Carbon offsets are instruments that allow an entity to compensate for its emissions by financing a reduction or removal of carbon dioxide elsewhere. This is a transaction that typically occurs in two types of markets: regulated compliance markets, in which governments or international agreements (like the Kyoto Protocol's Clean Development Mechanism) allow companies to use offsets to meet legal emissions caps, and voluntary markets, where companies and individuals buy credits to offset their emissions outside of a legal obligation, often for corporate social responsibility or "net-zero" claims. In theory, it is described as highly efficient. However, in reality this mechanism rests on a fragile foundation.

This article explores the critical discrepancy between, on the one hand, the promise of carbon offsets and, on the other hand, their practical performance, revealing how structural flaws can lead to increased emissions and how we must rethink their limited role in climate policy.

I. Why the economic promise of offsets fails in practice

A. The theoretical appeal

A simple economic intuition underlies carbon offsetting: global emissions should be reduced where it costs the least. Therefore, allowing companies located in high-cost jurisdictions to finance their reductions in developing countries or in sectors with less expensive absorption capacity should, in principle, minimize the overall cost of emissions reductions.

Following this logic, a regulated emitter can thus increase its own emissions if it purchases credits that reflect an equivalent reduction achieved elsewhere. In this sense, the mechanism also aims to equalize the marginal costs of emissions reduction across regions, which therefore enables a given level of global mitigation to be achieved at a lower cost.

This is the reason why offsetting mechanisms have played a major role, whether in compliance markets—such as the Clean Development Mechanism (CDM)—or in the rapidly expanding voluntary carbon markets. In abstract models of perfect information and application, compensation without any competition improves efficiency without harming environmental integrity.

B. The core fragility: additionality

Any functional compensation system relies on the principle of additionality, which means that the recorded reductions must be the actual result of the compensation payment. A project that would have been completed regardless of the payment—because it would already be profitable, or even because it is mandated by policy or technologically unavoidable—if it is the subject of a credit, results only in a net increase in global emissions. The explanation behind that phenomenon is that the credit buyer emits more even though the seller does not reduce emissions any more than they would otherwise have.

However, guaranteeing additionality is very difficult in real markets. This is due to three classic information and incentive problems that systematically undermine it:

- Adverse selection: economically attractive projects are overrepresented among credit applications, crowding out activities that are genuinely additional.

- Moral hazard: Baseline projections can be manipulated or even inflated by companies seeking loans or delaying investments while awaiting compensation.

- Unverifiable baselines: The counterfactual scenario (which answers the question "what would have happened without the project?") is by definition impossible to observe, and therefore relies on highly subjective assumptions.

In this sense, since additionality stems from a counterfactual scenario that, by its very nature, cannot be directly observed, it remains both conceptually fragile and empirically contested, thus constituting the central point of divergence between theoretical models and the implementation of compensation mechanisms.

C. Structural flaws that push systems toward non-additionality

While additionality remains a conceptual challenge, the structural characteristics of offset markets also systematically encourage the over-approval of non-additional projects.

1. Information asymmetry

While project developers possess private information on project profitability, expected returns, and baseline emission trajectories, regulators and auditors rely on disclosed (not just held) data and unverifiable projections. This imbalance also makes it systematically difficult to eliminate non-additional projects.

2. Diverging incentives

Whether they operate under compliance systems or voluntary markets, auditors and registries generally receive their compensation from project developers. It is their financial incentives that promote project approval and increase credit volumes, a system that weakens the oversight role upon which environmental integrity depends.

3. Uncertainty of References

References are based on hypothetical business-as-usual scenarios (which are unobservable). Since these scenarios are inherently uncertain, an optimistic/exaggerated projection can easily generate credits that do not correspond to actual emission reductions.

In the so-called voluntary market, several high-profile cases (including credits certified by Verra) have been heavily criticized for relying on deforestation scenarios that significantly overestimated the threat of forest loss. Consequently, when subsequent satellite data proved that actual deforestation was well below projections, it also demonstrated that many of the issued credits did not correspond to emission reductions (or at best, to insignificant reductions). If this type of case does not stand for the entire market, it remains a perfect illustration of how structural uncertainties around reference scenarios can inflate the supply of credit, with detrimental consequences for the overall level of emissions.

II. Empirical evidence: when offsets increase rather than reduce emissions

‍A. India: large scale misallocation in the CDM¹

Empirical analyses of the Clean Development Mechanism (CDM) often reveal significant misallocation of credits, as demonstrated by the Indian wind power sector. The BLIMPs (Blalantly Inframarginal Projects) methodology identifies projects that are significantly more profitable than comparable unsubsidized alternatives, based on observable criteria (capacity, wind resource quality, proximity to the grid, etc.). More than half of the wind power projects registered under the CDM meet this definition, showing that they would have been economically viable without carbon offsetting.

The extent of this non-additionality is considerable: approximately 27 million tons of CO2 have been credited to BLIMP projects, which is as much as the annual emissions of several coal-fired power plants.

The institutional consequences are striking. Compared to a hypothetical random allocation mechanism, the CDM approval process is less efficient: a simple lottery would have allocated fewer credits to non-additional projects, a result that underlines the systemic weaknesses of the project selection process as well as the limitations of a project-by-project evaluation within the framework of a large-scale international program.

B. China: technology upgrades that enabled more output and more emissions²

Data on Chinese manufacturing firms participating in the CDM show that, contrary to expectations, offset projects can lead to higher emissions at the firm level. Indeed, on average, CDM participation resulted in a 49% increase in firm emissions over a four-year period. This differs significantly from the reductions anticipated in the project documentation, raising important questions about the structural effects of offset-financed investments.

