Biodiversity offsets, their effectiveness and their role in a nature positive future

Biodiversity offsets are used as a strategic tool to mitigate impacts on biodiversity in the context of the mitigation hierarchy¹ — a framework that determines the sequence of approaches to be taken when mitigating environmental impacts. Under the mitigation hierarchy, efforts should first be made to avoid impacts on biodiversity. If avoidance is not possible, impacts should be minimized; next, any temporary impacts must be repaired and the affected biodiversity restored on-site. Finally, if any impacts remain, offsets are used to at least fully counterbalance them through equivalent gains elsewhere. Thus, the full application of the mitigation hierarchy should achieve at least no net loss of biodiversity, and preferably a net gain, for a given project. Offsets have become an appealing solution to ‘unavoidable’ impacts on biodiversity, and have increasingly been embedded in impact-mitigation policy over the past two decades² (Box 1). However, despite increasing integration into policy frameworks, the design and execution of offset policy — and offsets themselves — often fall short of recommended good practice^3,4.

Offsets are no panacea. Biodiversity decline is driven by myriad factors over which humans have varying degrees of leverage. However, at least for impacts that are directly and immediately caused by project development — such as the removal of a forest for an industrial development, or the dredging of sea floor for a shipping lane — biodiversity offsets promise to at least counterbalance the damage done. By this standard, offsets often fail^5,6. Some of these failures are because of limited policy scope: by design or otherwise, most impacts proceed without any requirement for associated offsets (or impact assessment), even in jurisdictions where offset policies exist. Second, where offsets are required, failures result from policies specifying offset requirements that, even if implemented as intended, could not be expected to counterbalance the losses. Finally, failures of implementation are commonplace, involving promised offsets not proceeding or not achieving the outcomes intended⁴. Collectively, these shortcomings contribute to the perspective that offsets are a ‘licence to trash’ and that they enable greenwashing, providing cover for damaging projects that might otherwise be subject to greater scrutiny or even not proceed⁷. On the other hand, evidence exists that the prospect of offsets sometimes leads to avoidance of impacts or early-stage project cancellation⁸, and some offsets have achieved important biodiversity outcomes^9,10.

Beyond the controversy over offsets and their effectiveness, the context in which offsets exist is evolving rapidly. First, over the past ten years, the requirement for ‘no net loss’ as the outcome of offsetting has started to be superseded by ‘net gain’¹¹ and ‘net positive’ goals¹². Second, although typically applied to individual project-level impacts, the need to extend the logic to addressing all impacts associated with economic activity, including throughout value chains, is increasingly recognized^13,14. Now, the novel concept of ‘nature positive’^{15,16,17,18,19} looks set to reshape how sectors that traditionally used biodiversity offsetting and the mitigation hierarchy engage with their impacts on nature^14,20,21. The role of offsets in this context is uncertain and contested, providing an opportunity to take stock of the role and potential of offsets as a tool for better biodiversity outcomes, to identify barriers to their effectiveness, and to identify how they must evolve if they are to be a useful part of the transition ahead.

In this Review, we trace the evolution of biodiversity offsetting from its inception to present-day practices, including the recent enthusiasm for net gain approaches in offsetting, broader calls for nature positive goals, and the emergence of biodiversity credit markets^22,23. We summarize the reasons behind the widespread underperformance of offsets in achieving their intended conservation goal of at least no net loss of biodiversity^6,24,25,26, and explore whether and why offsetting is peculiarly prone to failure relative to other conservation approaches, such as the establishment of protected areas and payments for ecosystems services^27,28,29,30. Given the complex sociopolitical environment in which any conservation interventions must operate³¹, we outline approaches that could help to safeguard offset schemes against failure^28,32,33. Looking ahead, we explore how biodiversity offsets interact with, and can contribute to, a nature positive future^34,35.

Box 1 A brief history of biodiversity offsetting

The development of offset policy and practice is most often traced back to the USA in the 1970s; however, key concepts underpinning offsets are much older (see figure). A 2021 review¹⁰⁶ analysed the emergence and spread of biodiversity offsetting discourses through contemporary history, starting with rising environmental concerns in the 1960s (the publication of Silent Spring by Rachel Carson¹³⁷) and the adoption of the US National Environmental Protection Act in 1969 (ref. ¹³⁸). These developments paved the way for the mainstreaming of environmental impact assessment (EIA) and various sectoral policies requiring the mitigation and compensation of environmental impacts, including the US Clean Water Act of 1972 (ref. ¹³⁹), the US Endangered Species Act of 1973 (ref. ¹⁴⁰), and the French Nature Protection Law of 1976 (ref. ¹⁴¹), which was the first such European law, prior to the adoption of the European Union’s EIA Directive in 1985 (ref. ¹⁴²) and the Habitats Directive in 1992 (refs. ¹⁴³,¹⁴⁴). Sectoral policies also contributed to the development of offsetting; for example, India’s Forest Conservation Act of 1980 (ref. ¹⁴⁵) included a requirement for compensatory afforestation¹⁴⁶. Such legislation provided the context in which requirements for offsets and the achievement of no net loss or net gain were later introduced.

