We’re thinking about carbon credits all wrong

This story originally featured on Hothouse. Subscribe to the climate newsletter on Substack.

Netflix is offsetting your movies with mangroves. Along the Caribbean coast of Colombia, the tropical trees rise above hot, humid mud flats, home to manatees, otters, and needle crocodiles. They also house 20,000 tons of carbon bought and paid for by streaming dollars.

Netflix, which operates in Colombia’s capital of Bogotá, protected this particular forest after committing to net-zero emissions last year. Since then, Netflix has procured 1.5 million tons of carbon credits from Kenyan grasslands to Brazilian rainforests.

Netflix’s approach is the strategy du jour for American companies: Cut some internal emissions and offset the rest by buying up carbon credits that protect natural ecosystems. Of the 50 largest US companies, more than a third have voluntarily purchased carbon credits, a figure only expected to rise, reports Capital Monitor, a sustainable finance research firm. 

With companies doubling down on offsets, it’s worth asking: Are we thinking about carbon credits all wrong?

The trade in CO2 has been denominated in tons ever since the Kyoto Protocol, a 1997 climate treaty establishing the international carbon market. Today, companies agreeing to net-zero targets do the same—offsetting one ton of emissions for every one they emit.

But the devil is in the details. This market paradigm has led to endless debates about how to define, issue, and sell a quality carbon credit. Tracking a ton of CO2 from its creation to its (alleged) final resting place, it turns out, is infernally difficult. It demands complex monitoring, math, and near-metaphysical feats of prediction.  

A better question may be: “What is the most effective way to spend our climate money right now?” That answer may be decarbonizing the future, not offsetting the past. 

Today, almost every dollar spent on offsets pays for a project to zero-out past carbon emissions. For now, few other ways exist to reach “net zero” goals before mid-century.

A second (or complementary) option is to decarbonize the future as rapidly as possible. Instead, it would pool money to place massive bets on decarbonizing the economy, and measure success by the carbon intensity of the resulting technologies. That alleviates the need to buy offsets at all. This backs away from the entire idea of a “carbon market” denominated in tons.

If it sounds implausible, we already have a successful model: the vaccine for the leading cause of childhood death, pneumonia. 

Ghosts of carbon past 

In 2020, Microsoft decided to eliminate all its emissions by embracing clean energy and energy efficiency, and impose an internal carbon tax. The rest would be offset. Then it announced something no other major company had done before: it would zero out every ton of emissions since its 1975 founding. In practical terms, that means Microsoft must now remove 600 million tons of carbon dioxide from the atmosphere between 2030 and 2050.

Microsoft has two choices when it comes to carbon credits: removals and avoided emissions. Avoided emissions prevent the release of emissions into the atmosphere in the future. For example, an Amazonian forest the size of Vermont burns every year, protecting swaths of this forest, as Netflix preserves its mangroves, prevents that carbon from entering the atmosphere. Those avoided emissions—essentially, lower deforestation—are turned into credits and sold. A similar model works in the renewable energy market by displacing fossil fuels.  

Removals, on the other hand, pull carbon out of the atmosphere and stash it somewhere. Trees and marshlands turn carbon into cellulose, wood, and soil. Carbon dioxide captured in the air is liquified and pumped underground into permanent storage.

Microsoft chose removals. “The reason is simple,” the company said in its first offset report. “Looking ahead 10 years shows we simply can’t meet our global climate goals without carbon removal.” That is the conclusion from a US National Academy of Sciences report suggesting as much as 10 gigatons per year, roughly double the U.S.’s annual emissions, will be needed to meet climate targets by 2050.

As a result, carbon removals feature prominently in corporate climate strategies. But so far, only about 10,000 tons of carbon have been permanently sequestered from the atmosphere by man-made solutions. The most proven (and affordable) of carbon removal technology remains trees. More than 99% of Microsoft’s portfolio relies on nature-based solutions, predominantly forestry and soil.

Only Microsoft ran out of offsets. After screening 189 carbon removal projects for “high-quality, long-lasting, and immediate carbon [credits],” it ultimately approved just 3% of the 55 million offsets it reviewed, ultimately purchasing more than half of those for 2021. That covered barely a tenth of the company’s annual emissions.

For now, that suggests, there aren’t enough high-quality carbon removal credits to support a market. “If everybody thinks like Microsoft does, there’s no possibility that they’ll get enough removals generated by the carbon market to meet that,” says Owen Hewlett, the chief technical officer of Gold Standard, an independent carbon credit verification agency. “It does show the absurdity of focusing on one thing to the exclusion of all else, because the competition for it isn’t going to add up.” 

Microsoft believes the market will catch up. Its $1 billion climate fund is backing research into direct air capture technology, noting “the world needs to invent substantially stronger technology-based solutions than are available today.” 

That’s why Silicon Valley entrepreneurs are shoveling corn husks in Kansas.  

Is Kansas the future of emission reductions?

