent annual cost is =PMT(8%,C2,-B2), the cost per tonne in USD is =(PMT(8%,C2,-B2)+D2)*1000/E2, and the NPV at a carbon price in G1 is =-B2+PV(8%,C2,D2-G1*E2/1000). Sort the table by cost per tonne and add a running total of abatement for the curve.

Standards and reporting (dated facts)

The GHG Protocol has been consulting on revisions to parts of its standards; check ghgprotocol.org for the status of any revised scope 2 or corporate guidance before you build a baseline for a filing. How IFRS S2 and the EU rules apply in a given jurisdiction is set out in mba/mba-stakeholder-esg-card and finance/issb-sustainability-disclosure. This section lists dated facts only; it is not reporting or legal advice. Check the current official text and a qualified professional before you rely on it.

Assumptions and limits

• Level annual costs and abatement. The card uses a constant annual saving and constant abatement over the life of each measure, so the cost per tonne equals the break-even carbon price. With rising energy prices or a decarbonising grid, compute the NPV year by year and divide by discounted tonnes instead.
• Fixed implementation sequence. Abatement is computed in the order the company would implement the measures: demand reduction and electrification first, the renewable electricity contract last, covering all remaining residual-mix load. The sorting by cost is for display only. Another sequence gives other bar widths (Step 4).
• Market-based scope 2. Electricity abatement is measured with the residual-mix factor of the stakeholder and ESG card (0.450 kgCO2e per kWh). Under the location-based method (0.380), the renewable contract abates nothing and the efficiency measures abate less. Show both when you present the plan.
• Scope 3 abatement needs better data. Northfield's purchased goods are estimated with a spend-based factor. A supplier programme that cuts a supplier's emissions does not change a spend-based estimate at all; only supplier- specific or hybrid data will show it. Measures M7 to M9 assume that data exists.
• No financing, tax or subsidy effects. Grants, tax credits, accelerated depreciation and green financing lower the cost of many measures; add them as separate lines with their source and date.
• Single discount rate. All measures use 8%. A riskier measure (an unproven technology, a supplier programme with uncertain uptake) deserves a higher rate or a scenario range.
• No offsets or removals. Carbon credits and removals are outside the MACC of the company's own emissions; report them separately and do not net them against the baseline.

How companies use the MACC and an internal carbon price

Companies use internal carbon prices in three main forms. A shadow price is added to the cash flows of a project appraisal only; no money changes hands, but projects that emit more look worse. An internal fee charges business units for their emissions and funds a pool for low-carbon projects; it changes behaviour because it hits budgets. An implicit price is the cost per tonne of what the company actually did, computed after the fact; it tests whether stated prices are applied. A shadow price is easy to set high and ignore; a fee is harder to set but harder to ignore. IFRS S2 asks whether and how an entity applies a carbon price in decisions, so an unused shadow price becomes visible.

The build step by step

1. Baseline. Take the inventory with both scope 2 methods and scope 3 by category. Record factors and their dates.
2. Long list. Collect measures from operations, facilities, logistics, procurement and product teams; give each an owner and an ID.
3. Inputs. For each measure, record incremental capital, life, annual operating change and the activity change (kWh, litres, tonne-km, spend) with the factors before and after.
4. Sequence. Decide the order in which measures would be implemented, compute each abatement against the emissions left after the earlier ones, and check that the total equals baseline minus the planned end state.
5. Cost per tonne. Annualise capital with the capital recovery factor, add the operating change, divide by tonnes.
6. Curve. Sort and draw, with bar width equal to abatement.
7. Carbon price. Compute NPV at zero and at the internal price; list the break-even price of each measure.
8. Target check. Compare cumulative abatement with the target; the remainder is residual emissions to explain.

Worked example A (fully solved): Northfield Outdoor Co.

Example situation: Northfield Outdoor Co., the fictional outdoor-gear retailer of mba/mba-stakeholder-esg-card, has a market-based baseline of 99,494.4 tCO2e: scope 1 926.4 (natural gas 338.6, fleet diesel 562.8, refrigerant 25.1), scope 2 market-based 1,728.0 (location-based 2,432.0) and scope 3 96,840.0 in three categories (purchased goods 88,200.0, upstream transport 5,280.0, commuting 3,360.0). The board asks for a list of measures, their cost per tonne and the effect of an internal carbon price of USD 100 per tonne. The discount rate is 8%.