Policy momentum built over the following decades, and biodiversity compensation principles were included in the Convention on Biological Diversity (CBD) in 1992. In 2006, the CBD called for guidance on offsets (Decision VIII/17; ref. ¹⁴⁷) and in 2008 proposed offsets as a potential financing mechanism for conservation (Decisions IX/18 and IX/11; refs. ^148,149). Bilateral and multilateral financial institutions also began to impose no net loss and net gain requirements — specifying the use of offsets in achieving these aims — on developers seeking project finance, for example, the International Finance Corporation (IFC) in 2012 with Performance Standard 6 (ref. ⁶⁴) and the World Bank in 2016 (ref. ¹⁵⁰). Through this period, carbon offset mechanisms were also gaining prominence, with the UN Framework Convention on Climate Change (UNFCCC) introducing REDD+ (Reducing Emissions from Deforestation and Forest Degradation) in 2013.

As of 2019, biodiversity offsetting (mandatory or voluntary) was part of public policy in more than a hundred countries^2,151. Reference to no net loss is increasingly being replaced with language such as net gain and net positive, and calls are increasing to apply the mitigation hierarchy, including offsetting, to impacts incurred throughout value chains, beyond direct and indirect project footprints^13,14. The need to better align outcomes from offsets with jurisdictional targets for biodiversity is increasingly recognized, such as in the context of spatial planning or strategic environmental assessment¹¹¹.

At the same time, unfavourable perceptions of biodiversity offsets, heightened by criticisms of carbon offsets, have accumulated. In some cases this has resulted in emerging conservation discourses, such as those focused on nature positive and biodiversity credits, moving away from the traditional language of offsetting. Data in the Box 1 figure were taken from refs. ^2,13,⁶⁴,^{106,139,140,147,148,152}.

The textbook case for biodiversity offsetting starts from accepting that at least some project development involving negative biodiversity impacts is necessary and desirable for improving people’s well-being. This approach recognizes that, despite best efforts to avoid and minimize the impacts of such development and to restore affected biodiversity, some residual impacts might remain (Fig. 1). The only way for such development, with its residual impacts, to proceed without further eroding biodiversity is to ensure that the impacts are at least fully counterbalanced by equivalent biodiversity gains elsewhere, to achieve at least no net loss of affected biodiversity. Such equivalent gains could be achieved by, for example, restoring a degraded site or protecting a site from threats. Equivalence requires not only that the benefits from these actions accrue to the same species, ecosystems or other biodiversity components negatively affected by the proposed development, but that the benefits are at least as large as these impacts and persist for the same duration (often, in perpetuity). This section outlines how biodiversity gains are delivered on the ground through various mechanisms, and how the definition of success varies depending on important differences between alternative framings of the goal of offsetting as either a relative net outcome or an absolute net outcome. Finally, this section outlines the range of approaches to governance and implementation of offsets.

The mitigation hierarchy is a longstanding framework for mitigating impacts on biodiversity through a sequential series of steps: avoid, minimize, restore and offset. Potential impacts from a development project must first be evaluated, and any negative impacts first avoided and then minimized as far as possible. If some impact on biodiversity is expected to remain after these preventative steps, the two remedial steps must be used: first, repairing or restoring damage on-site and only then, as a last resort, seeking to offset any residual impacts through actions that benefit the same biodiversity at an off-site offset area. The net outcome is required to be at least no net loss, or, increasingly, net gain.

Full size image

Biodiversity offsetting practices

In practice, offsets rely on implementing conservation or restoration interventions, in targeted locations, to generate the measurable gains they aim to deliver for biodiversity. The types of intervention broadly aim either to prevent declines of biodiversity that would otherwise have occurred (‘averted loss’ or ‘avoided loss’ offsets) or to improve species populations or other components of biodiversity over time (‘improvement offsets’ or ‘restoration offsets’)³⁶. Some offsets combine interventions that aim to achieve both avoided loss and improvement gains.

Averted loss offsets involve mechanisms to decrease ongoing pressures or threats to biodiversity, through legal protection (for example, establishing a protected area) or working with local people, governments and businesses to stop or change harmful land use or natural-resource-extraction practices (for example, reducing deforestation, hunting pressure or harvesting of wild plants). These mechanisms often require alternative livelihood schemes or compensation payments, alongside conservation agreements or public–private–community partnerships³⁷. Effective enforcement of such agreements or partnerships is necessary for success. The role of government in such arrangements is highly variable, depending on local land-tenure regimes (for example, whether a developer can buy and privately own land for its offset needs) and governance of natural-resource use (for example, hunting or fishing rights). Because offsetting requires changes in human behaviours, the social and economic impacts on affected people need to be carefully considered and addressed³⁸.

