How to avoid carbon removal delaying emissions reductions

With global carbon budgets shrinking rapidly, carbon removal strategies have become an essential component of reaching net zero emissions and stabilising rising temperatures. However, concerns persist within the climate policy community that reliance on future carbon removal deployment could deter or delay urgent efforts to cut emissions in the near term. This phenomenon, known as ‘mitigation deterrence’ or ‘moral hazard,’ poses risks of overshooting carbon budgets if emissions reductions are reduced today based on assumptions about large-scale availability of carbon removal technologies in the future.

This report aims to enable policy-makers and practitioners to responsibly accelerate carbon removal deployment while avoiding mitigation deterrence risks. It provides a comprehensive framework for understanding and proactively addressing potential mitigation deterrence across three key use cases of carbon removal as defined by the IPCC. Based on this analysis, tailored solutions are proposed to remove the factors encouraging mitigation deterrence and establish the guardrails needed to limit its impact.

Accelerating progress towards net zero

The first use case examines deploying carbon removal to complement aggressive emissions reductions in accelerating progress towards reaching net zero targets. With current climate action lagging, carbon removal deployment must begin immediately in order to have time to scale up to the massive gigatonne levels needed by mid-century. Starting deployment now also reduces risks of carbon removal solutions failing to materialise at scale or underperforming expectations.

Removing obstacles to emissions reductions through incentives, infrastructure, and regulations is a key starting point to avoiding mitigation deterrence. However, safeguards are  also essential in the near-term to avoid mitigation deterrence, given the remaining uncertainty about carbon removal technologies. The report highlights the like-for-like principle where only permanent carbon storage can be used to neutralise fossil fuel emissions. Doing so avoids the risk of deterring fossil fuel phase-outs by relying on low-permanence biogenic storage.

In addition, the report advises instituting separate, fixed interim targets for emissions reductions and carbon removal ramp-up. These targets would prevent carbon removal deployment from displacing ambitious reductions.

Reaching net zero

Enabling net negative emissions

The third use case involves planning to use future carbon removal technologies to achieve net negative emissions and lowering temperatures after overshooting 1.5C or 2C climate goals. The theoretical availability of negative emissions to compensate for overshoot could deter the radical near-term emissions reductions required to avoid overshooting carbon budgets in the first place.

Several factors determine whether mitigation deterrence is occurring in relation to overshoot for a given country. These include the feasibility of further emissions reductions under current conditions, a country’s interpretation of their fair share of global effort, and their overall commitment to climate goals.

Although few countries explicitly plan to overshoot their fair share of the carbon budget, many could be said to implicitly do so. Countries without plans for near-term emissions cuts could be said to be particularly exposed to mitigation deterrence, and they provide particular cause for concern. Countries that already have 1.5C-compliant emissions reductions targets that are not overly reliant on CDR could be said to not be mitigation deterred.

The report puts forward solutions to avoid mitigation deterrence related to overshoot. Firstly, urging countries to set binding  near-term targets consistent with 1.5C carbon budgets and rapid mitigation is paramount. Any references to negative emissions solutions should also prompt demands for detailed roadmaps demonstrating feasibility and alignment with the like-for-like principle. Quantitatively allocating future negative emissions responsibilities across countries would also sharpen incentives for ambitious targets today. The last proposed solution is to make countries set targets for all of the emissions for which they are responsible, including their consumption-based emissions and the emissions from the export of fossil fuels, for example.

In addition to the policy-driven mitigation deterrence forms covered in the three use cases, the report acknowledges a more indirect discourse driven form of deterrence. Portraying removal as a ‘silver bullet’’ solution risks creating complacency and a justification for climate inaction. While this indirect deterrence risk is difficult to quantify, the solutions proposed in the report would significantly limit its impacts.

In summary, this report’s proposed solutions enable decision-makers to responsibly scale up removal while maintaining urgency on emissions reductions and avoiding a harmful tradeoff.

