The House of Cards: Why Forest Carbon Markets Will Never Work

Forest carbon credit markets have a fundamental problem: they’re trying to fix climate change while selling indulgences to polluters. After 25 years of attempts, reforms, and scandals, the evidence is overwhelming that these markets systematically overstate their climate benefits by factors of 5 to 10 times or more. Research from Berkeley, Oxford, Cambridge, and Penn converges on a stark conclusion—the problems aren’t just poor implementation, they’re baked into the structure of the market itself. When California’s supposedly “gold standard” forest offset program has already burned through 95% of its 100-year wildfire insurance buffer in less than a decade, and when 90% of the world’s largest certifier’s rainforest credits turn out to be “phantom credits,” we’re not looking at a market that needs tweaking. We’re looking at a market that was built on sand.

The voluntary carbon market currently processes $1-5 billion annually and was projected to hit $50-100 billion by 2030, but those dreams are collapsing as investigations reveal what researchers increasingly call an “intractable problem.” Companies from Shell and Volkswagen to Disney and Nestlé have purchased millions of worthless credits, allowing them to claim climate progress while continuing to pollute. Meanwhile, the forests these credits supposedly protect are burning, being logged, or were never at serious risk in the first place. The brutal reality is simpler than the complex accounting suggests: you cannot offset the permanent release of fossil carbon by temporarily storing carbon in trees that are increasingly vulnerable to the very climate change those emissions create.

Gaming the System: How IFM Projects Manufacture Carbon Credits From Thin Air

Improved Forest Management projects generate carbon credits by claiming to store more carbon than would have happened without the project. The operative word here is “claiming.” These projects—which account for 193 million credits globally and nearly half of all U.S. offsets—depend entirely on something called a baseline: a hypothetical scenario representing what would have happened to the forest without carbon financing.

Here’s the problem: that hypothetical is entirely made up, and everyone involved has an enormous financial incentive to make it as pessimistic as possible.

The most comprehensive analysis of these projects comes from CarbonPlan’s 2022 study of California’s program, which examined 65 IFM projects representing 102 million upfront credits. The findings were damning: systematic over-crediting of 29.4%, totaling 30 million phantom tons of CO₂ worth roughly $410 million. That’s equivalent to the annual emissions from 8.5 million cars—except these “reductions” never actually happened.

The mechanics of this fraud are almost elegant in their simplicity. Protocols allow projects to set baselines using regional averages across broad areas and different tree species. A project owner with a carbon-dense Douglas Fir forest in Northern California gets to claim credits based on the average for the entire Southern Cascades supersection—which includes much less carbon-dense Ponderosa Pine forests. The project owner receives credits simply for having naturally better trees, not for improving anything. CarbonPlan found that 90% of projects set baseline averages at or within 5% of the minimum allowed by protocol. They gamed the system perfectly.

Even worse is what researchers call the “lumber liquidator” baseline problem. Projects assume they would have aggressively logged without carbon financing, even when there’s zero evidence this would have happened. Barbara Haya’s research at UC Berkeley found that 29% of landowners surveyed were “not confident or unsure” their projects represented additional carbon sequestration. Translation: they were getting paid to not do something they weren’t going to do anyway. One Massachusetts Audubon Society project claimed it could have heavily logged hundreds of thousands of trees—despite being a conservation nonprofit with no intention of ever cutting them down. They still earned millions in credits.

The information asymmetry here is fundamental and unfixable. Landowners know their actual intentions; verifiers and registries don’t. A landowner can harvest aggressively right before enrolling to create an artificially low starting point. Timber Investment Organizations have been documented rapidly logging before selling to carbon developers, manufacturing a “threat” that existed only on paper. And because offset registries, auditors, and developers all earn more money when more credits are issued, everyone in the system profits from over-crediting.

Remote sensing studies have tried to cut through the noise by directly measuring what actually happened on the ground. The results confirm the worst fears: Coffield’s 2022 analysis of 37 California IFM projects found “lack of evidence that the offset program influenced land management.” Projects approaching their 10-year mark showed no difference in harvesting patterns compared to similar forests that weren’t enrolled in offset programs. The credits were generated not by improved management but by creative accounting.