Restoration interventions are focused on proactive repair and improvement of degraded ecosystems by removing barriers to natural processes (for example, by removing dams, dykes or drains to restore rivers or former wetlands) and enabling native vegetation to re-establish itself (through topsoil stabilization, afforestation and planting, or through assisted natural regeneration), or even recreating whole habitats (such as breeding ponds for amphibians). Restoration offsets can often require invasive species to be actively removed or managed.

Importantly, effective offsets address the appropriate processes or limiting factors that enable the targeted biodiversity to respond. Offset interventions must therefore be designed and implemented according to the species, habitat and/or ecological processes they target, and the baseline conditions at the offset site. As such, considerable information is often required to design an appropriate offset and, in many cases, that information is not available³⁹. Such knowledge gaps concerning effective conservation or restoration approaches — especially when working with already rare and highly threatened species or habitats — are a strong indication that offsets are not an appropriate response to potential impacts on those biodiversity features, and that avoidance is necessary. Offsets are simply not always feasible.

The fundamental requirement of an offset is that it generates a measurable benefit, or gain, for the affected biodiversity on the ground. As such, although actions such as research or monitoring might be necessary to enable or evaluate the performance of an offset, these actions alone do not constitute an offset according to best practice⁴⁰. Further, the reintroduction of locally extinct species and the restocking or genetic rescue of severely depleted populations might be appropriate as an offset, but captive breeding alone would not qualify.

Finally, for losses and gains to be equivalent, the gains must endure for at least as long as the loss. As most losses are permanent, offset sites tend to require some form of in-perpetuity protection (and often, ongoing funding) if they are genuinely to achieve at least no net loss^41,42. Although various mechanisms for achieving these outcomes exist (such as protected area designations, easements, on-title conservation agreements and trust funds), these are not always put in place. In principle, however, securing offset sites for the long term can mean they are suitable as Other Effective Area-Based Conservation Measures and, therefore, could contribute towards achieving international targets for protected and conserved area coverage^43,44, although using offsets to achieve such targets can introduce perversities^44,45.

Relative versus absolute net outcomes

Although the goal of at least no net loss appears to be straightforward, in practice its meaning can vary substantially depending upon the frame of reference^46,47 (Fig. 2). Offsets have typically required no net loss, or net gain, to be achieved, relative to the outcome if neither the impact nor the offset had occurred. Owing to widespread, ongoing biodiversity loss, this ‘counterfactual scenario’ is generally presumed to be one of biodiversity decline (Fig. 2a). This assumed decline enables averted loss benefits to be generated through the prevention of future damage to biodiversity that already exists⁴⁸. As such, a successful offset exchange that achieves no net loss across impact and offset sites can result in less biodiversity after the impact than before — if the loss in biodiversity from the impact is equal to or less than the loss that would have occurred at the offset site, had it not been made an offset⁴⁷.

a, No net loss and net gain, relative to a counterfactual scenario of continuing biodiversity loss. b, No net loss and net gain as absolute outcomes over time, relative to a fixed baseline.

Full size image

The use of counterfactual scenarios poses many challenges to the practice of offsetting^25,49. For example, the benefit at an offset site (which must be sufficient to at least counterbalance the loss at an impact site) is calculated as the difference in biodiversity outcomes between two scenarios: the scenario in which the offset action occurs, and the counterfactual scenario in which the offset action does not occur⁵⁰. Estimating this benefit therefore involves an assumption about a non-observable, hypothetical trajectory of biodiversity⁵¹. Ideally, this assumption ought to be drawn from the recent trajectory of biodiversity at comparable sites, adjusted for any site-specific constraints⁵²; however, counterfactual scenarios are often not established appropriately⁵³. To further complicate matters, estimating a counterfactual scenario requires consideration of effects beyond the offset site itself. For example, if the habitat at a potential offset site had a 50% chance of being destroyed if not protected as an offset, then the additional benefit of protection might be 50% of the value of the site. However, if the hypothetical destruction of that site would itself have been required to be offset, the net benefit would be zero^47,52. Clearly, an incentive exists to overstate the counterfactual decline at, or risk to, a potential offset site, as doing so would increase the estimated benefit from that offset and thereby reduce the size of an offset obligation⁵⁴.

Offset approaches that aim for relative no net loss or relative net gain can be valid, although challenging to design and manage. These approaches aim only to neutralize the impact attributable to the project, without regard to any wider goals or targets for biodiversity conservation. Net outcomes from such offsets, across impact and offset sites, can therefore be misaligned with broader policy goals such as nature positive approaches, in which biodiversity must be improved in absolute terms over time⁵⁵. To address this misalignment, alternative approaches to counterfactual-based offsetting exist. For example, offsets can require ‘absolute’ net outcomes relative to fixed baselines (Fig. 2b) rather than dynamic counterfactual scenarios. Target-based ecological compensation is an approach that aligns offset outcomes with the absolute outcomes desired for the wider landscape⁵⁵. This approach is less frequently used, although offset policies in South Africa, England and Australia’s Northern Territory, for example, require absolute outcomes across impact and offset sites^23,56,57, and the latter two require absolute net gain. The key difference between relative and absolute net gain policies is that the latter cannot generate gains from averting losses, only from recreating biodiversity or improving its condition at the offset sites.

