The Commons

Note: article is in process of being edited and restructured

Months into my first post-college job at a climate mitigation non-profit, I emailed an update to family and friends that included a half-joking reference to ‘carbon trading.’ A godmother figure warned me off it. “A good way to legislate some profiteering for a few without real benefit for mother earth,” she replied.

That was 2011. Carbon offsets’ image problem in the intervening years has worsened. From 2019 – 2023, as corporate use of voluntary carbon offsets surged, waves of high-profile news articles trumpeted carbon credits’ shortcomings. Many credits appeared to be hot air, such as the tree-planting efforts in the Lower Mississippi Valley where landowners received carbon credits for trees they had already planted (Bloomberg) or an anti-deforestation project in Zimbabwe that issued tens of millions of tonnes of bogus credits (New Yorker). The resultant exposes can be diagrammed like a Dan Brown conspiracy novel: (1) in situ portrait of a ‘carbon cowboy’ in an exotic location (2) ominous unveiling of carbon project shortcomings with (3) zoom out to reveal the structural failings of climate policy, especially that offsets might be cover for rapacious business as usual. They fed a paranoia that carbon credits might be not just ineffective but damaging. 

Yet despite doubts from journalists, academics, and swaths of the public, voluntary carbon markets ballooned tenfold over a decade to an estimated $2b-dollar-per year market, before falling somewhat in the past few years. There’s a holographic quality to their size: they are a rounding error compared to most markets, far short of the wild expectations of a few years ago, yet still enough to potentially transform the non-profit ecosystems around niche industries like clean cookstoves. And carbon markets might be on the verge of another upswing even though, still, no one quite agrees what a carbon credit even is.

**

“Priming the pump” for big funding flows

Starting in earnest around 2018, reputationally sensitive companies, anticipating a sea change in regulatory and social stances to greenhouse gas emissions, made ‘net zero’ climate commitments to eliminate their greenhouse gas emissions over time. But they had few good options for dealing with left-over emissions once they made the easy emissions cuts within their control. This produced demand for tradable carbon credits, which supposedly represent one metric ton of CO₂ reduced or removed from the atmosphere from mitigation activity anywhere in the world.

This was an artificial market of invisible gases. Demand existed within a broader enabling architecture that involved stamps of approval from environmental guidance-makers, advisors, and policy influencers. One underappreciated but critical catalyst were the climate professionals who decided that carbon markets could be an important source of funding for protecting and restoring nature, transitioning industrial systems, or removing carbon from the atmosphere using novel technologies. There was a sense that this corporate voluntary market was building the infrastructure that would enable regulatory markets hundreds of times larger, and that a failure to establish favorable rules during this nascent phase could mean that a preferred approach might get shortchanged. Many of the most thoughtful market influencers focused less on the risk that carbon markets might fail to deliver ironclad climate gains than that they should fund deserving causes, provided they could improve over time.

Consequently, the chattering class that helped to stoke demand, prime supply, and create the architecture and rules within which the two met was turbocharged into a temporary feeding frenzy. The ratio of ink spilled on carbon markets versus tangible outputs rivals cryptocurrency. (Carbon markets and blockchain even had an erstwhile conjunction.) Because carbon markets’ value is defined by their ambiguous legitimacy, they are defined, structured, and accounted for by voluminous documents that also support a robust archeology into how they work. Such an archeology is complemented by firsthand experience: I was, from 2020 – 2023, excited and nauseated to be in the thick of the carbon market gold rush.

**

Jungle diligence

Somewhere deep down a jeep road in the Southern Yucatan, past the overgrown ruins of a Mayan household, was the sample plot where the ejidarios showed us how they inventoried regrowing trees. Here, despite the high canopy, the broadleaf forest felt more sparse and open than jungle-like. We drove past a set of beehives that produced high-value forest honey, past old rubber trees, then walked on foot the rest of the way to the unremarkable plot.

During and after the Mexican revolution communities were granted communal ownership over land in a structure known as ejidos. Ejidos control most forestland in Mexico. In some, forest management is light touch; in others, timber management is comparatively sophisticated. An increasing number of ejidos have been generating carbon removal credits as their forests fill out following historical harvesting. Similar projects had been noted for being community-based and beneficial (though not in uncomplicated ways).