The False Promise of Harvested Wood Products

Forest carbon accounting faces a philosophical tension that protocols have failed to resolve: should we maximize carbon by leaving trees standing, or can we harvest them for wood products that store carbon while allowing young forests to regrow? This isn’t just an academic debate—it determines whether millions of credits represent real climate benefits or accounting fiction.

The harvested wood products argument sounds reasonable at first. Cut trees become lumber for buildings, furniture stays in homes for decades, and regrowing forests sequester carbon faster than mature ones. Meanwhile, wood products substitute for carbon-intensive concrete and steel. Win-win, right?

The reality is far messier. Only 1% of carbon originally in standing trees remains in products still in use at 100 years post-harvest, according to Harmon’s 2019 analysis. Another 13% sits in landfills. The rest? Released back to the atmosphere through processing waste, short-lived products, and decomposition. When you account for processing and transport emissions, the carbon footprint approaches the amount stored in long-lived products.

Yet the deeper problem is how IFM protocols account for HWP. They use a “storage factor” approach that immediately discounts harvested carbon to its 100-year average value. This creates massive over-crediting in early project years—the very period when those credits are used to offset current emissions. Projects receive credits upfront based on the difference between their current carbon stocks and a 100-year average baseline, while leakage deductions get spread evenly across a century. As Haya’s team calculated, this timing correction alone would reduce credits by 35%, and combined with proper leakage accounting, would cut crediting by 51-82%.

The protocols also fail to account for critical carbon pools. Soil carbon—which represents 40% of total forest carbon—isn’t included in most crediting calculations, despite substantially increasing in forest conservation projects. Meanwhile, when projects do cause soil disturbance through harvesting or site preparation, losses of 8-11% in the top meter of soil can over-credit projects by 10% or more when ignored.

Perhaps most troubling is the “biomass carbon neutrality myth” that underpins HWP accounting. A 2020 paper in Scientific Reports demolished this assumption, showing that wood product carbon footprints are systematically underestimated by pretending forest carbon dynamics following harvest and limited product lifetimes don’t matter. The time to “sequestration parity”—when the regrowing forest finally compensates for the harvest—often exceeds 30-80 years, well beyond climate policy horizons. During that entire period, the offset credit claims to have eliminated emissions that are still warming the atmosphere.

When Forests Burn: The Permanence Crisis That Everyone Knew Was Coming

California designed what was supposed to be the world’s most rigorous forest offset program, complete with a “buffer pool” insurance mechanism. Projects contribute 10-20% of their credits to this pool, which compensates for reversals when things go wrong. The state set aside 6.8 million credits to cover wildfire risks over 100 years. That insurance pool was supposed to protect the climate integrity of 231 million total credits issued across 149 projects spanning 5.5 million acres.

It lasted less than a decade.

By 2022, wildfires had already consumed 5.7-6.8 million credits—95-114% of the entire wildfire buffer. By 2024, total wildfire losses reached 10.7-13.6 million credits, representing 39% of the entire buffer pool across all risk categories. The math is brutal: California bet it could insure a century of climate risk with the equivalent of 10 years of actual climate damage.

The individual project failures tell the human story behind these numbers. In July 2021, Oregon’s Bootleg Fire tore through Green Diamond’s Klamath Basin carbon offset project, releasing 3.3 million metric tons of CO₂—equivalent to the annual emissions from 700,000 cars—in a matter of weeks. Only 20% of the 570,000-acre project burned, but that was enough to terminate the entire project. The Colville Reservation in Washington lost 3.7 million credits over multiple fires that burned through half of its 500,000-acre project, representing the largest single hit to the buffer pool. The reservation had sold those credits to BP, which used them to claim carbon neutrality for some of its operations.

The 2024 fire season brought five more projects in California to ash, including 45,000 acres destroyed in the Park Fire alone. This happened despite the total burned area being only 125% above average—meaning a relatively “normal” bad fire year in the new climate reality can wipe out multiple projects simultaneously.

What makes this crisis truly existential for forest carbon markets is that the climate change these offsets supposedly address is making forests increasingly vulnerable. Extratropical wildfire CO₂ emissions have increased 60% globally since 2001. In boreal forests, which hold 30-40% of land-based carbon, emissions have tripled. The UN projects a 30% rise in extreme wildfires by 2050 and 50% by 2100. You cannot insure against an accelerating threat with a fixed buffer pool sized for historical risk.