Biodiversity offset governance and implementation

Offsets require that the costs of compensation fall upon those responsible for the damage, consistent with the ‘polluter pays’ principle^58,59. Offsets also incentivize the avoidance and minimization of impacts on biodiversity by internalizing the cost of the compensation⁸ and, in theory, setting limits on impacts based on what is considered ‘offsettable’^55,60. The concept of offsetting is well established in the context of pollutants (such as greenhouse gas emissions) and has been extended to include no net harm to people affected by the development and accompanying offset^37,38,61,62.

Biodiversity offset systems have a variety of forms⁶³. Some jurisdictions have no government-mandated offset requirements, but offsetting might still occur voluntarily or under requirements set by financiers of developments^28,63,64. However, at least 37 countries have laws that require or enable offsets for some types of biodiversity impact, and over 100 have provisions in some form for offsetting².

Offsets are generally delivered through three broad pathways (Fig. 3): (1) developer-led pathways, in which the developer directly provides an offset through activities on its own land or through individually negotiated contractual arrangements with a third party; (2) market-based pathways, in which (ideally) pre-certified units of biodiversity gain can be purchased by a developer (sometimes multiple credits from single sites) to acquit their offset obligation; and (3) in-lieu fee pathways, in which developers pay a required amount into a trust fund, to which the obligation to procure the required biodiversity gains is transferred. Many systems involve a mix of these approaches or hybrids, and the involvement from private sector brokers and states as a buyer or seller varies⁶³ (Fig. 3). In the case of government-mandated offsets, the general role of the state is to mandate that certain biodiversity impacts must be compensated, enabling willing land managers to sell biodiversity gains to buyers who cause (or have assumed responsibility for) residual biodiversity loss. The state also sets relevant exchange rules by which trades must abide (such as equivalence requirements)⁶⁵ and regulates the trades, often through a statutory body such as a government department. Ideally, the state also ensures adequate monitoring and compliance, and transparent reporting on net outcomes for biodiversity.

Legislative frameworks and/or policies set the requirements for offset provision and, sometimes, the way in which they are delivered. Many delivery pathways are possible and vary among jurisdictions. A project developer responsible for impacts incurs the obligation to provide an offset. This may be delivered directly on land controlled by the developer, but offsetting is more often delivered by a third party. Connecting the developer to this offset provider might involve several types of intermediary, who provide services ranging from information about potential offset providers through to a platform through which the developer can purchase existing offset ‘credits’ already generated by a provider to meet their offset obligation. In some cases, the developer can pay into a trust, which then incurs the obligation to deliver the required offset. Regardless of the delivery mechanism, the regulator or agency responsible for the overarching offset regulation or policy ideally monitors and enforces compliance with the policy, although the extent to which this occurs in reality is often limited.

Full size image

Over the past two decades, biodiversity offsets have become a contentious part of conservation policy. Offsets can be contested on a range of grounds: those subject to offset requirements might see obligations as too onerous or inefficient; others see offsets as ethically dubious, reducing nature to a commodity to be traded⁶⁶. These perspectives each require a focused discussion, which is outside the scope of this Review. Rather, this section considers a third major category of critique: the extent to which offsets are effective in achieving their stated goal of (at least) no net loss of biodiversity. Myriad factors can limit offset effectiveness³ (Supplementary Table S1, Supplementary Information), although changes and safeguards to address these limitations are possible.

The consistent underperformance of biodiversity offsets against their stated objectives can be due to poor policy design (that is, even when implemented perfectly in line with policy the offsets would not achieve the required net outcomes for biodiversity⁴⁷), and compliance or implementation failures. Often, both shortcomings are present, and most associated issues (Table 1) can emerge at either the design or implementation stage. Here we consider the effectiveness of biodiversity offsetting in terms of (1) the extent to which offsets generate gains at the site level; (2) the extent to which offset trades achieve their stated goal of at least no net loss across the impact and offset sites (the exchange level); and (3) the potential broader net effect on biodiversity of the inclusion of offsets in the policy mix (the system level).

Site-level outcomes

Most assessments of ecological outcomes of biodiversity offsets at the site level have concluded that the evidence available to evaluate offset performance is very limited^{5,6,28,67,68,69,70,71,72}. As a result, offset outcomes have rarely been evaluated relative to a robust counterfactual scenario to explore whether the biodiversity gains delivered by the offset would not have been delivered in the absence of the offset. Centralized offset registers or databases are crucial to enable the effective tracking, management and transparency of offsetting activities, ensuring accountability and facilitating informed decision-making by regulators, stakeholders and the public⁷³. For example, RIBITS (Regulatory In-Lieu Fee and Bank Information Tracking System) has a pivotal role in monitoring and managing mitigation banking activities across the USA, providing valuable centralized data for regulatory compliance. However, the effectiveness of RIBITS is hindered by issues such as inconsistent data quality, complex and differing regulatory requirements, and missing files or reports, which collectively contribute to a reduction in transparency and accessibility of this otherwise instrumental system^71,74.

