Governments and companies are betting heavily on the voluntary carbon market (VCM) to achieve decarbonisation goals. However, many carbon credits are not only risky (because saving carbon is not guaranteed) but highly uncertain in the sense that the chance of success simply cannot be measured. Such uncertainty leads to endless debate over the value of credits and the methodologies used to measure that value – as evidenced by a recent investigation by journalists at The Guardian, Die Zeit and SourceMaterial into rainforest carbon offsets issued by the largest voluntary standard, Verra.

For the VCM to support climate goals, the net impact of a carbon credit from issuance to use must be a reduction in the volume of greenhouse gases in the atmosphere. In this commentary, we explore how these net mitigation benefits can be achieved despite the variable and uncertain quality of credits.

Net impacts depend on substitution

The history of domestic recycling programmes offers useful insights into how to evaluate the net impacts of carbon offsetting. As concern about growing volumes of plastic waste rose throughout the 1970s, recycling programmes (implemented following lobbying by polluting firms) helped defuse momentum towards regulation and kept consumer demand high. Unfortunately, these programmes removed incentives for firms or policymakers to invest in alternatives such as bottle return schemes.

Over time, the inability of recycling to solve the plastic waste problem (seen in accumulating waste stocks and rising microplastic pollution) led to renewed public concern. Roughly a generation after recycling went mainstream, firms are investing in low-pollution technologies (e.g. biodegradable plastics) and policymakers are pursuing other regulatory options (e.g. bans on single-use plastics, or exchange schemes). 

The history of plastic recycling demonstrates that the success or failure of a pollution reduction programme depends on substitution for technological innovation, alternative policies or reduced demand, as well as the effectiveness of the programme itself. The same considerations apply to carbon markets – but with the added complication that the effects of a given carbon credit on atmospheric greenhouse gas stocks are extremely difficult to calculate.

Short of altruism, firms purchase carbon credits to substitute for other, more costly actions. These may involve investing in emission-reducing technology, or spending more on marketing goods or services as ‘green’.

Substitution also occurs at a systemic level. Time and resources spent developing carbon markets come at the cost of pursuing other policy initiatives, and the existence of markets may alter economic and political incentives to pursue other solutions.

We argue that what carbon credits substitute for and uncertainty over their effectiveness are issues that are fundamentally related. If uncertainty can be pinned down as quantifiable risk, or if it can be resolved at some future point, then buyers can use a range of credits to offset their emissions. However, when uncertainty is irreducible, regulators can design market frameworks to incentivise net-positive mitigation by understanding substitution effects.

Planning for uncertainty: a substitution approach

With the door closing quickly on a world that avoids warming of 1.5oC, uncertain carbon credits are controversial – but they may still be indispensable in halting climate change. Understanding substitution could promote net mitigation benefits despite high uncertainty. In the extreme case, a credit whose true effectiveness will never be known can still have net benefits provided it does not replace other actions to reduce emissions.

Uncertain credits cannot be used in accounting systems that treat each tonne reduced or removed as a certainty, because doing so distorts accounting and may weaken incentives to decarbonise. At least one standard-setting body – the Science-Based Targets initiative (SBTi) – responds to this reality by ruling out certain kinds of substitution. Under SBTi rules, highly uncertain ‘avoidance credits’ (from projects that reduce emissions against a hypothetical baseline, like preventing deforestation, rather than removing carbon from the atmosphere) can only be used for ‘beyond-value-chain mitigation’. SBTi and the Voluntary Carbon Markets Integrity Initiative (VCMI) are actively developing further guidance on the legitimate use of credits.

We argue that such market stratification is the right response to uncertainty, and should be based on matching the use to which credits are put with their degree of uncertainty (an idea developed in more detail in this working paper). Only credits with highly certain effects should be used to substitute for emissions that could have been avoided with equal certainty (such as through switching production processes or taking a train rather than flying). When credits substitute for actions with more uncertain mitigation consequences (such as spending more on marketing), uncertain credit quality is more acceptable.

With the ‘rules of the game’ still in development, the VCM could move past the dominant offsetting paradigm, where credits are matched with some portion of unabated and/or historical emissions on a ‘tonne-for-tonne’ basis. In a stratified market, uncertain credits would only be used for specific purposes: for example, to make claims about contributions to overall mitigation goals but not carbon neutrality or net zero emissions. The SBTi approach is an important first step but further nuance and stratification could unlock more finance for a broader suite of risky or uncertain interventions.

Resolving uncertainty: a portfolio approach

Different remedies would be available if uncertainty about credit quality can eventually be resolved, allowing buyers to take bets on credits or reduce volatility by investing in a portfolio of risky offsets. More complicated arrangements such as options contracts also become possible, increasing liquidity in the VCM by providing flexibility for buyers.  

The core of a portfolio approach is matching multiple credits to each unit of emissions that a buyer seeks to offset. The problem is how many credits to buy. If uncertainties about credit impacts can be quantified into risks, then equivalence can be calculated and credits can be appropriately priced (as shown here). But without good data on credit risk, what can be done? One option – which we put forward to stimulate debate – is to use internal corporate carbon prices or the social cost of carbon (SCC) as a benchmark. At a SCC of $50 per tonne and a credit price of $5 per tonne, 10 credits could be retired to offset a single tonne of emissions.

For a portfolio approach to work, monitoring and accounting systems need to be improved. Continuous observation is necessary to resolve uncertainty over each credit’s impact, and to ensure appropriate consequences for market participants based on the evolving value of their credit portfolio. These requirements are technically possible but would be a substantial advance on current VCM accounting practices. Over time, effectiveness can be monitored by comparing projects against similar peers or using quasi-experimental controls, and the availability of satellite data for monitoring land-use change projects offers a low-cost solution for some important project classes. Ratings agencies like Sylvera and BeZero Carbon are already working to provide independent, rigorous assessments of the quality of select credits using this kind of data.

Is a bad carbon credit better than none?

Let’s take the common problem of whether to offset business flights using carbon credits. If credit quality is known, then it makes sense to buy multiple (but risky) projects’ credits. However, as long as uncertainty over quality has not been resolved, then credits cannot substitute for the certainty of extra emissions without destroying the integrity of carbon accounting.

Flights on aircraft powered by conventional aviation fuel should first be avoided when possible, rather than offset: alternatives such as teleconferencing or lower-carbon forms of transport should be considered, but if these are not a reasonable option then purchasing credits should be on the table. However, uncertain credits should not be treated as a direct substitute for emissions or used to claim that the journey’s net emissions are zero or negative when in fact the true effect on greenhouse gas stocks is unknown. Instead, credit purchases can be used to make claims about a commitment or contribution to global goals like protecting forests. The VCM – and people’s faith in company claims – would benefit from a shift towards honest treatment of this fundamental uncertainty.

The authors would like to thank Luca Taschini, Frank Venmans and Ben Groom for their comments on an earlier draft of this commentary. The views in this commentary are those of the authors and do not necessarily represent those of the reviewers or the Grantham Research Institute.

Keep in touch with the Grantham Research Institute at LSE
Sign up to our newsletters and get the latest analysis, research, commentary and details of upcoming events.