There is broad agreement that carbon dioxide removal (CDR) at scale is needed to stabilise climate change well below 2°C (Paris Agreement) by reaching net zero greenhouse gas emissions and then enabling a net negative phase to compensate for any overshoot and bring atmospheric CO2 concentration back to a safer level. There is, however, disagreement around the size of CDR’s contribution and how and when different actors should engage with it. Mitigation deterrence, or moral hazard, is a key theme in these disagreements and refers to the risk of delaying or stopping emissions reductions due to the availability of alternative mitigation options, among them carbon removal.

Mitigation deterrence is not a new concept and it applies equally to other mitigation options such as carbon capture and storage (CCS) and offsetting, along with broader climate interventions such as solar radiation management (Merck 2019, Austin and Converse 2021). The appearance of new technological alternatives mostly deployed in the future creates the fear that they might interfere negatively with ongoing efforts and divert resources or political will away from existing/preferred near term options.

Such fears are particularly justified by a series of deeply-embedded societal drivers that favour mitigation deterrence. First, the fossil fuel industry and its associated infrastructure is still at the heart of how our society works. The transition away from fossil fuels requires huge changes that may potentially create significant amounts of stranded assets. Therefore, any alternative that would allow reducing emissions while continuing to use the fossil fuel system is naturally attractive.

Second, as pointed out by Carton et al. (2023), a significant driver dates back to 1992 and Nordhaus’s early conceptualisation of what today is called the marginal abatement cost. Since then, collective thinking on mitigation is dominated by an economic approach where mitigation options compete mostly on cost efficiency. Policy-makers often favour economic considerations over other such important dimensions as impacts on ecosystems and communities, long-term sustainability and climate justice, thereby enabling mitigation deterrence.

A third important driver is the nature and temporalities of the challenge we are facing. Actors are supposed to make rational decisions if they are well informed and exposed to the consequences of these decisions. Climate change deviates from this ideal situation; individual decisions made today generate shared consequences in the future. Therefore, today’s decision-makers are largely protected from the consequences of a failure in the substitution of one mitigation by another. This situation reduces their aversion to the risk they take, creating further room for mitigation deterrence.

Some authors argue that in the case of CDR, the fear of mitigation deterrence is further justified by the fact that current existing mitigation options are compared with future theoretical techniques that are unproven, adding the risk that they may actually turn out to be impossible. This argument is increasingly becoming less true as CDR methods are deployed and acquire more data that supports them as a proven approach.

McLaren (2020) reviews three scenarios that are helpful for understanding where mitigation deterrence can come from:

• Substitution and failure: refers to situations where CDR pathways fail to deliver the expected removed tonnes required to replace emissions reductions. There could be many reasons for such failure: lower technical performances, higher reversal, higher cost than expected, or even a more systemic failed scaleup due to policy shortcomings.
• Rebounds: refers to ‘indirect and typically unintended effects in which GGR [Greenhouse Gas Removal] might trigger additional emissions,’ for example by maintaining fossil value chains with all the associated emissions beyond those directly compensated for by CDR.
• Mitigation foregone: or ‘imaginary offsets’ refers to the indirect reassurance effect that CDR may create, resulting in actors lowering their engagement on emissions reductions, even unconsciously.

In determining where mitigation deterrence happens, Carton et al. (2023) suggest that mitigation deterrence occurs mechanically in Integrated Assessment Modelling (IAMs) used by the IPCC, since any CDR option introduced in the model mathematically displaces emissions reductions. Their conclusions are more nuanced at an individual level; where diverging conclusions are reported, they are largely linked to the design of the studies and have potentially low applicability to the more complex and subjective nature of real-world decision making.

For some, deploying carbon removal before net zero is a distraction and creates room for increased mitigation deterrence. Yet, if CDR is deployed today in addition to emissions reductions, it can contribute to minimising the CO₂ increase in the atmosphere. However, the main reason to deploy carbon removal today is to increase its deployment toward the scale needed to reach and maintain net zero. As CDR is scaled up there is some possibility it becomes ‘too attractive’ an option and replaces emissions reductions.An example of mitigation deterrence under this use case for CDR could be offsetting the use of fossil fuel powered internal combustion engine (ICE) cars. If permanent removal credits would cost 200 USD per tonne, the cost of offsetting a litre of petrol would be around 50 cents. This amount would be a significant increase in petrol prices, but still just around 25% of the price at the pump in many European countries. Many consumers might see this asymmetry as an acceptable tradeoff to be able to continue to use their ICE cars. The problem is that such neutralisation would create a risk of lock-in and slow the adoption of far more resource-efficient electric vehicles. This risk in turn would lead to the world being unnecessarily dependent on carbon removal.