Where robust evaluations of offset performance have been conducted, averted loss offsets (the majority of the offset system) have been found to deliver no or little additional biodiversity gain in New South Wales²⁵ and Victoria²⁶ (Australia), respectively, although offsets targeting increases in vegetation cover starting with low baselines did deliver additional increases. Additionally, the majority of wetland area gains in the USA wetland mitigation banking system were found to have been unlikely to occur in the absence of offsetting⁶⁹, but the ecological quality of those compensatory wetlands could not be assessed. In some cases, offsets may even be harmful; some species conservation banks (used to compensate impacts on threatened species) in California were concluded to potentially have prevented improvements that would probably have occurred in their absence⁷⁵.

Exchange-level net outcomes

Even when outcomes at an offset site can be evaluated, assessing the net outcomes associated with the offset exchange might not be possible. This discrepancy can occur because, even if outcomes at offset sites are discoverable, the impacts being offset are not known or linked to the offset site, or differing methodologies or metrics are used for assessing impacts and offsets²⁶. Such traceability challenges mean that few examples exist of comprehensive evaluations of net outcomes. One of the most comprehensive recent evaluations showed that a large mine in Madagascar was on track to deliver no net loss of forest through its offsets⁹. This was achieved through averted loss offsets and was only possible because the background rates of deforestation in the region were very high, so the overall loss of forest over time was similar to what was expected in the absence of the mine’s activities. Net outcomes for the numerous and often endemic species of animals and plants that were targeted by the mine’s biodiversity action plan could not be evaluated. Similarly, the ecological equivalence of wetland losses and gains in the USA could not be assessed, but the area of wetlands established fell short of that lost by on average 1,600 acres per year⁶⁹.

System-level outcomes

Beyond the direct effects of offsets, their presence in the policy mix has been suggested to potentially undermine impact avoidance. Broadly, two strands of empirical evidence support this ‘licence to trash’ argument. The first suggests that offsetting has quickly become the norm over impact avoidance in several systems where it has been applied (for example, Australia and North America^24,76). The second has demonstrated that introducing the option of offsetting into the impact-mitigation process increases the scope for negotiation between project proponents and planners⁷⁷, suggesting that some projects which would previously have been rejected outright might instead be considered favourably in the planning system. On the other hand, the prospect of offsets probably incentivizes early impact avoidance that might never be detected or attributed, because it occurs prior to a formal project proposal or through informal discussions during the environmental impact assessment (EIA) process⁷⁷. Allowing project proponents to try a variety of project plans in software that allowed them to calculate their offsetting liability associated with different spatial configurations of development was shown to result in final submitted plans associated with substantially lower biodiversity impacts than their initial plans⁸. Additionally, two studies that tested for changes in deforestation rates or development approvals coinciding with the introduction of offset policies (in New South Wales, and Florida) found no association between the introduction of offsetting policies and rates of ecological harm^25,78.

Given the widely recognized challenges to offset success (Table 1) and frequently reported failures to achieve their objective, it is tempting to conclude that perhaps offsets are not worth persisting with. However, before reaching this conclusion, it is worth exploring whether offsets are any more peculiarly prone to failure than other types of conservation mechanism — and if so, what the potential reasons are for this discrepancy.

Relative susceptibility of offsets to failure

Offsets often fail to achieve their goal of at least fully counterbalancing damaging impacts on components of biodiversity. However, the (appropriately) strict definition of success for offsetting is unusually clear and measurable: the requirement is to achieve at least no net loss, and any outcome that falls short is, by definition, inadequate⁷⁹. Other categories of conservation mechanism, such as establishment of protected areas or payment for ecosystem services (PES) schemes, often have less explicit or more complex criteria for ‘success’⁷⁹, and rarely state a fixed quantitative requirement for additional biodiversity gain. It could therefore be that offsets are similar to other interventions in terms of their typical impact/effect, but that their clear success/failure criterion means that they are more easily judged^71,74.

The lack of quantification of additional benefit from conservation interventions other than offsets is starting to be addressed. Overall, robust counterfactual-based evaluation reveals that most conservation interventions of many types have modest and context-specific impacts^80,81,82,83 and that all fail at least some of the time. In one review, conservation actions were found to improve biodiversity outcomes relative to a counterfactual scenario in about two-thirds of cases; however, no evidence of effectiveness was found for one-third of interventions and effect size for the remainder varied depending on factors such as the degree of targeting for each action⁸². For example, although overall positive effects on forest cover and deforestation reduction have been observed for PES schemes⁸⁴, the effects are heterogeneous, some robust evaluations have found no effect on forest cover⁸⁵, and selection bias frustrates the quality of evidence. Overall, evidence for the effectiveness and cost-effectiveness of PES in achieving environmental gains remains poor^84,85,86,87.

