In 2021, a major European airline announced that it had become the first in its region to achieve carbon neutrality. The announcement generated considerable press coverage and a round of congratulations from sustainability commentators. The mechanism behind the claim was a portfolio of carbon offset projects — forestry schemes, clean cookstove programmes, and methane capture projects spread across three continents. The airline had purchased certificates representing the carbon absorbed or avoided by these projects, and those certificates, on paper, cancelled out its emissions from flying.
By 2023, an investigation by the Guardian, Zeit, and SourceMaterial had examined the forestry offset projects behind many similar corporate claims. Their findings, drawing on analysis by researchers at Cambridge and other institutions, were striking. Of the projects studied, more than 90 percent were found to have overstated their carbon benefits. Some had no discernible benefit at all. Forests certified as protected were being logged. Calculations of what would have happened to the forest without protection — the counterfactual baseline that gives an offset its value — were based on assumptions that bore little relationship to actual deforestation patterns. The carbon credits were real. The carbon they claimed to represent largely was not.
This investigation was not an isolated exposé. It was the most prominent entry in a growing body of research that has been documenting the systematic weakness of the voluntary carbon offset market for more than a decade. The question the research raises is not simply whether specific projects have been fraudulent. It is whether the offset mechanism, as currently constituted, is capable of delivering the emissions reductions it promises — and if not, what legitimate corporate climate action should look like instead.
The Origins of Carbon Offsetting
The concept of carbon offsetting entered mainstream climate policy through the Kyoto Protocol in 1997. The Clean Development Mechanism established under Kyoto allowed developed-country governments and companies to fund emissions-reduction projects in developing countries and count the resulting reductions against their own targets. The logic was economic efficiency: if it costs less to reduce a tonne of emissions in a developing country than at home, the same global outcome is achieved at lower total cost. The carbon atoms do not know where the reduction happened.
The voluntary carbon market, which operates outside the Kyoto framework and allows companies and individuals to purchase offsets without a regulatory obligation to do so, grew substantially through the 2000s and especially through the 2010s as corporate net-zero commitments became common. By 2021, the voluntary market had reached approximately $2 billion in annual transaction value, with projections suggesting it could reach $50 billion or more by 2030 if corporate climate commitments were to be met through offsetting.
The appeal for corporations is straightforward. Most companies have emissions that are genuinely difficult and expensive to eliminate in the short term — aviation fuel combustion, industrial process heat, supply chain transport — and offsetting provides a mechanism to claim progress on those emissions while the hard work of structural decarbonisation is, in theory, underway. The offset becomes a bridge: a way to make a credible climate claim today while longer-term solutions are developed. In theory, there is nothing wrong with this framing. A well-functioning offset market that delivers verifiable, permanent, additional carbon reductions could play a legitimate supporting role in a genuine decarbonisation strategy.
The problem is that the voluntary carbon market has proven extremely difficult to regulate in ways that ensure those properties hold.
The Measurement Problem
The fundamental challenge in carbon accounting is that you are measuring a counterfactual. A forest offset does not absorb extra carbon by existing — forests absorb carbon regardless of whether they are certified. The offset value is the difference between what the forest absorbed with the project in place and what it would have absorbed without it. This counterfactual — the project's baseline — is necessarily a model, not a measurement. And models require assumptions.
For a forest protection project, the baseline assumption is the rate at which the forest would have been cleared in the absence of the project. In a region with high deforestation pressure and clear economic incentives to clear land, this rate can be estimated with reasonable confidence. In a region where the forest was never at serious risk of clearing — perhaps because it is remote, legally protected, or simply not commercially attractive to loggers — the baseline deforestation rate is low or zero, and a project claiming protection credits is claiming credit for preventing something that was unlikely to happen anyway.
A 2023 study published in Science analysed 26 REDD+ forest protection projects — the most widely used type of forest carbon offset — using satellite imagery and causal inference methods to estimate actual deforestation rates in project areas against matched comparison areas. The study found that actual avoided deforestation was approximately 2.7 million tonnes of carbon over the study period. The credits issued for these projects claimed avoided deforestation of 89 million tonnes. The ratio was not a rounding error. The projects had collectively delivered roughly three percent of the carbon benefit claimed in the credits they sold.
The researchers were careful to note that the projects may have delivered genuine co-benefits — biodiversity conservation, community income, ecosystem service protection — that have real value independent of their carbon claims. The finding was specifically about the carbon accounting, which was disconnected from reality on a scale that cannot be explained by methodology disagreements or measurement uncertainty. It was explained by systematically optimistic baseline assumptions built into a verification system that had structural incentives to approve projects rather than reject them.
The Additionality Question
Additionality is the principle that an offset should only receive credit for emissions reductions that would not have happened in its absence. A renewable energy project in a country where renewable energy was already growing rapidly due to falling costs and policy support is not additional — it would have been built anyway. A forestry project protecting land that was never going to be cleared is not additional. Credits sold for non-additional projects represent, in the language of climate accounting, hot air: a certificate with no underlying physical reality.