In the short term, one of the main drivers for mitigation deterrence is arguably the lack of accessible mitigation options to decarbonise existing activities. Removing obstacles to emissions reductions should be the starting point of all efforts to address mitigation deterrence. If companies and consumers cannot make choices that reduce their emissions due to various barriers, carbon removal might end up being the only actionable option, even if sub-optimal. For instance, consumers or businesses may not opt for electric vehicles due to a lack of charging infrastructure or higher upfront costs. Similarly, companies may continue to rely on fossil fuels if clean energy is not accessible due to a slow build out of the electric grid. Therefore, policy efforts must increase the accessibility and affordability of sustainable alternatives, improve infrastructure, provide incentives and subsidies for green choices and implement regulations that discourage high-emission activities.

One of the most efficient solutions for avoiding mitigation deterrence in this use case for CDR is to set separate fixed targets for CDR and emissions reductions in the short to medium term. Doing so forces companies and countries not to offset any specific emissions, but instead scale up CDR according to a set target and reduce their emissions in line with science. This solution restricts the substitution of emissions reductions with carbon removal, so that if a company or country increases their CDR deployment, they avoid the risk of displacing other emissions reductions efforts. One example could be the EU setting a target of 55% emissions reductions and, in addition, a (hypothetical) target for permanent carbon removal of 50 million tonnes for 2030. For long-term targets however, it would be sub-optimal to set fixed proportions between CDR and reductions today, since we do not know how different removal and mitigation options will be developed. On the other hand, as is discussed later under use case B, decision-makers today should assume that the amount of CDR available by mid-century will be limited.

Poorly or unregulated net zero or climate claims are also a major incentive for mitigation deterrence since companies can claim additional recognition or value on weak mitigation strategies. There are legions of companies buying large amounts of cheap CDR credits instead of reducing emissions and claiming to be carbon neutral. Eventually, such mechanisms divert significant amounts of money and create the false impression of climate action (see our position paper on Green Claims here).

The solution to this issue is to carefully regulate climate related claims and increase the requirement for transparency. Examples already exist where such claims must be supported by a genuine decarbonisation strategy that imposes yearly updates to inform the public about the company’s progress.

Mitigation deterrence may also occur when a company seeks to avoid expensive emissions reductions measures by categorising most of its fossil emissions as ‘residual’ or ‘hard-to-abate’ and compensates for them with cheap, low durability carbon removal. Such an approach contributes to creating both unsustainable (fossil carbon stored in biogenic pools) and high flows of carbon at net zero.

Implementing the like-for-like principle (see Figure 3) as defined above is a crucial part of avoiding low integrity strategies and mitigation deterrence. Fossil CO₂ from the long carbon cycle cannot be allowed to be compensated with CO₂ restored to the short carbon cycle.

For now, high permanence CDR remains very costly and is available in limited amounts. Playing on the economic approach to mitigation described above, these high costs can serve as a natural disincentive for mitigation deterrence if countries are able to carefully regulate compensation following the principles of like-for-like removal, maintaining durable net zero. Indeed, any fossil emission would have to be compensated by high permanence removals, making their high price tag a natural disincentive for mitigation deterrence.

Another risk is CDR not growing at the pace required to fulfil its role. CDR at scale requires innovation, talent, permitting, new policies, land, engagement with local populations and a lot of electricity.  To grow durable CDR from the tens of thousands of tonnes removed today to the gigatonnes needed at net zero, rapid deployment needs to start today (for example reaching a removal capacity of just 1.5Gt/yr would require building one 1 Mt/yr removal facility every week for 28 years.) If we wait until all cheap and actionable emissions reductions have been made, we will be decades too late deploying CDR, resulting in a large overshoot.

