Voluntary Carbon Offset Markets Are in a Credibility Crisis, and Buyers Are Quietly Switching to Direct Removal

Voluntary carbon offset markets, where companies purchase credits representing a claimed reduction or avoidance of greenhouse gas emissions elsewhere to counterbalance their own, spent years growing on the strength of low-cost credits from forestry conservation and avoided-deforestation projects. A sustained sequence of investigative analyses and academic studies examining those project categories found that a substantial share of credits issued did not represent the emissions reduction claimed — forests that would likely have been preserved anyway, baseline assumptions calibrated to maximize credit issuance rather than accurately reflect a counterfactual, and monitoring that in some cases failed to catch project reversals like wildfire or logging that erased the claimed carbon storage entirely. The credibility damage from these findings has been broad enough to affect the market beyond the specific projects and registries named in any individual investigation, because the underlying problem — verifying a counterfactual that, by definition, cannot be directly observed — applies to the entire avoidance-based credit category, not just the specific instances that happened to get scrutinized. Corporate buyers who had built public sustainability claims on offset purchases found themselves facing accusations of greenwashing regardless of whether their specific credits were among the ones shown to be flawed, simply by association with a category that had lost general credibility. The market response has been a shift in composition rather than an abandonment of offsetting entirely: corporate buyers, particularly larger companies with public net-zero commitments and the scrutiny that comes with them, have moved purchasing volume away from cheap avoidance-based forestry credits and toward smaller volumes of engineered carbon removal — direct air capture, mineralization, and similarly verifiable, physically measurable removal methods — that cost substantially more per ton but carry a fundamentally different verification profile: a ton of carbon dioxide captured and stored can be directly measured, unlike a ton of deforestation that a project claims it prevented. That shift has not solved the volume problem it created: engineered removal capacity remains a tiny fraction of the volume avoidance-based forestry credits could supply at their claimed (if disputed) scale, and corporate buyers who want their offset portfolio to be both credible and large enough to match their actual emissions face a real supply constraint that cheaper, more scalable, less verifiable credits do not have. Several buyers have responded to that gap not by returning to the discredited cheaper category but by lowering the total volume of offsets they claim and shifting more of their public commitments toward direct emissions reduction targets instead — a smaller and more defensible claim, but a genuine one, rather than a large claim resting on credits an increasing share of the market no longer trusts.