Carbon offsets explained: The science behind why offsetting does not reduce emissions

Key points:

• Carbon offsetting refers to a country, company or individual using carbon credits to “cancel out” their emissions. A carbon credit theoretically represents one tonne of CO2 or equivalent gas emissions avoided, reduced or removed elsewhere. 
• Supporters of carbon offsets see them as an cost-effecitve way for countries and businesses to reach their climate targets by paying for emissions to be avoided in cheaper places, whilst continuing to emit themselves.
• However, by design, carbon offsets do not cause a net reduction in emissions. Their use is not aligned with IPCC guidance on limiting warming to 1.5°C or 2°C and further delays emission cuts.
• Academic research and reputable investigations have found that carbon offsetting projects have not delivered emissions reductions. In fact, carbon offsetting has caused an overall increase in emissions. 
• This is because most carbon offsets come from projects that aim to reduce or avoid emissions compared to what would happen if no project were put in place, which is very difficult to prove. As a result, projects often overstate the quantity of offsets.
• Carbon offsetting assumes that all carbon is equal and that one tonne of carbon emissions removed or avoided is equivalent to one tonne of carbon emitted. However, earth systems research has shown this is not the case. For example, carbon stored in nature has a much higher risk of leaking back into the atmosphere than if it had stayed in the ground as an unburned fossil fuel.
• The IPCC outlines a theoretical need for durable carbon removals to offset a very small amount of residual emissions after we reach net zero. However, carbon removal faces significant feasibility challenges and cannot be relied on for meeting climate targets. 
• Carbon offset schemes have often been promoted as a means to increase climate finance, but projects often provide little funding for implementers. 
• Investigations have linked carbon offset projects to fraud and human rights abuses. As a result, companies face reputational and litigation risks when using carbon offsets.
• Although a popular concept, carbon avoidance and reduction offsets have no place in genuine and ambitious climate policies or corporate climate plans. To reach climate goals, mitigation should always be the first priority. 

What are carbon offsets?

Carbon offsetting refers to when a country, company or individual uses a credit to “cancel out” their emissions elsewhere. A carbon credit is theoretically equivalent to one tonne of CO2 or equivalent gas emissions avoided, reduced or removed from the atmosphere.

Carbon credits can be bought and sold on two types of carbon markets: 

• Compliance markets are regulated by governments and involve emissions reductions that are mandated by law. Examples include cap-and-trade or emissions trading schemes (ETS), such as the EU ETS. Some of these compliance schemes allow the restricted use of offsets to meet emissions-reduction obligations. 
• The voluntary carbon market (VCM) is largely unregulated and enables companies or individuals to buy offsets to count toward their emissions cuts, where there is no legal obligation to do so. 

As of the end of 2024, the majority of carbon credits issued were for forestry and land use projects (34.6%) and renewable energy projects (31.8%), followed by household and community projects, such as clean cookstove projects (12.4%).

In 2023, the voluntary carbon market was reported to have a transaction volume equivalent to trading 110.8 million tonnes of CO2 equivalent (CO2e), worth USD 723 million. 

The fundamentals of carbon offsetting do not align with what climate science tells us about reaching net zero

Supporters of carbon offset schemes argue that they are an economically efficient way to meet climate targets, as they enable emissions reductions to occur where they are cheapest. 

For example, countries or companies struggling to meet their climate goals can pay for emissions reductions in places where they are cheaper, rather than reducing their own, more-expensive-to-abate, emissions. Theoretically, this could also channel financing for mitigation towards developing countries, where much of the ‘low-hanging fruit’ for cheap offsetting projects is located.

However, examining the basic principles of carbon offsetting shows that by design, offsetting does not reduce emissions and is, at best ,a zero-sum game. Done properly, offsets merely compensate for emissions in one setting by reducing emissions elsewhere – there is no overall reduction. 

When projects used to offset emissions elsewhere lack environmental integrity (i.e. do not represent real emission reductions), research has shown that they actually lead to an overall increase in emissions. Indeed, peer-reviewed assessments of past carbon offset projects over decades have shown that many projects have failed to deliver the emissions reductions they promised.