Disease presents an even more sobering scenario. Twenty California offset projects contain 14.2 million tons of credited carbon stored in tanoak trees. The entire disease and insect buffer pool for all projects over 100 years? 5.7 million credits. Sudden oak death, caused by Phytophthora ramorum, has already killed over 40 million trees in California and Oregon, and the pathogen is considered too widespread to eradicate. Under conservative mortality estimates for tanoaks in these projects, the disease alone could consume 82% of the disease buffer. Under moderate scenarios, it exceeds the buffer by 10%. The fungus hasn’t even reached most of the projects yet—three are within 1 km of confirmed detections, and all 20 are within 50 km.

CarbonPlan’s actuarial analysis concluded the buffer pool is “severely undercapitalized.” That’s putting it mildly. The program set aside enough insurance to cover maybe 10-15 years of a climate emergency that will last centuries. The fundamental asymmetry is inescapable: these projects store carbon for decades while the fossil fuel emissions they offset persist in the atmosphere for millennia. Temporary storage cannot compensate for permanent emissions, no matter how sophisticated the accounting.

Leakage: The Shell Game of Shifted Emissions

Here’s a question California’s forest offset program struggles to answer: if a landowner reduces harvesting in one forest to earn carbon credits, what happens to the timber market? Do mills just shut down? Do construction companies stop using wood? Or does harvesting simply shift to someone else’s forest—possibly halfway around the world—to meet the same demand?

The answer is obvious to any economist. It’s called leakage, and it fundamentally undermines the climate logic of forest carbon offsets. Yet the protocols treat it like a minor accounting adjustment rather than the deal-breaker it actually is.

California’s protocols assume a 20% leakage rate—meaning they deduct one-fifth of credited harvest reductions to account for increased harvesting elsewhere. The problem? Academic literature consistently finds rates of 60-100% depending on context. Wear and Murray’s 2004 analysis of Pacific Northwest harvest restrictions found 84% leakage—43% regional, 15% elsewhere in the U.S., and 26% in Canada. Gan and McCarl calculated 75-78% leakage for U.S. forest offset programs. When Norway reduced harvests, Kallio and Solberg found 60-100% of the “protected” carbon was simply logged across the border in Sweden instead.

Barbara Haya’s 2019 analysis of California’s program exposed an even more devastating problem: a timing mismatch that renders the accounting physically impossible. Projects receive massive credits in Year 1 based on the difference between current carbon stocks and a 100-year baseline average. But they deduct leakage evenly over 100 years. This means projects claim immediate carbon storage at baseline levels while simultaneously claiming they’ll maintain constant harvesting rates for a century. Both cannot be true.

The real-world effect is stark. Analyzing 36 projects representing 80% of California’s forest credits (97 million tons CO₂), Haya found that 82% of credits likely don’t represent real achieved reductions. They’re promissory notes: polluters emit today, and forests must sequester additional carbon for decades to compensate. But there’s zero evidence these projects are actually changing management to deliver those future reductions.

One project in Alaska generated 15 million credits in the first year alone—equal to 60% of California’s expected annual cap-and-trade emissions reductions from 2021-2030. Fifteen million tons of immediate “greenhouse gas debt” piled onto a program supposedly designed to reduce emissions. The landowner can continue the same timber harvesting practices they’ve always used, claim the harvest would have been even higher without offset payments, and collect millions while functionally doing nothing different.

The leakage problem gets worse the more you examine it. Most protocols “control” for leakage by accepting forest management plans prepared 2+ years before enrollment as proof nothing changed. That’s not verification—that’s an invitation to game the system. Projects can shift harvesting to non-enrolled parcels owned by the same landowner with minimal oversight. They can sell timber that was always going to market at the planned rate while claiming they would have clearcut without offset financing. The only limit is how aggressive a baseline the protocol will accept—and as CarbonPlan documented, 90% of projects push that limit to the maximum.

Even the concept of “controlling” leakage is borderline absurd. Timber is a global commodity. If U.S. forests reduce supply, prices rise, and harvesting increases in Canada, Brazil, Indonesia, Russia—anywhere mills can access trees. You cannot eliminate market-mediated leakage through project-level accounting tricks. It requires jurisdictional or global coordination that voluntary offset markets will never achieve.