Step 1: measure inputs (Example)

Prices: electricity USD 0.12 per kWh, diesel USD 1.40 per litre, natural gas USD 0.05 per kWh. Factors as in the stakeholder and ESG card: residual-mix electricity 0.450, diesel 2.68 per litre, natural gas 0.183 per kWh, refrigerant blend 2,088 per kg (all kgCO2e).

Abatement formulas: M1 =1200000*0.45/1000; M2 =14400*2.68/1000; M3 =(90000*2.68-260000*0.45)/1000; M4 =12*(2088-150)/1000; M5 =(1000000*0.183-333333*0.45)/1000 (the heat pump delivers three units of heat per unit of electricity); M6 =(3840000-1200000+260000+333333)*0.45/1000; M7 =5280*25%; M8 =88200*4%; M9 =40000000*0.42*30%/1000.

Step 2: cost per tonne and NPV (USD thousand unless stated)

Operating changes: M1 saves =1200000*0.12/1000 = 144.00; M2 saves 20.16 of diesel and pays a 12.00 software fee; M3 saves 126.00 of diesel and 15.00 of maintenance and buys 31.20 of electricity; M5 saves 50.00 of gas and buys 40.00 of electricity; M6 pays a premium of USD 0.012 per kWh, =3233333*0.012/1000 = 38.80; M7 saves 150.00 of freight and adds 90.00 of inventory carrying cost; M8 is a co-funded programme of 400.00 a year; M9 is a 3% premium on 40,000 of fabric spend.

Check M1 in Excel: EAC =PMT(8%,10,-900) = 134.13; cost per tonne =(134.13-144)*1000/540 = -18.28; NPV at USD 100 =-900+PV(8%,10,-(144+100*540/1000)) = 428.60.

Step 3: the curve

Sorted by cost per tonne, with the running total of abatement:

Reading: two measures (M7 and M1) pay for themselves and abate 1,860.00 tonnes; at USD 100 a third (M6) joins, for 3,315.00 tonnes. Everything the company can do on this list abates 12,102.05 tonnes, 12.2% of the market-based baseline (=12102.05/99494.4), for an annual net cost of 1,776.04 thousand. Scopes 1 and 2 fall by 2,214.05 tonnes, 83.4% of their baseline of 2,654.40, leaving 440.35 tonnes of residual emissions: gas heating in the stores (155.55), the long-haul diesel still burned after route optimisation (283.01) and refrigerant leakage at the new factor (1.80), with 0.01 of rounding carried from the inventory. The large scope 3 tonnes sit in M8 and M9, which cost more than the internal price; the decision is whether to raise the price for supplier measures, to fund them as strategy (brand, regulation, customer demand) or to accept a slower path.

Step 4: sequence and interactions

The curve looks as if heat pumps (M5) and electric vans (M3) are the worst measures. That is partly a sequence artefact. Both were computed before the renewable contract, so the electricity they add is counted at the residual-mix factor. If the renewable contract were signed first, the electricity they use would be zero-emission under the market-based method:

The total does not change: the tonnes move from M6 to M3 and M5. That is the rule to remember. Bars in a MACC can be added only if every abatement is computed against what is left after the measures before it. A MACC built from stand-alone estimates double-counts (here, the electricity of M3 and M5 would be both added by them and removed by M6), and its total can exceed the baseline. Under the location-based method the picture changes again: M6 abates nothing, M1 abates 456.00 (=1200000*0.38/1000), and electrification is judged on the grid average.

Step 5: an internal fee on top of the shadow price

The board also considers an internal fee of USD 25 per tonne on scopes 1 and 2 (market-based), charged to business units and paid into a low-carbon fund. On the baseline of 2,654.40 tonnes the fund collects 66.36 thousand a year (=25*2654.4/1000). At USD 25 the renewable contract still has a negative NPV (-16.27), but at the shadow price of USD 100 used for appraisal it is clearly positive. The fee and the shadow price do different jobs: the fee changes budgets and funds projects, the shadow price ranks them.

Worked example B (half solved): Tallow Creek Medical Center

Example situation: Tallow Creek Medical Center, the fictional 380-bed not-for-profit regional hospital of the MBA library, considers three measures. Electricity costs USD 0.13 per kWh and the market-based factor is 0.40 kgCO2e per kWh.

Questions to answer:

1. Rank the three measures. Which pay for themselves without a carbon price? [ ]
2. Compute the NPV of B3 at a carbon price of USD 50. Hint: =-2400+PV(8%,25,-(175+50*600/1000)). [ ]
3. B1 is small in capital but large in tonnes. Why do anaesthetic gases matter so much in a hospital inventory, and what non-financial barrier could slow the measure? [ ]

Exercise C (unsolved): a least-cost target and break-even prices

C1. What internal carbon price makes the heat pumps (M5) break even under the sequence of Worked example A, and what price if they run on renewable electricity? What does the gap tell the board about the order of the plan?