In one ejido we visited, we smiled for group photos at an overlook where an ocean of forest stretched to the horizon in multiple directions. Driving around the patchwork of farmland, buildings, and roads, though, it was easy to imagine how development could encroach over time, and how payments for carbon-rich forests might subtly shift the decision-making. Not a lot, but a bit.

In this ejido, carbon credits were loosely associated with a new forest management plan that, counterintuitively, involved extracting more wood than before. The harvest was modest, and was complemented by a program to replant high-value hardwoods after each area was cleared. Our guides pointed out how leaving some mature ‘mother’ trees helped foster better regrowth. Still, it was hard to get my head around how these seedlings and improved growth in the surrounding forest were sequestering enough carbon to quickly counterbalance the loss associated with the harvest. I was thinking about this dynamic as we rode a massive, rumbling tractor through the muddy forest, perched precariously above the treads. ‘What do you do differently now that you have the carbon project?’ I asked when we stopped in a clearing to see some logs.

‘Nothing, it is the same as before,’ the ejidario answered.

It was as I had feared: even if the forest was in fact sequestering more carbon, the causality  was too murky to satisfy quality-focused buyers who demanded a tight link between the project and the climate benefits. A bust.

**

Roots of carbon markets

The standard story told by carbon markets observers is that market-based environmental solutions rooted in the Reagan-Thatcher soil of the 80s and grew as the U.S. acid rain program in 1990, which enabled trading of sulfur dioxide allowances among U.S. power plants. Such a solution was efficient in theory – the market would find and reward the lowest cost options –, reasonably effective in practice (due in part to some lucky coincidences), and more politically palatable than command-and-control regulation. Similar market environmentalism logic was incorporated into the Kyoto Protocol, the first international climate agreement, in 1996, and bloomed as the first wave of carbon market projects in the 2000s under the international Clean Development Mechanism, where buyers in rich countries paid for (mostly phantom) emissions reductions by HFC-producing factories in China and wind projects in India. It seemed like a similar architecture might emerge domestically, with voluntary efforts helping lay the groundwork ahead of a push for a nationwide carbon price during the early Obama years. Though federal efforts stalled, state efforts progressed: California climate policy began to be implemented in earnest in 2010 – 2012, including a moderate role for carbon credits within the state’s cap-and-trade program. The idea was that large emitters could trade a declining number of carbon allowances, similar to the acid rain program, but could also meet some of their obligations by buying carbon credits produced from sectors outside the emissions cap, like agriculture and forestry. This is also the time I became a climate professional, working to reduce greenhouse gas emissions from agriculture and forests in Sonoma County, California.

Many patterns evinced by carbon markets today were in evidence during that initial wave. First is the weirdly tight link between voluntary and compliance markets. The architecture of voluntary carbon markets is often meant to frontrun compliance markets. Indeed, one of the leading voluntary carbon credit registries – the non-profit organizations making the rules for how credits get generated –, the Climate Action Reserve, was created by the state of California to help prepare for carbon credit’s potential incorporation into regulatory architecture. Companies buying voluntary carbon credits have perhaps done so out of a sense that they are getting out ahead of something that might soon be required, or that voluntary compliance might forestall regulation. 

Second are carbon credits shortcomings, which some observers say are structurally inevitable rather than merely kinks that can be ironed out with tight rules. Alongside stories of clear climate failures, such as Indian and Chinese factories boosting their HFC-23 production simply to get money for destroying it, sit murky philosophical questions about what a carbon credit should do, i.e., what constitutes an acceptable bar for carbon credit robustness. Is it enough to subsidize climate beneficial activities or should each credit be ironclad? Do the climate benefits need to be permanent? In the 2000s, the Chicago Climate Exchange was flooded with cheap soil carbon credits from midwestern farmers for plowing less, which many farmers were already doing. Prices were propped up by speculators before falling to pennies when federal climate legislation was pronounced dead. Clearly that wasn’t the way forward. Now California regulators were debating whether abandoned mine or rice methane could create good credits; whether to allow avoided deforestation credits from Brazil; and other technocratic questions whose answers hinge on risk appetite as much as science. (Fast forward: credits in California’s compliance markets have run into some of the same credit quality problems as voluntary ones.)