The complex way carbon emissions are accounted for means that misalignment between carbon accounting schemes could lead to carbon credits being double-counted. “Double counting” occurs when emission reductions are claimed by the country that developed the credit as well as by the country that purchased it – the reduction is counted twice, even though it only happened once. This is being discussed as part of negotiations for the use of carbon trading under the Paris Agreement.

Offsetting is not a valid solution for reducing global emissions. According to the IPCC, countries and companies need to achieve deep emission reductions now. Keeping warming to 1.5°C requires achieving net zero emissions by 2050, meaning CO2 emissions must fall steeply from the 41.6 billion tonnes estimated in 2024. 

The use of carbon credits not only delays emission cuts, but can cause continued investment into non-1.5°C-aligned technologies and business models, which, according to Climate Analytics, could lock in those practices for decades. The UK Committee on Climate Change, which advises the UK government on climate change policy, writes that the purchase of offsets to enable the use of high-emitting technology “would in the long run lead to higher emissions than if funds used for ‘offsetting’ were used to invest in low-carbon technology.”

Carbon removal will play a very small role in limiting global warming, but scientists say it cannot be relied on for carbon offsetting schemes

As we have already emitted too much, all modelled pathways limiting warming to 1.5°C or 2°C in the most recent IPCC reports include the removal and storage of some level of carbon from the atmosphere – highlighting a theoretical need for carbon removal in the future. However, IPCC scientists warn that the small amount of carbon removal that will be possible must be saved for activities that absolutely cannot be decarbonised. As a result, carbon removals have no place in company strategies as a way to compensate for avoidable emissions.

Alongside emissions reduction and avoidance, a very small proportion of carbon offsetting projects include carbon removals  – around 3% of offsets in the VCM in 2023. Although some see these promised removals as a way to offset their emissions, carbon removal is currently unfeasible at scale and relying on future large-scale carbon removal is incredibly risky and uncertain. Technological approaches to removals are currently unproven at scale and face significant implementation challenges. These approaches are, and will likely remain, very expensive, according to the IPCC. 

The IPCC sets out two purposes for which removals should be reserved: to cancel out the residual emissions of very few ‘hard-to-abate’ activities and to remove some of the excess CO2 we have emitted historically to improve our chances of staying within the carbon budget, which research suggests will be exceeded in 6-10 years.  

Similar technology used predominantly by oil and gas companies to capture carbon from the source of emissions (carbon capture and storage) has been slow to develop, with a history of “unmet expectations”, according to a 2023 International Energy Agency report. Even if it can be made to work at scale, this technology perpetuates the burning of fossil fuels.¹While the technology used is very similar, carbon removal captures differs in that it captures diffuse CO2 emissions from the atmosphere, while carbon capture and storage (CCS) is used to capture emissions from the point source of an emitting process. 

This reiterates the need to implement readily available mitigation approaches and pursue deep, rapid emissions cuts now, rather than relying on carbon offsetting.

Research shows that carbon offsets have actually increased emissions, not reduced them

Each carbon credit purchased by a government or company is theoretically equivalent to preventing one tonne of CO2 emissions, or an equivalent amount of another gas, from being released into the atmosphere. However, academic studies and authoritative investigations have found that the environmental integrity of most carbon credits is low, meaning this is often not the case.

A systematic review of carbon offset projects published by Nature in 2024 looked at a fifth of all carbon offsets issued to date and found that “that less than 16% of the carbon credits issued to the investigated projects constitute real emission reductions.” The remaining 84% did not constitute real emissions reductions – thereby instead resulting in the release of 812 million tonnes of emissions, equivalent to “more than annual German CO2 emissions”.

Another investigation conducted by The Guardian and Corporate Accountability in 2023 found that 39 of the 50 offset projects that have sold the most credits likely did not result in the emissions reductions they claimed, and the others were either problematic or could not be assessed due to lack of information. 