On the edge of a Kansas field, workers toss bales of corn stalks and tassels into an anaerobic furnace. Rather than incinerate the waste, the lack of oxygen and high temperatures reduce it into two distinct substances: a gooey black oil (bio-oil) and a form of charcoal called biochar. Biochar returns to the fields to fertilize the next crop, while storing carbon in the soil. The bio-oil, also rich in carbon, is pumped into old salt caverns left behind by gas and oil extraction. 

All this is the work of Charm Industrial. The San Francisco-based carbon removal startup pays farmers for each bale of their agricultural refuse, and charges climate customers such as Microsoft and Stripe $600 per ton for permanently sucking CO2 out of the atmosphere. Charm pays for CO2 capture in plants, and pumps it back into the ground. It’s a bit like running an oil company in reverse. Charm has sold about 416 metric tons of carbon removal—equivalent to two rail cars worth of coal—for about $250,000 total.

Charm Industrial is one of a dizzying number of carbon removal companies testing techniques, from algae to spreading the carbon-capturing minerals on beaches. And almost all are very, very expensive—as much as $700 per ton of sequestered carbon, nearly 50 times the average price on the voluntary market. 

But the promise is that the price will come down dramatically over time with enough investment, following the cost of solar power has fallen from $1,825 per Watt in the 1950s to just $2.94 per Watt today, the cheapest source of energy in history.

In April, the online payments company Stripe launched an “advanced market commitment,” essentially a $925 million fund to buy credible carbon removal offset. This “guarantees a market for a product, even before it exists,” says the company. All projects must permanently sequester carbon for about $100 per ton by 2050, and eventually reach a global scale. Those that show progress compete for a jackpot: the guaranteed purchase of nearly a billion dollars in offsets (Shopify, Facebook’s parent company, Meta, and others have all funded an initiative called Frontier, a Stripe subsidiary).

The first companies get the highest price per ton, rewarding the earliest to market. Prices fall over time. This June, Frontier awarded $2.4 million to six projects as part of its first purchases.

This “advanced market commitment” was inspired by a way to fund affordable vaccines for critical diseases in poor countries. Few saw investment. That changed about a decade ago when the Gates Foundation, the US, UK, and other countries promised manufacturers a high price if they produced affordable vaccines for pneumonia, the leading cause of infectious death in children under 5. The faster a pharmaceutical company brought an effective vaccine to market, the higher the price.

Vaccine producers were soon competing to manufacture hundreds of millions of doses of a pneumococcal vaccine. The first pharmaceutical companies to bring vaccines to market were paid $10.50 for a set of shots. By 2019, the price had dropped to $6. Today, prices are still falling and an estimated 700,000 children have been saved.

The same model could work for the climate. 

What the future holds

There is, and likely always will be, a role for offsets, sold by the ton, to account for historical emissions. For example, the UN estimates we will need 10 GT of carbon removals by mid-century (a number reiterated by the oil giant Shell). That’s true assuming we follow our business-as-usual path of decarbonization. But these high numbers constrain how we think of the future—and the present.

Rather than focus on eliminating their past emissions, we can invest in ensuring future operations emit less—or zero—carbon. This demands a sea change in how we think about “net zero.” Toly Rinberg, an applied physics doctorate student at Harvard, argues such estimates are “weaponized by polluting interests” to preserve the status quo. “People calling for 10 or 20 gigatons of removal per year, why are they doing that?” he says. “It reduces political pressure to decarbonize.”

If we only budgeted carbon removals for the most challenging industries like aviation or concrete, the annual need would be as low as 1.5 to 3.1 gigatons per year, roughly a third of UN estimates (and half the estimate published in the US National Academy of Sciences).

That means redesigning our tool for the goal. Right now, we’re remodeling our house with whatever tool happens to be in our hand. The favored tool—offsets— abate the past, not decarbonize the future. “It’s completely backwards, right?” says Hewlett of the carbon credit monitoring agency Gold Standard. “If you want a kitchen, then you want to make sure you’ve got the right tools to build the kitchen.”

If we asked ourselves the most impactful thing we can do with limited resources, offsets that incentivize a decarbonized economy, instead of just carbon removals, might be it.

Among climate leaders and researchers, there is near universal agreement about the continued need to deploy green technologies and green the grid. Changing the infrastructure of our economy from cars to power grids to buildings will take a lot of capital, anywhere from three to six times current levels by 2030 to put us on the path to net-zero by 2050. 

The advanced market directive points the way. Instead of only paying to offset the past, companies can pool an equivalent amount of money to develop technology that makes their operations carbon-free—and future offsets unnecessary. There’s no shortage of technology targets: electric arc furnaces for green steel in cars, carbon-free aluminum smelting in phones (something Apple is funding through its 2017 Green Bonds program), and transforming agricultural practices (as much as 80% of emissions could be cut just be commercializing existing technologies).

Harnessing the carbon market to get us there, says Anne Finucane, former vice chair of Bank of America, would be transformative. “Reimagining carbon offsets means treating them as a catalyst; not just to grow trees, but to grow new technologies,” she wrote. “We need to build a new carbon market that broadens the definition of offsets to accelerate innovation in green technology at the same time it helps companies decarbonize faster.”

That means denominating a “carbon credit” in more than just tons.