Third is their semi-opacity. These markets exist in and by a series of rule-making processes that are ostensibly open but whose complexity rewards deep expertise and complicates meaningful public participation. This creates the necessary conditions for a technocratic class who help to structure and translate the rules of the system to others. Enter the carbon market professional. Who is the carbon market climate professional? They might circulate from a government agency to a non-profit and then into a philanthropic funder. They structure the rules of the system as part of a carbon registry and then go to work for a company generating credits, or work as a consultant and then as a sustainability professional in a big company. Because of this, the smart professional leaves doors open. The ones drawn into the bubbling frontier of carbon markets during the second-wave boom 2019 – 2023 seemed to have some common Gatsby-esque attunement to the ‘promises of life’ even though they had arrived by diverse paths.

**

Microsoft: carbon removal’s patron saint

Until recently Microsoft was one of the largest single buyers in the voluntary carbon market. They announced ambitious public targets in a splashy launch in 2020. Insiders describe a long process of building C-level support for a ‘moonshot’ approach to not just eliminate net emissions but also to remove enough carbon to account for historical emissions, funded by an internal carbon price of $15 charged to all business units. But to do it they would have to buy not just credits, but specifically good credits, and specifically carbon removal.

Much has been made of carbon removal as a category distinct from standard carbon offset credits, which are mostly reductions. Modeling suggested that, to meet Paris Agreement climate goals, the globe has to reach ‘net zero’ – the point at which net emissions into the atmosphere fall below zero – by 2050, and that this will require removing a significant volume of carbon from the atmosphere via a variety of approaches. This inspired a rash of corporations to commit to their own net zero equivalents under the Science Based Targets Initiative. Ambitious companies like Microsoft set target years as early as 2030.

Initially, the full contours of such a long-term commitment were fuzzy, including how fast companies would have to reduce their own emissions, how narrowly the boundaries for emissions would be drawn, and how offsets could or should work in service to such commitments. But companies would have quickly concluded that (1) it would be impossible to drive their own emissions to zero and (2) for the remaining emissions, they would need to secure good credits rather than bad credits or their commitments would be seen as hollow. From the perspective of the atmosphere, carbon removed from the air is hardly different from carbon not released in the first place. But a removals-oriented approach both flowed logically from the mid-century orientation – removals have to be massively scaled to meet global net zero goals– and helped to differentiate it from the problematic carbon credits of the past through unsupported claims that it would be easier to establish rigorous counterfactuals in a situation where carbon dioxide was being sucked from the atmosphere than when a forest was being threatened to be cut down. 

Microsoft set up a serious operation for procurement of credits, hiring a team of staff who initially focused on buying existing credits but over time worked on more than 30 separate commercial deals for long-term streams of credits that would help them meet their 2030 net zero goals, roughly 6 million tonnes worth per year. They also fostered an ecosystem of experts to help them distinguish good credits from bad, including a nascent firm called Carbon Direct with a handful of fulltime staff but a growing stable of consulting academic experts who were eager to shape carbon markets, get rich, or both. I was one of them.

It is hard to overstate the gravity of Microsoft in the carbon removal space. They are estimated to represent roughly 80% of the voluntary funding committed to carbon removal thus far, if you count commitments for future credits. They have committed to deals involving an estimated $5-15 billion in carbon credits over the coming decades They have provided public playbooks for their approach, guidelines for how they assess quality, and funding for an array of for profit and non-profit partners. Other credit buyers carefully watch what they do, even though the scale and resourcing of their approach is seen as hard to replicate.

Thus, the entire market seemed to shudder when it was reported Microsoft had ‘paused’ their carbon removal buying a year early as they redirect resources toward the same AI efforts that have become an uncapped emissions liability. Microsoft downplayed it, but has admitted to facing headwinds. “The moon has gotten further away,” wrote Microsoft Chief Sustainability Officer Melania Nakagawa. Months later, Carbon Direct laid off many of their carbon market staff.

**

Are carbon credits insidious?

One of the idealistic experts drawn into the Carbon Direct orbit quickly bounced out due to concerns about who the organization might advise. He was worried that major oil and gas producers were trying to normalize carbon credits to decrease public and regulatory pressure on their fossil fuel extraction business. Many others were concerned too: one philanthropic leader told a meeting that an oil and gas major was spending $40m just on advertising their tree planting efforts. Oil industry figures were sometimes explicit about their hopes for carbon removal. “If [oil] is produced in the way I’m talked about, there’s no reason not to produce oil and gas forever,” the CEO of Occidental Petroleum, Vicki Hollub, told NPR in 2023. Environmentalists eager to make sure that carbon removal and carbon credits are not seen as a valid alternative to reducing emissions have sued oil and gas companies and airlines for making carbon neutrality claims based on carbon credits, with mixed success.