In addition, a 2024 assessment of over 4,000 carbon credit projects conducted by US investment company MSCI rated 47% as low-integrity, meaning they did not meet most environmental and social criteria.

Research on offsets used under compliance mechanisms has shown similar results. A 2015 study by the Stockholm Environment Institute on the environmental integrity of offsets issued under the Kyoto Protocol’s Joint Implementation mechanism suggests that they enabled global GHG emissions to be about 600 million tonnes CO2e higher than they would have been if countries had met their emissions targets domestically.

Similarly, a 2016 study found that 73% of the potential supply of 2013-2020 credits used under the Clean Development Mechanism (CDM) offered a low likelihood that emission reductions are additional and not overestimated. A separate assessment of clean cookstove projects under the CDM, which replaced traditional stoves with more efficient ones, suggests that these may “fail to realize expected carbon reductions or anticipated health and climate cobenefits”.

Government offset schemes have also shown limited emissions reductions. For example, a review of 182 projects focused on native forest regeneration, from which the Australian government has purchased USD 204 million worth of credits, were found to have a negligible impact.

Analysis shows that few carbon offsets are truly additional

For carbon credits to be issued, project developers need to illustrate that the emission reduction promised would not have taken place in the absence of the project, and therefore that the offset is “additional”. However, analysis has shown that true additionality is difficult to verify and often not the case.

Almost all carbon credits issued to date have been for the avoidance or reduction of carbon emissions. Only around 3% of credits in the VCM in 2023 were for the removal of carbon emissions. The additionality of emissions avoidance or reduction projects is difficult to verify, as it requires the development of counterfactual scenarios to demonstrate that emissions reductions only took place due to the incentive from the carbon credit revenue generated.

Research has shown that it is likely that many projects would have occurred without carbon credit revenue. For example, given that renewable energy technologies are now cheaper than fossil fuels in most countries, hundreds of millions of tonnes of credits from renewable energy projects that are currently traded are likely not additional. A review of 1,350 wind farms across India found 52% of these projects would have been built without additional income from CDM credits, misallocating scarce resources and likely leading to increased global emissions. 

Renewable energy projects are now so unlikely to be additional that many of the largest VCM registries, like Verra and Gold Standard, no longer allow these projects. However, it is still possible to purchase renewable energy credits: A Bloomberg report found that in 2021 “one third of the carbon offsets purchased from the 100 highest-selling projects were tied to renewable energy”.  

As of February 2025, all grid-connected renewable energy project methodologies assessed by the carbon offset governance body, the Integrity Council for the Voluntary Carbon Market (ICVCM), had their applications for the Core Carbon Principles label rejected because they lacked additionality. 

Investigations have revealed that carbon offset quantities are frequently overstated, especially forestry offsets 

Forestry offset projects issue credits based on commitments to reduce or avoid deforestation or forest degradation. However, these types of offsets are challenging to verify, as this would mean demonstrating that existing forests would have been cut down or degraded at higher rates if the project were not funded. 

Investigations by authoritative sources have exposed businesses selling forestry offset credits for fraudulent business practices, such as inflating the baseline rates of how much deforestation would have occurred in the absence of the project to generate more credits, taking credit for trees that were already planted and selling offsets for preserving forests that were not in danger of being harvested. 

An investigation by the Guardian, Die Zeit and SourceMaterial found that one of the world’s leading carbon credit certifiers, Verra, overstated the threat of deforestation by 400% on average. As a result, more than 90% of its carbon forest offsets “do not represent genuine carbon reductions”. 

A 2023 investigation published in the New Yorker into an avoided deforestation project in Zimbabwe that generated 42 million carbon credits found that only 15 million of these generated a legitimate avoided emission due to inflated baselines. EU countries have also been exposed by the media for overstating logging targets in order to pocket additional carbon credits, worth 120 million tonnes of CO2e. 

Academic papers highlight that emissions reduction efforts from deforestation and forest degradation (REDD) projects and California’s forest offset programmes have not reduced deforestation. A review into a type of credit called Improved Forest Management offsets, which have produced 11% of all voluntary offsets to date, found that issuance of these credits deviates from scientific understanding and risks significantly overestimating the amount of carbon removed.