The Fraud Isn’t Just in the Accounting—Sometimes It’s Just Fraud

In January 2023, The Guardian published the results of a nine-month investigation into Verra, the world’s leading carbon credit certifier. The findings were catastrophic: more than 90% of Verra’s rainforest carbon offsets were “phantom credits” with no genuine climate benefit. Of 29 avoided deforestation projects analyzed in detail, only 8 showed meaningful deforestation reductions. The threat to forests had been systematically overstated by roughly 400%.

Companies including Gucci, Shell, Disney, Salesforce, and easyJet had purchased these worthless credits to make “carbon neutral” claims. Verra’s CEO resigned four months later. The organization phased out its criticized REDD+ program. The voluntary carbon market, already reeling from academic critiques, lost what little credibility remained.

But the Verra scandal wasn’t an isolated incident—it was the industry’s public relations Chernobyl moment that revealed the systemic rot underneath.

The South Pole Kariba project in Zimbabwe stands as perhaps the most brazen example of offset fraud. This 785,000-hectare project generated approximately $100 million in revenue by claiming to protect forests from deforestation. Verra’s own review in September 2024 found that 15.2 million of the 27 million credits issued (57%) were fake—credited “in excess” of any actual protection. The threat had been inflated 5-30 times depending on the analysis.

What makes Kariba particularly egregious is how the money flowed. South Pole, the Swiss climate consultancy that developed the project, kept 42% of revenue—not the 25% disclosed. They purchased credits from project owner Carbon Green Investments for as little as €0.50 per credit, then sold them to corporations for €20 or more. That’s a 40x markup on phantom climate benefits. When auditors tried to track where CGI’s $11.4 million in revenue went, they could only account for 10%. The rest vanished into a web of related-party transactions and shell companies. Local communities, promised development assistance, received almost nothing.

Corporate buyers read like a who’s who of greenwashing: Volkswagen, Nestlé, McKinsey, L’Oréal, Delta Air Lines, Porsche, Shell. All purchased millions of credits to offset their emissions while doing business as usual. When the fraud was exposed, most quietly retired the credits and moved on. Almost none faced legal consequences. The market simply absorbed the loss and kept functioning.

California’s supposedly rigorous compliance market fared little better. ProPublica and MIT Technology Review’s 2021 investigation, backed by CarbonPlan’s analysis, found that nearly one in three credits issued had no real climate benefit—ghost credits worth $1.8 billion. The Mescalero Apache Tribe’s New Mexico project earned over $50 million in credits based on a regional baseline erroneously set at zero because of missing Forest Service data. Nearly all of their credits were fake. Updated rules issued two weeks later would have eliminated the credits entirely, but the tribe got to keep their windfall.

The fraud extended to projects that simply never existed. In August 2024, Verra permanently rejected 37 Chinese rice cultivation projects for over-issuing 4.56 million credits, including at least 19 projects that “appear to never have existed.” Shell had used over 1 million of these phantom credits toward climate targets. China-based auditors had rubber-stamped certifications for projects without verifying they were real. Four validation and verification bodies were suspended—the first time in Verra’s history.

Papua New Guinea became synonymous with “carbon cowboys”—opportunistic developers making quick cash from Indigenous land. The Australian Broadcasting Corporation’s investigation found logging continuing in forests supposedly protected, landowners receiving as little as $80 USD per person while promised schools and health centers never materialized, and the same forests being sold to both carbon projects and logging operations simultaneously. After dozens of allegations, PNG imposed a national moratorium in March 2022 and didn’t lift it until April 2025 with new regulations.

A Northern Kenya grassland project won awards at COP27 and was called “exemplary” by Kenya’s president. Yet Survival International’s investigation found over 100 compliance concerns raised by validators before Verra approved it. The project couldn’t accurately account for carbon removal, altered Indigenous herding practices, had no way to prevent herders from taking livestock outside boundaries, and maintained shoddy records where grazing maps didn’t match satellite data. Verra suspended it from issuing new credits in March 2023, but not before companies including Netflix and Meta had already purchased credits.