Critics on the left hate carbon credits, which are seen as corporate greenwashing that perpetuates business as usual and delivers little value to people on the ground while perpetuating environmental injustice. Proponents counter with claims that companies investing in carbon credits are more serious about decarbonization, without disentangling causality. Both sides agree that decarbonization without carbon credits is more expensive, either because companies can no longer purchase illusory reductions or because they can’t use cost-efficient reductions outside their direct control. How all this impacts the political economy of climate ambition is difficult to say. Is climate policy less ambitious without carbon credits? Mostly, carbon credits have been used as a reluctant fudge factor in policy spheres. For example, in California, carbon credits were sometimes talked about as a pressure release valve in case cap-and-trade prices rose too high.

Some picture a world in which governments’ policies support climate mitigation without need for the indirect and unsettled accounting of carbon credits. For example, governments could collect revenues from companies and invest those in policies that support climate-friendly activities. So elegant. But it is also easy to imagine those funds getting rerouted to budget holes; to picture the U.S. supporting illusory reductions from Kansas CAFOs rather than paying to protect the Brazilian Amazon; or to envisage companies torpedoing the policy altogether.

**

How carbon markets really work

One remarkable thing about a young, burgeoning industry is that almost no one knows what they are doing. Dynamics are changing fast enough that legacy operators struggle to keep pace, whilst demand for experienced personnel far outstrips supply. Mark Carney, before he was elected prime minister of Canada, led a premature push to turn carbon credits into a mature financial asset; ex-techies and consultants created buzzy companies to plant trees using cocoons and drones, only to be stymied by fires, flooding, and rampaging wild boars; philanthropic quarterbacks tried to nudge the carbon markets toward protecting tropical forests. Because the first wave of carbon credits was viewed with suspicion, it created a business opportunity for new rule-makers, project developers, and intermediaries to usurp incumbents by claiming a higher quality bar – but, almost by definition, these new players lacked experience. (These same players benefit from continued presence of bad credits because a uniformly high-quality market would destroy their edge.)

There is no single rule maker in voluntary carbon markets. A handful of non-profit carbon credit registries issue methodologies that define the rules for how carbon credits can be generated. There are hundreds of such methodologies covering everything from industrial gases to hydropower, each of which has their own intricacies and loopholes, all purporting to produce credits representing one tonne of climate pollution eliminated from the atmosphere. These methodologies are written by consultative public processes involving the registries, the companies who hope to develop projects, and outside experts. So far, so good. But engaging with these rulemaking processes is pedantic and tiresome, and many parties benefit from laxer rules that produce more abundant, cheaper credits: registries that get paid on a per credit basis; developers that can produce and sell more credits, even quality-agnostic carbon credit buyers that get lower prices.

Cross-cutting efforts have sprung up to try to combat the inevitable erosion of quality, most notably the Integrity Council for Voluntary Carbon Markets, as well as carbon credit ratings agencies that help buyers sift good credits from bad. Quality-focused buyers, tired of bad credits, have helped power these efforts to both eliminate the most egregious credits and allow buyers to better know the strength of any given credit. They have helped. Though carbon markets are still dogged by their questionable reputation, a number of indicators suggest that carbon credit quality is improving.

There remains, however, a fundamental problem with efforts to sift good from bad: quality is subjective. Pretty much everyone agrees that some credits are better than others. “A tonne is not a tonne is not a tonne,” carbon removal expert Julio Friedmann says. But in about a third of cases raters disagree substantially. The most discussed difference between credits is how long their climate benefits last – purists say that the climate benefits should be more-or-less permanent over thousands of years – but the thorniest is ‘additionality,’ the idea that the climate-beneficial activity couldn’t have occurred without the carbon credit and associated funding. Oftentimes additionality involves scrying unknowable counterfactuals. Someone says they planted a tree to generate and sell a carbon credit. But maybe they were already planning on planting the tree. Or maybe they would have left the land undisturbed, and trees would have regrown naturally after a time. We can say confidently that the promise of future carbon finance somewhat shifted the economic calculus to make planting trees more attractive. Purchasing the ‘verified tonne of carbon removed’ might increase the chances the tree planter will reinvest and plant further trees for additional profit, but not on a perfect 1-to-1 basis. What gets treated as a binary to be proven (“this project is clearly additional”) might be better thought of as shifting the supply curve.