Moreover, at this stage (near term), CDR still lacks credibility, and uncertainty on the feasibility of the different CDR pathways remains large, increasing the risk of ‘substitution and failure’ with actors betting too much on it. Instead, policy-makers must invest in the emerging CDR industry to accelerate its deployment, build test facilities and acquire the data and knowledge required to reduce uncertainty on what CDR may or may not deliver and reduce the risk of ‘failure’.

A company or country that sets a net zero target but does not plan to substantially reduce its emissions and instead relies on future removals is engaging in mitigation deterrence. Here, the belief that carbon removal will be available and affordable in the future that potentially keeps the company from making near-term emission cuts, while they still get good publicity from what might look like an ambitious target.

To prevent such situations, short-term emissions reduction targets that consider carbon budgets should be required to accompany any long-term targets. A target that is only concerned with reaching net zero irrespective of the path taken contributes to leading to an overshoot and a ‘high flows’ net zero state. Moreover, such an approach also endangers the net zero target since very large quantities of CDR are not likely to be available.

While CDR must be scaled as fast as possible, countries and companies should also purposely plan for a limited amount of removal to be available in reaching their long-term net zero targets. This restraint is to account for the high uncertainty that remains on the feasibility of large scale deployment of CDR, and is in line with the precautionary principle enshrined in various international treaties, including the Treaty on the Functioning of the European Union.

The Science-Based Target initiative’s (SBTi) Net Zero Standard is an example of such an approach where companies need to reduce emissions by at least 90% and only use carbon removal to neutralise up to the last 10%. However, for corporate and national long-term targets, a range rather than a fixed percentage is likely best since it leaves room for technological development both in emissions reductions and carbon removal.

In considering hard to decarbonise sectors, a clear principle of when to deploy carbon removal instead of emissions reductions becomes essential. The remaining emissions to be compensated with removal are referred to as residual emissions. Currently, net zero pledges refer to residual emissions without a clear definition, leaving room for companies to argue that most of their emissions are ‘unavoidable’ (Buck 2022).

Strictly speaking, no emissions are physically impossible to eliminate. However, for some sectors/activities such as aviation, there may be limited to no options to decarbonise, and emissions will probably not be capped to 5 or 10 % of their 1990 levels. There is a large possibility that carbon removal will be the option that is cheapest and with the fewest negative side effects for aviation. Of course, not emitting at all is always the cheapest option but also carries alternative costs.

Ideally, the decision of whether or not to deploy CDR instead of emissions reductions should be based on:

• the cost, availability, side-effects (including effects from continuing to use fossil fuels) and co-benefits of the carbon removal methods;
• the alternative cost, availability, side-effects and co-benefits of emissions reductions options, including demand reduction.

This analysis is complex and not all effects might be known. Therefore, it may be hard to set rules that let the market decide when CDR should be used, which may lead to rules designating certain sectors for CDR compensation and not others.

Once the level of residual emissions is controlled by regulations like the ones proposed above, the compensation of the ‘credible’ residual emissions should follow the like-for-like principle mentioned in use case A to guarantee sustainable carbon flows.

A different approach would be to implement solutions like a carbon takeback obligation where all fossil fuel use would be required to be removed and durably stored, but where there is disagreement on what the fossil fuel is used for.

Finally, we must consider who controls emissions. For example, some corporations could be compelled to use carbon removal to reach net zero for their Scope 3 emissions, even when emissions reductions would be theoretically cheaper, but these corporations lack the necessary control of the supply chain. For example, Scope 3 reductions are very challenging for a company whose emissions are spread over thousands of suppliers for whom they are just one of many customers. As more company and product specific emissions data becomes available, it will be easier for companies to select suppliers with lower carbon intensity, yet it is likely that a significant share of Scope 3 emissions will not be addressed without carbon removal.

Carbon removal could, in theory, be used to reach net negative emissions and reduce temperatures. This use case is an entirely different one for CDR than using it to reach net zero, requiring deployment at a scale several orders of magnitude larger if done in a short time span. Researchers estimate that we need on average 222 Gt of carbon removal to lower temperatures by 0.1C. To put that number into context, humanity has emitted 1736 Gt of CO₂ since the start of the industrial revolution. To lower temperatures by 0.1C today, ~13% of the CO2 emissions ever emitted would need to be removed (given no more CO₂ is added), showing the extreme difficulty of dealing with overshoot through carbon removal.