Carbon offsets assume that all carbon is equivalent, but carbon stored in nature is not permanent – it is better to leave fossil fuels in the ground

Nature-based offsets assume that all emissions are equivalent and fungible, meaning they are mutually interchangeable. Proponents of carbon offsetting believe that carbon emissions from one source, such as industry, are equivalent to the emissions from other sources, like cutting down trees. 

However, this assumption – although foundational to the idea of offsetting – is incorrect. Research published in Nature Climate Change indicates that removing one tonne of carbon does not have the same effect as releasing one tonne of carbon. When researchers tested this theory using Earth system modelling, the results suggested this would result in a slightly higher CO2 concentration than before.   

Additionally, releasing a tonne of carbon from fossil fuels cannot be exactly offset by taking up one tonne of carbon by planting a forest. The carbon released from fossil fuels was stored for millions of years under the earth in a stable form, while the carbon taken up in the forest will – at best – be stored for decades, until the trees begin to degrade. So while storing carbon in nature could help to delay warming in the near term, nature-based carbon credits should not be treated as equivalent to fossil fuel emissions.

Some types of carbon credits, particularly those that are nature-based, also lack permanence. They risk the premature release of carbon back into the atmosphere due to natural or human-caused events, for example via wildfires or clearing of forests. A report on avoided deforestation projects in Cambodia used satellite imagery to show that four years later, half of these areas had been cleared. In 2021, the Los Angeles Times reported that a forest carbon offset project in California continued to sell credits even after many trees burnt down in wildfires.

Offset projects that aim to reduce or avoid emissions in one area can also cause emissions to increase or ‘leak’ to another area to compensate. A 2023 review of nature-based offsets published in the journal One Earth shows that “leakage is vastly underestimated in practice and argue[s] that current efforts to improve accounting methods are unlikely to deliver the accuracy required.”

Poor-quality credits mean there is an oversupply of credits in carbon markets

While generating as many carbon credits as possible may appear to make business sense, an oversupply of poor-quality credits has pushed the carbon price – and therefore the value of the credits – down. 

So many credits with a low likelihood of achieving actual emissions reductions have been issued that there is an oversupply in the market. According to the Berkley Carbon Trading Project, there is currently a surplus of 829 million credits on the VCM, just counting those from the four top carbon registries. As a result of being oversupplied, credits are priced far too low to reflect the true cost of carbon, which means it is much cheaper for companies and countries to purchase ‘junk’ credits than to reduce their emissions.²Offsets on the VCM are currently priced at USD 1-10 (as of February 2025) The 2017 Report of the High-Level Commission on Carbon Prices estimated that carbon prices should be USD 40-80 by 2020 and USD 50-100 by 2030 to be consistent with achieving Paris Agreement temperature goals. Other studies estimate that the real cost of carbon is much higher – around USD 185 per tonne of CO2.

For example, research published in Nature analysing VCM activity found that the 20 companies retiring the most offsets from the VCM over 2020-2023 predominantly sourced low-quality, cheap offsets, 87% of which “carry a high risk of not providing real and additional emissions reductions.” Most of these offsets came from forest conservation and renewable energy projects. Purchasers come from a range of sectors, including automotive, aviation, oil and gas, consumer goods, and transport and logistics firms, all of which have access to meaningful – albeit more expensive – decarbonisation activities within their own supply chains. 

Offset projects have resulted in fraud, human rights abuses and carry reputation and litigation risks for companies

Offset projects are often purportedly designed to benefit local communities and protect their environment, and are promoted as a way to channel climate financing towards developing countries, which is much-needed, particularly for the protection of nature. 

However, carbon offset projects have been found to provide little funding for project implementers on the ground and have been linked to human rights abuses and land conflicts that can push Indigenous People and local communities off their land.  The money designed to be channelled into communities is pocketed by intermediaries and project developers, many of which commit fraud, for example by exaggerating the baseline risk of deforestation, to entitle themselves to more funding. 