C2. The board sets an Example target of a 50% cut in scopes 1 and 2 (market-based) against the baseline of 2,654.40 tonnes, at the lowest annual cost, without M3 and M5. Which measures do you choose, what is the abatement and what is the annual net cost? Remember that M6 must be recomputed: without M3 and M5, the residual-mix load is 2,640,000 kWh.

C3. In one paragraph, explain why a plan whose cuts in scopes 1 and 2 come mainly from a renewable electricity contract can be both correct under the GHG Protocol and weak as a transition story, and what you would show next to it.

Connections to other cards

Common mistakes and misconceptions

1. Adding stand-alone abatement estimates. Interacting measures counted separately double-count; the total can exceed the baseline. Fix the sequence first.
2. Using total instead of incremental capital. An electric van replacing a van that would be bought anyway costs only the premium over the diesel van.
3. Forgetting the added electricity of electrification. Heat pumps and electric vehicles move emissions from scope 1 to scope 2; abatement is the difference.
4. Reporting only the market-based scope 2. The GHG Protocol requires both methods; a plan that only works on the market-based method should say so.
5. Expecting a supplier programme to move a spend-based inventory. Spend factors do not respond to supplier improvements; plan the data change with the measure.
6. Treating a negative cost per tonne as free money. Barriers (split incentives between landlord and tenant, management attention, disruption, hidden costs) explain why such measures are not already done.
7. Mixing annual and lifetime figures. Cost per tonne is annual cost over annual tonnes, or lifetime cost over lifetime tonnes, never one over the other.
8. Ignoring the sign convention. A saving is a negative operating change; a sign error turns the best measure into the worst.
9. Setting an internal carbon price nobody uses. A shadow price that no decision applies is a disclosure risk, not a tool.
10. Netting offsets against the baseline. Credits and removals are reported separately from the company's own abatement.
11. Reading a static MACC as permanent. Costs fall with scale and learning; refresh the curve each year.

Self-test questions

1. Define scopes 1, 2 and 3 in one sentence each.
2. What does the width of a MACC bar show, and what does its area show?
3. Show that the break-even carbon price of a measure with level costs equals its cost per tonne.
4. Why does M1 have a negative cost per tonne, and what is its NPV at zero carbon price?
5. Which three measures clear an internal carbon price of USD 100 in Worked example A, and how many tonnes do they abate?
6. Why does the heat pump's cost per tonne fall from 1,998.16 to 360.32 when it runs on renewable electricity?
7. What happens to M6's abatement under the location-based method, and why?
8. Name two differences between a shadow price and an internal fee.
9. Why might a supplier energy programme not appear in Northfield's reported scope 3, even if it works?
10. What did Weitzman (1974) show about prices against quantities, in one sentence?
11. Name two limits of MACCs from Kesicki and Ekins (2012).
12. What is the capital recovery factor at 8% over 10 years, and what is the EAC of 900?

My notes and next steps

Sources and further reading

Primary works:

• Weitzman, M. L. (1974). Prices vs. quantities. The Review of Economic Studies, 41(4), 477-491. https://doi.org/10.2307/2296698
• Kesicki, F. and Ekins, P. (2012). Marginal abatement cost curves: a call for caution. Climate Policy, 12(2), 219-236. https://doi.org/10.1080/14693062.2011.582347
• Gillingham, K. and Stock, J. H. (2018). The cost of reducing greenhouse gas emissions. Journal of Economic Perspectives, 32(4), 53-72. https://doi.org/10.1257/jep.32.4.53

Official pages for dated facts (checked 2026-10-30):

• GHG Protocol, Corporate Accounting and Reporting Standard (revised edition), https://ghgprotocol.org/corporate-standard
• GHG Protocol, Corporate Value Chain (Scope 3) Standard (2011), https://ghgprotocol.org/corporate-value-chain-scope-3-standard
• IFRS Foundation, IFRS S2 Climate-related Disclosures (June 2023), https://www.ifrs.org/issued-standards/ifrs-sustainability-standards-navigator/ifrs-s2-climate-related-disclosures/
• IFRS Foundation, Amendments to Greenhouse Gas Emissions Disclosures (Amendments to IFRS S2), December 2025, https://www.ifrs.org/projects/completed-projects/2025/amendments-to-disclosure-of-greenhouse-gas-emissions-s2/

Further reading (by name only): the global greenhouse gas abatement cost curve published by McKinsey & Company (Pathways to a Low-Carbon Economy, 2009), which made the MACC format widely known; Berk and DeMarzo, Corporate Finance, for annuities and NPV.