The pattern repeats endlessly: Cambodia projects where 88% forest cover dropped to 46% in nine years, one site going from 90% forested to zero while Verra certification remained valid. Brazilian Amazon projects connected to illegal timber laundering while selling credits to Boeing, Nestlé, Toshiba, and Spotify. Acre, Brazil’s model REDD+ program, where government officials admitted to ProPublica that “the priority is getting foreign aid to protect forests; the validity of the offsets is an afterthought.”

What’s remarkable isn’t that fraud occurred—it’s how systemic and rewarded it was. A comprehensive 2024 study analyzing 2,300 carbon mitigation projects found less than 16% represented actual emissions reductions. For cookstoves: 0.4% legitimate. Renewable energy: 0% legitimate. Forestry projects, the “best” category, still had only 25% legitimate—meaning 75% over-credited or fake. This isn’t a market with a fraud problem. This is fraud pretending to be a market.

Carbon Indulgences: Paying for Permission to Sin

In the 16th century, the Catholic Church offered a simple transaction: purchase an indulgence, reduce your time in purgatory. The more you paid, the more sins you could commit without spiritual consequence. The abuse of this system became so brazen—with clergy literally pricing different sins and offering bulk discounts—that it triggered the Protestant Reformation.

Today’s carbon offset markets offer an eerily similar bargain: purchase credits, continue polluting. The more you pay, the more emissions you can claim to have “neutralized.” And just like medieval indulgences, offsets don’t actually require you to stop the behavior—they just let you pay someone else to absolve you of responsibility.

Environmental writer George Monbiot crystallized this critique when he compared carbon offsetting to “the sale of medieval Catholic indulgences,” arguing that from sellers of offsets, “you can now buy complacency, political apathy and self-satisfaction.” The comparison isn’t just rhetorical—it cuts to the moral core of what offsets represent.

Yale Environment 360 addressed the analogy directly: “The standard trope is that [carbon offsets are] like the old practice of buying indulgences from the Catholic Church: You get to commit environmental sin—driving your SUV or living in your McMansion—and still sit at the right hand of God.” The problem isn’t that the comparison is unfair. It’s that it’s accurate.

Consider the logic. A company burns fossil carbon that was stable underground for millions of years, releasing CO₂ that will remain in the atmosphere for centuries to millennia. They then pay someone to plant trees or not cut down a forest, claiming this “offsets” their emissions. The forest might store that carbon for decades—if it doesn’t burn, if disease doesn’t kill it, if a future owner doesn’t log it, if climate change doesn’t push the ecosystem past a tipping point. Even in the best case, the forest storage is temporary while the fossil emissions are effectively permanent. The offset doesn’t cancel the emission; it just provides a narrative that allows the emission to continue (while allowing a certain class of people to feel good about themselves).

Michael Kramer, an investment advisor, framed the moral problem precisely: “You can’t make up for the use of carbon by buying your way to a position of freedom from responsibility for emitting it in the first place—once the damage is done, it’s done.” Like indulgences, offsets operate on the principle that sin can be balanced by paying someone else to be virtuous on your behalf, rather than requiring you to change your own behavior.

The “license to pollute” dynamic is most visible in how companies actually use offsets. Research from the University of Florida analyzed hundreds of large publicly listed firms and found that 70% of retired offsets were priced below $4 per ton. This isn’t serious climate action—it’s the cheapest possible way to buy a “carbon neutral” marketing claim. Companies literally shop for the lowest-cost offsets regardless of quality, because the goal isn’t reducing emissions but purchasing social license.

A 2025 study in Nature Communications analyzing 89 multinational companies found no significant difference in climate performance between companies that offset emissions and those that don’t. The top offsetters spent only 1% of capital expenditures on carbon credits on average. For airlines Delta and easyJet, there was evidence that “investments in decarbonisation are crowded out” by offset purchases—the offsets actively prevented real climate action.

Carbon Market Watch examined 24 major companies’ net-zero pledges and found they committed to reducing actual emissions by only 36% by mid-century while claiming net zero. To contribute their fair share to global climate goals, they would need to slash emissions by at least 90%. The 54-percentage-point gap? Filled with offsets that allow them to continue emitting while claiming climate leadership.