Statistics offers sophisticated but imperfect tools to deduce causality and resolve counterfactuals. One common approach is to show that the carbon project performed better than equivalent control areas with similar characteristics. A notable improvement after the project starts helps show it is the project, not other factors, causing the performance bump. This approach has become increasingly common in relevant methodologies. 

Still, projects often had a holographic quality: good in one light, bad in another. One of my favorite projects received middling scores from a ratings agency who feared that East African farmers might be making too much money from the trees they planted to qualify as “additional.” Those with the greatest command of technical material are in a privileged position to define quality but, when disagreements emerge, the final judgement represents less an emergence of scientific truth than a political-scientific negotiation. The way that we choose to measure a criteria becomes the very definition of that thing. For example, some registries say that an activity is de facto additional if it has less than 20% market penetration; this definition bears only a loose relation to the idea that a project is additional if carbon finance was the determining factor in causing the activity to occur. Rubrics can hide the myriad fundamental judgments one layer back, whilst the ground truthing that might help calibrate that judgement instead focuses myopically on whether projects follow the stated rules.

My own experience reviewing dozens of forestry projects and editing dozens more reviews across project types is that, despite a notable spectrum in quality, many credits fall into a middle ground where they are supporting productive work even if they aren’t definitively reducing or removing emissions. They are better than they have a right to be given they are subject to the same brutal capitalist realities as everything else. This is because there are people who care about climate embedded throughout the system; because credits are subject to considerable scrutiny; and because an increasing number of projects are going forward as a direct result of offtakes from buyers, making additionality more clean-cut than when buyers were focused on pre-existing credits. The butter-churning of the climate professionals has done some good.

**

Market volatility, market opportunity

CFOs have been curtailing sustainability spending the past couple years. Voluntary ambition has ebbed as companies continue to walk back sustainability commitments in the Trump era. Within the carbon credit subspace, the quality wars have taken a toll, with companies leery of spending millions of dollars on carbon credits that might just land them in a bad news article. Corporate board members read the same articles as the rest of us: one asked me if forests were a real climate solution (yes, with caveats); another pressed on the fundamental squishiness of additionality. SBTI’s revised rules seem positioned to make carbon credits an optional bonus for high achiever corporations until at least 2035. Heady predictions for explosive growth in the voluntary carbon market have been tempered as the overbuilt industry shrinks and consolidates. Carbon removal in particular is at an emotional ebb following Microsoft’s pullback.

But there are also signs of life. Country-to-country carbon trading under the Paris Agreement allows, for example, Switzerland to purchase credits from Peru to help meet Swiss climate goals. The rules around this have only been firmed up in the past few years after multi-year negotiations at the COP and now transaction volume is beginning to ramp. Rules are relatively permissive about what types of carbon credits are allowed; there’s arguably even less oversight in bilateral transactions than in the voluntary carbon market. Another source of emergent credit demand is international aviation. There, rules are tighter and credits are in short supply, though that doesn’t mean that credit quality is higher.

Microsoft, belying the reported ‘purchasing pause,’ recently bought credits from a Danish bioenergy and carbon capture project. This came shortly before Europe announced that carbon credits would eventually become part of the European Union’s massive compliance Emissions Trading System (ETS), which covers a large swath of the continent’s emissions. Although carbon credits had continued to have a modest role in California, Japan, and elsewhere, the EU decision could eventually represent a step change in regulation-driven demand – the culmination of the expectations and dreams of many carbon credit architects, developers, and even buyers. Credits will only be allowed from direct air capture and bioenergy with carbon capture and storage as 2040 approaches and industries face an increasingly challenging time dealing with residual emissions. (Forest and farming credits are excluded, in line with the common perception that they are less robust, though they are supported by incentives under other programs.) If past experience is any guide, it is highly likely that some of the methodologies and muscles built up in the voluntary carbon market will get adapted into the regulatory system. For market boosters, hope reemerges. The hologram shifts again.