Assessing additionality rigorously is difficult because it requires a credible answer to a counterfactual question. Verification bodies — the organisations that certify offset projects — typically assess additionality through a financial analysis showing that the project would not be economically viable without carbon revenue, and through a barriers analysis showing that the project overcomes non-financial obstacles that would otherwise prevent it. Both methods are vulnerable to strategic manipulation by project developers who have an obvious financial interest in demonstrating additionality whether it exists or not.
Academic analysis of additionality claims in certified offset markets has consistently found high rates of failure. A 2016 study published by the Stockholm Environment Institute examined projects certified under the Clean Development Mechanism and found that 85 percent of projects and 73 percent of potential credits had low or questionable additionality. More recent analysis of voluntary market standards has found similar patterns, with some researchers arguing that the current structure of financial incentives in the offset market makes systematic over-crediting inevitable regardless of the verification methodology in use.
The implication is uncomfortable for the corporate sustainability industry. Many of the most widely-purchased offset credits — the ones that underpin the net-zero claims of major airlines, consumer goods companies, and financial institutions — have been shown to have additionality problems of the kind that would, if accounted for honestly, substantially or completely eliminate the carbon benefit they claim. Companies that have built their climate communications around offset-based neutrality claims are, in many cases, making claims that the underlying science does not support.
Permanence: When the Forest Burns
Even a project that genuinely reduces or avoids emissions at the time of certification faces a third challenge: permanence. Carbon stored in a forest can be released at any time if the forest burns, is logged, or dies due to drought or disease. Carbon stored in soil can be released by changes in land management. Emissions avoided by a clean cookstove project are not stored anywhere — they simply did not occur, and there is no physical reservoir to protect.
The permanence problem for forest offsets became vivid in 2021 when California's buffer pool — a reserve of offset credits set aside to cover losses from wildfires and other reversals — was revealed to be dramatically undersized. The buffer pool had been calculated on historical fire frequency and severity. As climate change accelerated wildfire activity in the western United States, forests that had been certified as long-term carbon stores began burning at rates that the buffer calculations had not anticipated. Studies by CarbonPlan estimated that fires in 2020 and 2021 alone had consumed a significant fraction of the entire California buffer pool, representing a permanent loss of the carbon previously attributed to those forests.
This is not a failure of the specific projects involved. It is a structural problem with the attempt to treat forest carbon as a long-duration store equivalent to the long-duration storage that occurs when fossil carbon remains underground. Fossil carbon has been sequestered for millions of years. Forest carbon cycles on timescales of decades. Using forest carbon to offset the release of fossil carbon is, in the long-term carbon accounting, not an exchange of equivalents. It is borrowing against a volatile asset and treating the loan as repaid.
How Offsetting Enables Structural Inaction
The technical problems with offset quality would matter less if offsetting were genuinely functioning as a bridge — a temporary mechanism giving companies time to make structural changes — rather than as a permanent substitute for those changes. The evidence that it is functioning as the latter is substantial.
Analysis of corporate net-zero commitments by the NewClimate Institute and Carbon Market Watch, published in their 2022 Corporate Climate Responsibility Monitor, found that the net-zero pledges of major corporations relied heavily on offsetting and forest sinks to achieve their stated targets, often while projecting little or no reduction in absolute operational emissions. In some cases, companies were projecting absolute emissions growth through 2030 while simultaneously claiming to be on track for net-zero. The gap between the headline commitment and the underlying operational trajectory was bridged entirely by offset purchases and accounting assumptions about forest sequestration.
The report examined 25 major companies across sectors including consumer goods, retail, automotive, and financial services. It found that their combined pledges, if achieved as stated, would reduce real-world emissions by only about 40 percent of what a genuine net-zero trajectory would require. The remaining 60 percent was attributed to offsets and carbon removals whose quality and deliverability were, in the researchers' assessment, highly uncertain. The headline "net zero by 2050" commitments, in other words, were not what they appeared to be to a reader who did not examine the methodology behind them.
This matters for reasons beyond individual corporate reputation. Capital allocation follows climate commitments. Investors, lenders, and insurers are increasingly pricing climate risk and climate performance into their decisions. If the performance data underlying those decisions is inflated by offset accounting that overstates real reductions, the capital is flowing toward the appearance of decarbonisation rather than the substance of it. The financial system is, in effect, rewarding companies for purchasing certificates rather than for changing their operations.
What the Science Says We Actually Need
The scientific consensus on what is required to limit warming to 1.5 degrees Celsius above pre-industrial levels — the threshold identified in the Paris Agreement as the boundary above which climate impacts become substantially more severe — is unambiguous about the primary mechanism: rapid, deep, and absolute reduction in the emission of greenhouse gases from fossil fuel combustion and industrial processes. The IPCC Sixth Assessment Report is direct on this point. Achieving 1.5 degrees requires global emissions to be roughly halved by 2030 relative to 2019 levels and reach net zero around 2050.
Carbon dioxide removal — including natural sinks like forests and soils, and engineered approaches like direct air capture — has a role to play in this pathway, but primarily in the second half of the century and primarily to address residual emissions from sectors where full elimination is technically or economically impossible by 2050. The science does not support using carbon removal as a primary mechanism for offsetting emissions from sectors where reductions are technically feasible today. It supports using removal as a limited, late-stage complement to deep emissions cuts, not as a substitute for them.