This is precisely the moral hazard that made indulgences so pernicious. When wealthy individuals could purchase spiritual forgiveness, it removed the incentive to actually behave ethically. When profitable companies can purchase climate forgiveness, it removes the incentive to actually decarbonize. Stanford researchers explain the fundamental asymmetry: “Compare two hypothetical companies that could call themselves carbon neutral. One dramatically reduces its own emissions, except for a small amount of remaining emissions that it offsets with carefully vetted carbon credits; the other takes no action to reduce its emissions and, instead, buys a large amount of low-quality carbon credits as offsets. The first company will have a much larger effect on tackling climate change.”

Under current offset rules, both can call themselves “carbon neutral.” The second company—doing nothing but buying cheap credits—gets the same marketing benefit as the first company that fundamentally transformed its operations. This creates a race to the bottom where serious climate action is economically disadvantaged compared to greenwashing.

The Transnational Institute put it bluntly: “Instead of committing to deep decarbonisation by setting credible pathways to reduce their own emissions, many companies are choosing to ‘neutralise’ them, through the purchase of carbon credits.” The word “neutralise” does heavy lifting here, implying equivalence where none exists. You cannot neutralize emissions by paying someone to promise to maybe store some carbon temporarily in a tree that might not burn.

This matters because offsets actively delay real climate action. Every dollar spent on cheap offsets is a dollar not spent on actual decarbonization. Every year a company can claim “carbon neutrality” through offsets is a year they avoid the difficult operational changes required to eliminate emissions. Every government that accepts offsets in its climate accounting is a government allowing continued pollution under the fiction that it’s being compensated somewhere else.

Climate scientist Joe Romm published a 50-page paper titled “Carbon offsets are unscalable, unjust, and unfixable—and a threat to the Paris Agreement.” His conclusion, after analyzing decades of offset failures: offsets are “just another in a never-ending chain of things designed to let fossil fuel companies keep adding unsustainable levels of carbon dioxide, methane, and other climate-killing emissions.” Like indulgences before them, offsets exist to preserve a profitable status quo, not to drive the fundamental change required.

Why the Market Can Never Be Fixed

After 25 years of carbon offset markets, the evidence of failure is overwhelming. The Clean Development Mechanism, the first major program, had 85% of projects with low likelihood of actually reducing emissions. California designed a “second generation” program explicitly to fix those problems—and still ended up with 29% systematic over-crediting. Verra has updated methodologies repeatedly, yet 90% of its rainforest credits turned out to be worthless. The Integrity Council for Voluntary Carbon Markets launched “Core Carbon Principles” in 2024 to finally ensure quality—and Corporate Accountability found 93% of the largest approved projects still had “fundamental failings.”

The pattern is unmistakable: every reform attempt fails, triggering new reforms that also fail, in an endless cycle that researchers increasingly describe as “intractable.” A landmark 2025 review in Annual Review of Environment and Resources examined this 25-year history and concluded bluntly: “We have assessed 25 years of evidence and almost everything up until this point has failed.”

The reason is simple: the problems aren’t implementation failures, they’re structural features of how offset markets work.

Start with the baseline problem. Every project must establish what would have happened without the project—the counterfactual scenario against which credits are calculated. This is inherently unknowable. You cannot prove what would have happened in an alternate timeline. The best you can do is make educated guesses using historical data, regional averages, and economic models. But those guesses must predict timber prices, economic conditions, land use changes, trade policies, technology shifts, and owner intentions over 100 years.

It’s not just hard—it’s impossible to do accurately. And because landowners know their actual intentions while verifiers don’t, every flexible baseline rule gets exploited. Project developers have complete information; auditors have partial information. This information asymmetry is fundamental and cannot be overcome through better auditing.

Even worse, everyone in the system benefits from over-crediting. Developers profit when they have more credits to sell. Auditors, paid by developers, have incentives to be lenient so they’ll be hired again. Registries like Verra generate revenue by issuing more credits. Corporate buyers benefit from cheap, plentiful credits that let them claim “carbon neutrality” at minimal cost. As Stanford researchers concluded: “At their roots, offsets involve a basic asymmetry of information between offset sellers and buyers that, especially when combined with enormous complexities and uncertainties, will make them vulnerable to being gamed.”

This creates what Cambridge researchers call a “lemons market”—where buyers cannot distinguish quality, so sellers flood the market with bad products, trust breaks down, and the market fails. The voluntary carbon market is already experiencing this. Prices have been in race-to-bottom freefall because companies seek the cheapest possible offsets to meet PR goals, not actual climate impact.