Because reaching net negative emissions can address legacy emissions and restore some carbon budget, the responsibility to go beyond net zero will most probably fall onto countries with the largest historical emissions and ability to pay, adding further complexity in how countries plan their targets and negotiate the efforts needed to respect the carbon budget.

The availability of negative emissions for this purpose could, in theory, keep nations from setting targets that avoid or limit overshoot to a minimum. As Buck (2022) explains, currently, aggregated national climate targets lead, at best, to overshooting the 1.5C carbon budget and reaching at least 1.8C warming. However, answering who is being deterred is key in determining if mitigation deterrence is happening.

When attempting to determine if a country is subject to mitigation deterrence in relation to overshoot, several factors must be considered, including the perceived feasibility of stricter emissions reductions, the country’s interpretation of what constitutes a fair contribution to global emissions reductions efforts and their level of commitment to tackling climate change. For instance, countries that perceive it as impossible to further reduce their emissions beyond current plans are unlikely to primarily be influenced by the potential for CDR, since their lack of stricter targets is not driven by the availability of this technology. In this case, it is the perceived impossibility of stricter emissions reductions that makes the country fail to set a stronger target, not a reliance on CDR. Similarly, a country that is indifferent to whether the global temperature increase stays within the 1.5C limit is unlikely to have emissions reductions deterred by CDR.

Furthermore, a country that has set targets in line with the 1.5C limit based on their interpretation of a fair contribution could be viewed as not being influenced by mitigation deterrence. They may argue that they are already fulfilling their responsibility and any potential overshoot is the responsibility of others based on their interpretation of a fair allocation.

We find it helpful to differentiate between explicit and implicit mitigation deterrence. With explicit mitigation deterrence, we refer to planning to overshoot one’s own budget and then using negative emissions to compensate. This type of direct mitigation deterrence seems rare. No countries have such explicit plans to our knowledge. Implicit mitigation deterrence appears to be more common and we define it as not explicitly planning for net-negative emissions but overshooting budgets nonetheless.

It is important to note that even if countries only implicitly use carbon removal to deal with overshoot, it can still be intentional. Intentional implicit mitigation deterrence is hard, if not impossible, to prove. A country or company might overshoot their budget even if CDR did not exist. They might also use CDR as a convenient explanation for why they are overshooting. As Carton et al. (2023) explains, ‘no amount of modeling or experimental studies can tell us what targets companies or countries would have set in the absence of carbon removal, under actually existing social, political, and economic conditions’.

An example of implicit mitigation deterrence could be Norway’s continued oil production. While Norway is in many regards a leader in the green energy transition with ambitious domestic climate targets, they continue to extract and sell oil. The emissions from burning that oil are not attributed to Norway, but to the end consumers. Even though research clearly states that no new oil fields can open if the world is to stay within 1.5C without overshoot, Norway is steadily taking new fields into production. Norway does not explicitly say that carbon removal should be used to deal with the emissions from its oil production, but it could be said to implicitly ask for it.

Another example of implicit mitigation deterrence comes from countries with weak targets that do not plan to make any near-term emission cuts. Such examples include China, Russia, Mexico, Egypt and Singapore whose nationally determined contribution plans do not include emissions reductions until 2030. According to Climate Action Tracker’s analysis, these countries’ targets are critically or highly insufficient for 1.5C. Such countries are on their way to massively overshooting their share of carbon budgets (together with large historic emitters such as the US, but for actors such as the US and the EU, most emissions lie in the past rather than in the future). The insufficiency of the targets stands true even in a fair share analysis taking historic emissions into account. Such countries could be said to implicitly ask for large-scale carbon removal in the future to make up for the overshoot they are creating.