This framing has direct implications for how corporate climate commitments should be structured. A commitment to reach net-zero operational emissions by 2040, with interim targets requiring absolute emissions reductions of 50 percent by 2030, measured against a 2019 baseline, with offset use limited to genuinely residual emissions in hard-to-abate sectors — this is a structurally different commitment from one that projects minimal operational change and relies on offset purchases to bridge the gap to a net-zero label. The first is consistent with what the science requires. The second is not, regardless of the certification quality of the offsets involved.
What Better Corporate Climate Commitments Look Like
The Science Based Targets initiative provides the most widely-adopted framework for corporate climate commitments that are aligned with the Paris Agreement. SBTi-validated targets require companies to set emissions reduction trajectories consistent with 1.5-degree pathways, with specific requirements around the proportion of emissions covered, the baseline year, and the interim milestones. Critically, SBTi requires that near-term targets — those covering the period to 2030 — be met through genuine emissions reductions rather than offsetting. Offsets can be used to support beyond-value-chain claims but cannot substitute for operational reductions in the primary target.
Not all SBTi-validated targets are created equal, and the framework continues to evolve as scientific understanding develops and methodological gaps are identified. But the SBTi approach represents a meaningful improvement over offset-based net-zero claims precisely because it forces companies to engage with their actual emissions trajectory rather than their offset purchasing budget.
Beyond target-setting, credible corporate climate action has several distinguishing characteristics. It prioritises Scope 1 and Scope 2 emissions reductions — direct operational emissions and purchased energy — where companies have the most direct control and where reductions are most verifiable. It develops a credible plan for Scope 3 emissions — supply chain and value chain emissions, which typically represent 70 to 90 percent of a large company's total footprint — that goes beyond supplier surveys to include procurement requirements, supplier development programmes, and product design changes that reduce embodied emissions. It discloses absolute emissions data annually in a format that allows third-party verification. And it is honest about the portions of its footprint where reductions are not yet technically or economically feasible, rather than obscuring those portions behind offset accounting.
The Role of Policy and Individual Action
It would be misleading to suggest that the corporate climate problem is solvable through voluntary commitments alone. The structural incentives that make genuine decarbonisation expensive and offsetting cheap are products of policy choices — the absence of a sufficiently high carbon price, the treatment of forest carbon as equivalent to fossil carbon in accounting standards, the self-regulatory nature of the voluntary offset market — that individual companies cannot unilaterally change.
Policy reform is therefore the complement to better corporate practice, not a substitute for it. A meaningful price on carbon that reflects the true cost of emissions, robust accounting standards that distinguish real reductions from accounting offsets, regulated minimum quality standards for voluntary offset markets, and mandatory climate disclosure requirements that force accurate reporting — these are the systemic conditions under which genuine corporate decarbonisation becomes economically rational rather than merely ethical.
For individuals navigating this landscape, the most important insight is probably the simplest: be sceptical of any climate claim that relies primarily on offsetting rather than reduction. A company that has reduced its absolute emissions by 40 percent and offset the remaining 10 percent has done something meaningfully different from a company that has maintained its emissions and purchased enough credits to call itself neutral. The mathematics of the certificate are the same. The contribution to atmospheric carbon concentration is not.
Individual offset purchases — for flights, for personal carbon footprints — face the same structural quality problems as corporate purchases. The best available options tend to be those focused on carbon dioxide removal rather than avoidance, with long-term permanence guarantees and independent scientific verification. Biochar, enhanced weathering, and direct air capture, while currently expensive, store carbon in forms that are genuinely durable rather than forest-dependent. For individuals who want to act beyond consumption reduction, these represent more credible uses of money than the cheap forestry credits that dominate most consumer offset platforms.
Moving Beyond the Shortcut
The carbon offset has always been an appealing idea: a mechanism that lets progress happen where it is cheapest while those with harder problems buy time to solve them. In a functioning market with rigorous standards and honest accounting, it might even work as intended. The accumulated evidence of the past decade suggests that the voluntary carbon market, in its current form, has not been that mechanism. It has been a mechanism for generating legitimate-sounding claims without the commensurate physical reality.
That conclusion does not mean that all climate progress outside direct emissions reduction is worthless. Forest protection, clean cooking access, and renewable energy deployment in developing economies have genuine value for biodiversity, human health, and energy poverty — value that exists independently of their carbon accounting. The error is treating that value as equivalent to a tonne of avoided fossil carbon, and using that equivalence to delay or replace the structural changes that the physics of the climate system actually requires.
The companies that will have credible climate stories to tell in ten years are the ones that are changing their operations today — switching to renewable energy, redesigning supply chains, investing in electrification, rethinking products whose embedded emissions cannot be eliminated. These changes are harder than purchasing a certificate. They are also the only changes that move the needle on the number that matters, which is the concentration of carbon dioxide in the atmosphere. That number does not respond to accounting. It responds to physics. And physics, unlike corporate communications, does not allow for the purchase of a shortcut.