Most robust evaluations of the effectiveness of protected areas have focused on the extent to which deforestation is reduced by protected area establishment. In general, protected areas are found to reduce deforestation by a modest amount; for example, protected areas were estimated to have prevented deforestation across 10% of their extent over 37 years in Costa Rica⁸⁸, and 10.5% of their extent in Queensland over 30 years⁸⁹. Effective management of wildlife populations is another key objective of protected area establishment. Globally, protected areas have been found to result in improvements in about one-quarter of waterbird populations⁹⁰, but declines occurred in a similar fraction of populations. Effective stakeholder engagement and community partnerships have been identified as key factors associated with successful outcomes from protected areas^91,92.

Carbon offset credits through forest protection (for example, through REDD+) are one of the few types of intervention that have a similar quantitative success criterion to biodiversity offsets: these interventions must avoid an amount of deforestation and forest degradation which, if it had occurred, would have resulted in at least the amount of emissions that any credits sold claim to represent. However, independent evaluations suggest that shortfalls of 75% or more are common^93,94. Comparisons of the conservation effectiveness of carbon offsets with that of publicly funded conservation⁹⁵, as well as other comparisons between conservation on private versus publicly managed land⁹⁶, have shown a high degree of context specificity. In general, forest protection for carbon offsets achieves similar ecological outcomes to protected areas⁹⁵, but other land management such as Indigenous stewardship might outperform either public or private ownership^96,97.

Even though offsets (both carbon and biodiversity) might not be unusual in having limited effectiveness, certain features do set them apart from other common conservation approaches. One important distinction is that offsets are intended to deliver a specified amount of benefit that is used to compensate for that same amount of damage⁷⁹. Any shortfall therefore represents not just an opportunity cost, as is the case for a PES scheme or a new protected area that did not achieve the hoped-for biodiversity benefit, but a realized, on-the-ground loss of biodiversity that might not have been permitted to occur had the inadequacy of the offset been known. In combination with the bias towards conservation typically occurring in locations where the additional benefit from the intervention is relatively low⁹⁸, this potential for negative net outcomes increases the prominence and consequence of any shortfall in performance.

Slow progress in offset policy and practice

Perhaps a greater concern than the poor performance of offsets to date might be the lack of evidence of their continuous improvement over time. Although a range of concerns afflict offset policy and practice (Table 1), the solutions to these often appear straightforward, and many can be addressed through policy reform. In general, however, relatively little improvement over time has been observed (Table 1).

Some offsetting systems have improved after decades of policy trial and error. For example, the 2008 compensatory mitigation rule in USA wetland mitigation systems marked an important change from previous approaches, because it made all types of offset provider subject to the same standards and created a level playing field between mitigation types (previously, mitigation banking had been subject to higher standards than developer-led mitigation)^{99,100,101,102}. In France, international experiences were instrumental in improving decades-old legislation by introducing a no-net-loss requirement, rules for equivalence, a public register of offsets, and the option to use government-certified ‘offset banks’¹⁰³. In England’s new biodiversity offsetting system, policymakers learned from established failings of Australian averted loss-based offsetting systems to create a system based on a static baseline²³ (Fig. 2b).

However, this progress has not been universal. Many cases of backsliding exist, such as the downgrading of Victoria’s (Australia) aim of biodiversity net gain to no net loss, and consistent governance failures or under-resourcing of offset systems around the world leading to systemic compliance failure remain^4,104,105. If the only constraint on success of biodiversity offsetting policies was insufficient knowledge manifested in poor policy design, then the performance of offsetting systems would be expected to increase over time to a greater extent than has been observed. A proposed explanation¹⁰⁶ for this lack of improvement is that, when an effective offsetting policy in which the true biodiversity costs of development are fully internalized is proposed, potentially affected sectors organize themselves politically to counter such change. These potential conflicts between conservation and other interests influence the design and implementation of offset systems.

Developers are often heavily involved in offset policy design as one of the key interest groups engaged in consultation¹⁰⁵, and are therefore able to influence the direction of policy and its implementation. Important motivations for developers include reducing resistance to new development and simplifying the planning system¹⁰⁴, which are often at odds with the needs of a robust offset system. Offset procurers also have a vested interest in offsets not being costly; however, monitoring and enforcement costs comprise a substantial proportion of the overall costs of offsetting, and are necessary to ensure that the promised environmental outcomes are truly being delivered.