The greatest risk for implicit mitigation deterrence is arguably among the actors that currently have the least aggressive plans for reducing emissions, provided they are committed to achieving global climate goals in the first place. These actors have the most room to enhance their targets, which could potentially lead to a greater reliance on future carbon dioxide removal to deal with overshoot as a justification for not adopting more stringent targets. For instance, it is easier to make significant emissions reductions in the absence of ambitious reduction plans to begin with, as is the case with China and Norway’s oil production. On the other hand, for those who have already committed to aggressive emissions reduction targets, like the European Union with its goal of 55% emission cuts from 1990 by 2030, further improvements are relatively harder (albeit far from impossible) to achieve, since they are already planning significant reductions.

Countries that already have 1.5C-compliant emissions reductions targets that are not overly reliant on CDR could be said to not be mitigation deterred, at least not for their production-based emissions. The main concern for them is to deliver on their pledges, rather than on strengthening them. The obvious solution to the mitigation deterrence risks described above is to get countries and companies to set and implement targets that avoid or minimise overshoot. Targets that avoid mitigation deterrence are those that are followed, that are 1.5C compliant, and do not include excessive amounts of CDR. However getting actors to set and implement such targets is easier said than done, given that actors with relatively weak targets plan to emit a large share of future emissions and might not have much interest in strengthening them.

Another way to reduce the drivers for mitigation deterrence is to allocate the future responsibility for the negative emissions that are needed to reverse overshoot. Unallocated responsibilities for net negative emissions carry a much higher risk for permanent overshoot than allocated ones. Such allocation principles can also include the responsibility for historical emissions and principles of fairness. Fyson et al. (2020) explores allocations of negative emissions by looking at scenarios that take into account historical emissions, least-cost and the ability to pay. The approach either gives the US (historical emissions) or China (ability to pay) the greatest responsibility for CDR, illustrating the difficulty in agreeing on a principle for allocations.

The main advantage of allocating such quantified responsibility to countries would be that it would strongly incentivise them to police mitigation deterrence to ensure their CDR capacity is well used. Indeed, if they are committed to delivering certain amounts of removal, countries would probably monitor the uses of their CDR capacity to minimise the volumes of CDR dedicated to compensating avoidable emissions.

What else could actors who want to counter potential mitigation deterrence related to overshoot do? Convincing governments to disregard net negative emissions as a solution, as some CDR critics have suggested, would not be sufficient if mitigation deterrence is mostly implicit.

Currently, mitigation deterrence risks delaying both emissions reductions and CDR deployment, creating a dangerous situation. Yet the main takeaway of this discussion paper is that policy-makers can manage mitigation deterrence by removing the drivers and strengthening the required safeguards. There is a clear difference in the risk and type of mitigation deterrence depending on the use case of carbon removal, with proposed solutions to address each.  The solutions proposed can also help avoid more indirect forms of mitigation deterrence based on the public discourse of carbon removal.

The first use case involves using CDR as a complement to emissions reductions. To counter the risk of any associated mitigation deterrence, policy-makers must remove obstacles for emissions reductions and set separate targets for CDR and emissions reductions in the short to medium term. Furthermore, by supporting carbon removal technologies today, they can reduce the risk of CDR failing to materialise to fulfill its planned role. Policy-makers must also require like-for-like compensation and regulate CDR related claims.

In the second use case, CDR is used to compensate for residual emissions. Actors can handle associated uses of mitigation deterrence by setting short term emissions reductions targets and considering carbon budgets in addition to long-term net zero targets. Policy-makers can also counter mitigation deterrence by only assuming a limited amount of removal technologies will be available for net zero targets, and developing clear principles for when it is preferable to deploy carbon removal instead of emissions reductions.

The third use case deals with CDR used to avoid tackling overshoot and is the hardest one to address, but solutions include pushing those countries and actors that have less ambitious targets to strengthen them, asking for specificity in how they plan to use CDR and introducing supplementary targets for actors that encompass all emissions they influence. Assigning responsibility for the deployment of negative emissions can also sharpen countries’ monitoring and control of mitigation deterrence to preserve their CDR capacity.

Addressing mitigation deterrence today is critical to maximise the speed and scale of both emissions cuts and CDR deployment, but also to depart from the poor precedent set by the offsetting industry and set much higher standards to position CDR as a truly complementary and additional climate solution. It is critical that decision-makers start implementing these solutions today and tackle mitigation deterrence before it grows into a larger problem.