Answer key

Worked example B: B1 annual net cost =10.02-12 = -1.98, cost per tonne -6.39; B2 operating change =-700000*0.13/1000 = -91.00, abatement 280.00, annual net cost -16.49; B3 operating change -175.00, EAC =PMT(8%,25,-2400) = 224.83, annual net cost 49.83. Ranking: B2 (-58.88), B1 (-6.39), B3 (83.05); B1 and B2 pay for themselves. NPV of B3 at USD 50: -211.67 (at zero, -531.91), so B3 needs a carbon price above 83.05 or a subsidy. Anaesthetic gases are potent greenhouse gases released directly to air, so a small volume carries many tonnes; the barriers are clinical habit, training, the supply of equipment and the need for clinical leadership to own the change.

Exercise C:

• C1: the break-even price equals the cost per tonne: USD 1,998.16 under the sequence of Worked example A, USD 360.32 on renewable electricity. The gap says to contract renewable electricity before or with electrification, and to judge electrification on the electricity it will actually use.
• C2: the target needs 1,327.20 tonnes (=50%*2654.4). M1 (540.00, cost per tonne -18.28) plus M6 recomputed for 2,640,000 kWh (abatement 1,188.00, annual premium 31.68, cost per tonne 26.67) give 1,728.00 tonnes, a 65.1% cut, for an annual net cost of 21.81 thousand (=-9.87+31.68). M2 and M4 are not needed.
• C3: under the market-based method, contractual instruments that meet the Scope 2 quality criteria legitimately cut scope 2. But the location-based figure, the physical grid and scope 1 do not change, and investors increasingly ask how much of a plan is own operational change. Show both scope 2 methods, the share of abatement from contracts and the measures that cut scope 1 and scope 3.

Self-test:

1. Scope 1: direct emissions from owned or controlled sources. Scope 2: emissions from purchased electricity, heat, steam and cooling. Scope 3: all other indirect emissions in the value chain, in 15 categories.
2. Width: annual abatement in tonnes. Area: annual net cost (cost per tonne times tonnes).
3. Set $NPV = -K + (-\Delta O + P \cdot A) \cdot AF = 0$, divide by $AF$ and use $1/AF = CRF$: $P = (K \cdot CRF + \Delta O)/A$.
4. The electricity saving (144.00 a year) exceeds the equivalent annual cost of the capital (134.13); NPV 66.25.
5. M7, M1 and M6, for 3,315.00 tonnes.
6. The electricity it uses no longer carries the residual-mix factor, so abatement rises from 33.00 to 183.00 tonnes while the cost stays the same.
7. It falls to zero: the location-based method uses the grid average regardless of contracts.
8. A shadow price changes appraisals only and moves no money; a fee charges business units and funds projects, and so affects budgets.
9. A spend-based factor multiplies spend by an industry average, so a cleaner supplier with the same prices shows no change; supplier-specific data is needed.
10. When costs are uncertain, a price and a quantity instrument lead to different outcomes, and which is better depends on the relative slopes of marginal costs and marginal benefits.
11. Any two of: interactions ignored, uncertainty hidden, non-financial barriers left out, static view of costs, inconsistent baselines.
12. =8%/(1-1.08^-10) = 0.1490; EAC =PMT(8%,10,-900) = 134.13.

Revision history

Standard and licence note

Based on the GHG Protocol Corporate Accounting and Reporting Standard (revised edition, with the 2015 Scope 2 Guidance), the GHG Protocol Corporate Value Chain (Scope 3) Standard (2011), IFRS S2 Climate-related Disclosures (June 2023) and its December 2025 Amendments to Greenhouse Gas Emissions Disclosures, and the primary works listed above, with dated facts checked on 2026-10-30 on the official pages named in the sources. The text, the model, the numbers and the exercises are original; no standard text, consultancy material or published case is copied. Northfield Outdoor Co., Tallow Creek Medical Center and every price, factor and measure in this card are fictional or illustrative. The template is independent and not affiliated with or endorsed by the GHG Protocol, the IFRS Foundation or any company. Check the current official text of each standard and a qualified professional before you rely on it. Template licence: MIT.