Ultimately, the combination of these political economy issues with the evidence of low effectiveness of biodiversity offsetting policies means that offsetting cannot be relied upon as the sole, or even the main, solution for navigating trade-offs between economic expansion and the conservation or recovery of biodiversity. Even if the challenges outlined in Table 1 were to be addressed, in many jurisdictions, insufficient land is available to fully offset future infrastructure development plans¹⁰⁷. Avoidance of biodiversity impacts therefore remains the most critical component of the solution^49,108, and whether offsetting policies incentivize or hinder this (by legitimizing development) in the long term remains unclear.

After decades of increase in the prevalence of biodiversity offsets, their future now appears to be at a crossroads. An increasingly negative view of offsets is starting to lead to rejection of their validity, but also to increasing focus on related but distinct emerging concepts, terminologies and frameworks. Notable among such concepts is nature positive — an outcome-based societal goal to “halt and reverse nature loss by 2030 on a 2020 baseline, and achieve full recovery by 2050”¹⁰⁹. The vision of nature positive, consistent with that of the Kunming–Montreal Global Biodiversity Framework (GBF) adopted in 2022, is that the decline of nature will be not just slowed but reversed, so that more nature will exist in the future than today. The concept has been embraced not only by civil society organizations, but also by government and business, and pledges to contribute to nature positive are mounting^35,110.

Offsets in some form appear to have an essential role in a nature positive future, because the need remains to address legitimate residual biodiversity losses after rigorous application of the mitigation hierarchy (Fig. 1). Indeed, offsets could also have a proportionally larger role if used to expand the scope of losses being mitigated (for example, along corporate value chains¹⁴) and/or if they make meaningful contributions to biodiversity outcomes beyond no net loss¹¹¹. Target 19 of the GBF¹¹² and other reports cite offsets as a source of funding to support biodiversity recovery and a potential substantial contributor towards the current biodiversity funding gap¹¹³. However, the extent to which offsets will actually contribute to halting and reversing global biodiversity loss is unclear. A new wave of guidance, terminology and actors could instead result in the replacement of offsets by alternative actions that appear similar to offsets, but in reality enable an exemption from the strict outcome requirements and well established standards expected of (if not achieved by) offsets³⁵ (Table 1).

A spectrum of guidance on the role and use of biodiversity offsets is emerging to support private sector alignment with nature positive, in the form of biodiversity-related performance standards and risk disclosure frameworks. This guidance ranges from a complete exclusion of offsets from acceptable ‘nature-positive-aligned’ actions¹¹⁴, to agnosticism about whether and when offsets are acceptable¹¹⁵, to clear guidance on how to successfully implement offsets¹¹⁶. A particular area of interest is the expansion of compensation beyond direct, site-level footprints to the biodiversity impacts of a business’s value chain. However, a major challenge is that current life-cycle assessment methods used to determine business value chain impacts tend to produce aspatial outputs or outputs with very coarse spatial resolution, in biodiversity units that cannot currently be practically translated into the project-scale biodiversity gain measures required for site-scale offsetting¹¹⁷. Ongoing research is exploring how to establish equivalence for diffuse value chain offsets¹¹⁸.

Perhaps the area of greatest ambiguity relates to the rapidly growing interest in the potential for biodiversity credits (or certificates)¹¹⁹ to finance conservation and the improvement of biodiversity¹²⁰. The hope is that industry commitments to nature positive will drive the emergence of voluntary markets in biodiversity, in which business will seek to purchase biodiversity credits to demonstrate improvements in environmental performance, particularly if they must also disclose environmental damage¹²¹. However, despite obvious parallels, some proponents of biodiversity credits have sought to distance them from offsets^122,123,124. Some guidance explicitly states that use of biodiversity credits does not constitute offsetting¹²⁵, although some still use the language of net outcomes (implying a balancing of losses and gains)¹²⁶. Others state that purchased credits should not replace compliance-based offsets, but that credits could potentially be used for compliance offsetting depending on local jurisdictional requirements^127,128.

Suggestions have been made that the lowest-risk use of credits is to mitigate biodiversity losses along value chains when the type and amount of losses are unknown or unknowable^35,129. Such a use could not strictly be considered offsetting, as the net outcome cannot be known. Regardless, most demand for biodiversity credits will probably be driven by some form of counterbalancing of negative environmental impacts, whether or not this counterbalancing is referred to as offsetting. However, this conceptual ambiguity carries the risk that true offsets (that deliver biodiversity gains equivalent in type, amount and duration to the biodiversity losses caused by impacts) will be quickly replaced by other conservation actions that might deliver some gains but are unlikely to deliver the correct amount or kind, and without the traceability essential to support biodiversity outcome claims, such as no net loss, net gain and nature positive outcomes³⁵.

Contrasting trends are visible: on the one hand, increasing discomfort around offsets and their negative associations; and on the other, increasing interest in offset-like actions and terminology proposed as a solution to nature decline. In part, this disconnect might be consistent with the increasingly diverse backgrounds of those involved in bridging the worlds of biodiversity and business, many of whom have backgrounds not in ecology and conservation science, but rather in business administration, economics or political science²⁶. Others bring experience from the carbon markets, which face many of the same challenges as biodiversity markets, although biodiversity markets produce substantial new complexities. This growing diversity of actors engaging with the challenge of nature recovery should be beneficial, but also carries risks. Loss and dilution of institutional memory can lead to over-simplification of the complexities and wickedness of nature positive goals, and institutional amnesia¹³⁰. It is crucial that we learn from the past.