The permanence problem is similarly unfixable. Forests are biological systems subject to fire, disease, drought, pests, and human decisions. California’s buffer pool—supposedly conservative insurance against these risks—was depleted in less than a decade. Actuarial analysis shows it’s “severely undercapitalized” even before accounting for accelerating climate impacts. You cannot reliably insure century-scale biological storage when the climate itself is destabilizing. And you certainly cannot equate temporary biological storage with permanent fossil carbon releases, regardless of how much insurance you set aside.

Leakage is inherently unmeasurable at project scale. Timber is a global commodity. If harvest reductions in California increase timber prices, that incentivizes cutting in Canada, Brazil, Indonesia—anywhere. The market finds equilibrium. Project-level accounting cannot capture these global dynamics. Academic studies consistently find leakage rates of 60-100%, but protocols use 20% because acknowledging the real numbers would make most projects uneconomic. The economic incentive ensures the fiction persists.

Even when projects claim to follow rigorous protocols, verification failures are endemic. Auditors are paid by project developers, creating obvious conflicts of interest. Site visits happen years apart. Satellite monitoring can miss selective logging and degradation. Verra’s certification of the Cambodia project that went from 90% forest cover to zero while maintaining valid credits isn’t an anomaly—it’s what happens when verification depends on occasional document reviews rather than continuous monitoring.

Charles Canham of the Cary Institute captured the structural problem perfectly: “The flaws in the markets are structural and deep, and may be irredeemable.” Professor Raphael Calel at Georgetown stated simply: “I’m not seeing any examples of carbon offsetting programs that are working well.” Barbara Haya, who has studied offsets for over 20 years, concluded: “We need alternative ways of supporting climate mitigation because the current offset market is deeply not working.”

Perhaps most damningly, climate change itself undermines the system from within. The emissions these offsets supposedly compensate are making forests more vulnerable to fire, disease, and drought. We’re using forests to offset the very emissions that are destroying forests’ capacity to store carbon. It’s a positive feedback loop guaranteed to fail.

The voluntary carbon market has already lost 60% of its value between 2022-2024 as scandals mounted. Major companies including Nestlé, Gucci, and EasyJet have reduced or eliminated offset purchases. The Article 6 mechanism finalized at COP29, meant to finally fix offset quality problems, “did not substantially address the quality problem” according to the Oxford/Penn review. Even COP negotiators couldn’t agree on how to make offsets work.

The Only Solution is to Stop Pretending

Forest conservation is critical for climate, biodiversity, and human wellbeing. That’s not in dispute. What’s in dispute is whether paying landowners to generate carbon credits that claim to “neutralize” corporate emissions is an effective way to fund that conservation.

The evidence says no. Resoundingly, consistently, across every major analysis: no.

Leading researchers are now calling for phasing out forest carbon offsets entirely. The Oxford/Penn team recommends “urgently phasing out offsets that did not actively suck CO₂ from the atmosphere and shifting the focus of offset markets to high-quality carbon dioxide removal and storage.” Berkeley’s Carbon Trading Project is developing an alternative “contributions” approach that funds climate mitigation where it’s most needed without claiming to offset emissions. Nature published an editorial stating the idea “that emissions can be offset through projects that claim to avoid releases or to remove carbon dioxide from the atmosphere is fatally flawed.”

The core insight is simple: protecting forests is valuable, but calling it an “offset” creates the dangerous fiction that fossil fuel emissions can continue without consequences. Like the Protestant Reformation’s rejection of indulgences, we need to reject the logic that allows the wealthy to pay for permission to harm.

Companies should directly reduce their own emissions. Governments should directly fund forest conservation. Financial flows should support climate mitigation based on where they can do the most good, not based on creative accounting that generates marketable credits. Forest protection should be valued for what it is—critical ecosystem preservation—not twisted into a commodity that enables continued pollution.

The voluntary carbon market’s slow collapse isn’t a failure to be fixed. It’s a reckoning with a fundamentally flawed idea. You cannot shop your way out of a climate crisis. You cannot offset your way to sustainability. You cannot buy indulgences and call it virtue.

The house of cards is falling. What matters now is how quickly we stop building climate policy on top of it.