Two mechanisms can explain this result:

- The selection effect: Firms already experiencing rapid growth are more likely to apply for CDM accreditation, demonstrating that participation is correlated with increased emissions even before the project is implemented.

- The scale effect: While technological improvements financed by the CDM have increased emissions intensity, these efficiency gains have been reinvested by companies in expanding their production. The increase in total production leads to a rise in absolute emissions.

Consequently, the net impact on overall welfare is negative, meaning that the credits initially intended to offset emissions are now subsidizing business expansion, paradoxically contributing to an overall increase in emissions since the growth of companies in emission-intensive sectors is now being stimulated.

C. Voluntary markets: green claims without robust verification³

Voluntary markets also exhibit many of the same structural weaknesses observed in compliance systems. Forest carbon offsets, in particular, have demonstrated persistent benchmark inflation, concerns about their sustainability, and limited monitoring capabilities. Some studies have shown that deforestation projections used to generate credits are significantly higher than actual deforestation rates.

These discrepancies do not necessarily reflect deliberate manipulation but rather the inherent uncertainty in constructing counterfactual land-use scenarios. Nevertheless, they inevitably contribute to a harmful proliferation of credits, whose environmental integrity is compromised, as they allow companies to claim an illusory carbon neutrality devoid of the corresponding overall benefits.

"In some forest conservation projects (such as the Kariba REDD project, Zimbabwe), independent evaluations have found that the expected benefits for communities—such as improved livelihoods or income sharing—are not materializing to the extent described in the project documentation. While these shortcomings do not constitute a criticism of the concept of compensation itself, they do highlight, at the very least, governance and implementation problems that could significantly affect both the local effectiveness and the reliability of the verification process⁴."

III. Rethinking offsets: what they can (and cannot) dot for climate policy

‍A. The global offset architecture needs guardrails—not abandonment

While empirical data shows significant integrity issues in compliance and voluntary carbon offset markets, it does not necessarily advocate for the outright abandonment of offsetting. Indeed, recent governance initiatives emphasize the need for a more rigorous and, above all, targeted use of this instrument. Following this logic, frameworks such as the Voluntary Carbon Markets Integrity Initiative (VCMI) agree on the fundamental principle: "reduce first, offset last."

This approach highlights the benefits of offsetting only for residual emissions that are difficult to reduce—for example, in sectors where short-term technological alternatives are limited. This safeguard is sought to prevent companies from substituting offsetting for their emissions reduction efforts. In the meantime, it also ensures that offsetting complements—and does not replace—structural decarbonization efforts. By restricting the scope of application of compensation considered legitimate, these frameworks aim to realign the mechanism with its initial logic of efficiency (as touted in its theoretical effects).

B. Policy improvements grounded in economics

The structural vulnerabilities observed in current offset systems can be mitigated by reforms based on classical economic reasoning:

1. Stricter Emission Caps

Designing emission caps that anticipate the presence of non-additional offsets can limit the risk of adverse selection leading to excess emissions. They would also reduce situations in which offsetting is preferred to direct mitigation.

2. Programmatic Approaches

Transaction costs as well as information asymmetry can be reduced by shifting from individual project validation to sectoral or programmatic credit allocation. Indeed, sectoral baseline scenarios are generally more robust and less susceptible to manipulation compared to project-specific counterfactual scenarios.

3. Randomized Assessments

Integrating randomized trials or quasi-experimental comparisons can provide more reliable estimates of true baseline scenarios and help detect patterns of non-additionality. Empirical methods used in energy efficiency programs, for example, could be adapted for carbon offset markets.

4. Realigning Incentives for Auditors and Registries

Reforming the compensation structure for auditors and certification bodies is an essential improvement. Aligning their incentives with long-term environmental integrity—rather than the volume of credits—is a shift that would help to address governance weaknesses that have historically hampered the quality of offsets.

Collectively, these improvements can enhance the reliability of carbon offset markets, although they cannot completely eliminate the underlying challenges of counterfactual measurement and information asymmetry.

C. The inherent limit: offsets cannot replace systemic decarbonization

Even substantial reforms will not be enough to go beyond the inherent limitations of carbon offsetting to become the primary instrument for mitigating climate change. Indeed, it fundamentally is impossible for this tool to compensate for insufficient carbon pricing, regulatory exemptions, or weak national incentives for emissions reductions. Its role remains subsidiary, which means limited to accounting for residual emissions once all possible domestic reductions have been achieved.

While carbon offset markets are based on counterfactual scenarios and complex monitoring frameworks, structural decarbonization, through the deployment of clean energy, for example, or through industrial transformation and strict emissions caps, produces verifiable and, above all, sustainable reductions. In practice, therefore, an increased reliance on carbon offsetting can become a major obstacle by delaying the necessary transitions.

As summarized by a now widely accepted principle: "Reduce what you can; offset only what you must." The underlying economic logic corresponds to empirical data, in that carbon offsetting can slightly reduce the cost of deep decarbonization but cannot in any way replace it.

In theory, offsets optimize climate spending; in reality, their core weakness—proving "additionality"—makes them intrinsically prone to failure. Evidence has shown that reforms have the ability to sharpen this blunt instrument. However, offsets cannot replace systemic decarbonization. Indeed, their role must be strictly limited: they can account for a small fraction of residual emissions, but must never delay the essential work of cutting pollution at its source. Ultimately, verifiable, structural decarbonization driven by clean energy and robust regulation cannot be outsourced.