The future of offsets

The future of offsets in the context of nature positive and other shifts is uncertain. Alternative futures for offsets and biodiversity exist (Box 2), and decisions by business, governments and civil society will help to determine which future emerges.

In the best scenario, the persistent problems with offsets (Table 1 and Supplementary Table S1) are addressed, and newly emerging risks associated with dilution of offsets by cynical interpretation of nature positive goals and biodiversity credits are forestalled through strong integrity requirements and guardrails^35,131. The increased engagement of a wide range of sectors in the nature conservation discourse, coupled with genuine attempts to learn from the mistakes of the past, could prove a welcome first step towards reforming the role of offsets in achieving the genuine transformative change called for under the GBF. At worst, the pursuit of nature positive could become a flash in the pan, undermined by greenwashing, and could set back progress and decades of lessons learned on high-integrity implementation of the mitigation hierarchy and offsets.

Biodiversity offsets in practice suffer from many of the same flaws and challenges as other conservation interventions. Most of these flaws can, in theory, largely be addressed. However, some intrinsic factors make improving offset practice particularly difficult. A fundamental tension exists between instituting necessary controls on integrity and respecting the limits to the use of offsetting as a response to environment–development conflict, and the reality of the political influence of those with an interest in facilitating development projects with minimal impediments⁴¹. In this context, the recent raft of frameworks seeking to mainstream consideration of nature conservation through mechanisms such as biodiversity credits should be viewed with caution as well as optimism.

Research could support refinements to current practice in several ways. Understanding the limits to what can be offset is increasingly important — certain populations, ecosystems and places are fundamentally irreplaceable and should not be allowed to be negatively impacted, as no offset is possible. The need for further development of methods to evaluate and attribute impacts on biodiversity through entire value chains is increasingly recognized¹³². Similarly, a better understanding is required of which products, activities, and value chains can be substituted to enable avoidance of biodiversity loss, particularly where such losses cannot be addressed through offsets. However, in most cases, most of the solutions to the challenge of poor offset performance are well known, and the required actions have been identified for more than a decade (Supplementary Table S1). The fundamental problems that prevent continuing improvement in biodiversity offsetting are generally not those that technically focused research can address. Instead, the problems stem from resistance to genuine reform owing to conflicting motivations and values, and the associated power dynamics^106,133.

In light of these fundamental and continuing obstacles to obtaining improved outcomes from offsets, the most basic and necessary change requires widespread improvements in environmental regulation: simply preventing further conversion of natural systems, especially those of high environmental value^21,134,135. However, any such changes encounter fierce — and usually effective — resistance from the same actors that frustrate improvements in offset policy. Furthermore, legitimate and desirable human development needs, and even associated environmental goals (such as the shift to renewable energy), will be almost impossible to meet without causing at least some damage to biodiversity. Eliminating offsets as a response to these residual impacts risks shifting the responsibility for addressing this damage away from the proponents to broader society, violating the polluter-pays principle. Simple taxes on environmental damage might enable more ‘efficient’ investment in environmental benefits than insisting on like-for-like replacement of losses, as well as disincentivize impacts. However, if set high enough to disincentivize damage, such taxes would probably also be strongly opposed¹³⁶. A risk also exists that such approaches would become an avenue for the loss of ecologically irreplaceable ecosystems and habitat in exchange for those that are easily replaceable³⁵ — or worse, could simply replace existing environmental expenditures⁵⁴. In short, none of these alternatives is a promising replacement for biodiversity offsetting.

Despite inherent flaws, the mitigation hierarchy, with offsetting fixed firmly in its appropriate place, remains key to navigating the complex landscape of environmental challenges. Its transparent, quantitative outcome requirements and well established guidelines set the mitigation hierarchy and biodiversity offsetting apart from other conservation approaches. No obvious alternative can replace offsets without introducing an uncertainty about outcomes and accountability that would also enable greenwashing. However, acknowledging the continued role of offsetting does not imply passive acceptance of the serious flaws in how offsets are currently used; instead, it underscores the need for continued efforts to improve offset policy and practice²¹. Our focus should shift toward strategies that reduce reliance on offsets, and that use properly designed and fully costed offsets as the disincentive they ought to be.

Despite the mounting evidence of the poor performance of offsets and the systemic pressures and interests that impede their improvement over time, plausible and appealing alternatives are few. Furthermore, despite several emerging frameworks for mainstreaming nature that explicitly or implicitly reject offsets, most of these frameworks nevertheless rely on offsets to succeed. Therefore, the unceasing task of improving rigour and integrity in offset policies and markets remains critical to achieving anything close to a nature positive future.