Purchases of carbon offsets by companies are often seen as contributing to climate action, but misleading claims about the outcomes of using carbon offsets are increasingly leading to reputational and legal risks for companies, reflecting that courts agree that carbon offsets fail to act as a climate solution. For example, in 2023, a regional German court ruled that advertising claims of climate neutrality from the use of offsets by airline company Eurowings were misleading and breached German law. Similar lawsuits have been taking place around the world, causing companies to abandon carbon neutrality and net zero claims associated with the use of offsets. 

There are more effective ways to fund climate action than carbon offsetting, despite its popularity

Despite carbon crediting schemes being promoted as a means to increase climate finance, there are other, more secure ways for countries and businesses to deliver climate finance.

Businesses can contribute to emissions reductions and other climate projects without claiming carbon credits. NewClimate Institute suggests that companies can do this by setting an internal carbon price and using this to support climate initiatives outside their own value chains. This resulting ‘climate contribution’ represents “a financial commitment that is a complement – and in no way an alternative – to directly reducing one’s own climate footprint.”

The offsetting industry has helped shape a narrative around the scarcity of funding. In reality, the issue is less about the amount of money available and more about how it is distributed and who decides. In 2022, Global North countries spent almost 80 times more on fossil fuel subsidies than the USD 31.5 billion they provided in grant-based climate finance in the Global South. Research by Oil Change International suggests that over USD 5.3 trillion a year could be generated by rich countries ending public finance for fossil fuels and implementing fossil fuel and higher weather taxes, among other approaches.

The way forward: Untangling ideas about carbon offsets from what climate science says about their effectiveness

The use of offsets is not aligned with the scientific need to lower emissions and avert warming beyond safe levels. Although some policymakers and companies may see them as a cheap and easy solution to meet emissions targets, they do not durably remove carbon and instead allow companies and countries to continue emitting in line with business as usual, instead of investing in meaningful mitigation measures. 

Based on an analysis of 20 voluntary carbon offset registries for credits retired between January 2020 and December 2022, the top users of carbon offsets were fossil-fuel producers, car makers and tech firms. Companies such as Shell, Volkswagen and Chevron – the three largest purchasers of carbon credits – have routes to decarbonisation. A study that examined how carbon offsetting was used in four major oil companies’ net zero strategies found that none had plans to transition away from fossil fuels. 

Ultimately, offsetting has done more harm than good for overall emissions reductions. The VCM has failed to deliver real impact and continues to have integrity issues. The use of carbon offsets by companies and countries should continue to be closely scrutinised. Emissions avoidance and reduction offsets have no place in ambitious climate policies. 

Mitigation should always be the first course of action – and should always be prioritised over carbon removal. These tested routes to limiting warming (for example, the implementation of solar, wind, battery storage and energy efficiency improvements) are also more cost-effective. There are a broad range of mitigation options that are not yet fully utilised. 

Support for the restoration of degraded ecosystems and the protection of biodiversity, alongside other mitigation efforts and the theoretical need to incentivise real, permanent carbon removal, does not have to be serviced by a market mechanism in the buying and selling of carbon credits. Instead, funding can be sourced outside of carbon offset revenue, such as via public financing, including revenue generated from wealth and fossil fuel taxation, and company climate contributions. Direct and targeted support should, in particular, be provided for countries with high forest cover and developing countries.

This copy was redeveloped and reformatted in March 2026. 

• 1
  While the technology used is very similar, carbon removal captures differs in that it captures diffuse CO2 emissions from the atmosphere, while carbon capture and storage (CCS) is used to capture emissions from the point source of an emitting process.
• 2
  Offsets on the VCM are currently priced at USD 1-10 (as of February 2025) The 2017 Report of the High-Level Commission on Carbon Prices estimated that carbon prices should be USD 40-80 by 2020 and USD 50-100 by 2030 to be consistent with achieving Paris Agreement temperature goals. Other studies estimate that the real cost of carbon is much higher – around USD 185 per tonne of CO2.