Gotcha. Since you’re already in the UChicago network I think it’s definitely worth getting in touch with someone about Climate Vault, as in the early days they were aiming at permanent retirement—I don’t know the specifics of why they pivoted away from that model but I imagine you’ll learn something useful.
Giving Green briefly looked into some brokers with similar models here as well; ultimately we’re pretty skeptical. Even if your marginal additionality is 1 for small purchases, it’s inherently not a scalable model. If the core of what you’re doing is funding public goods and not selling carbon allowances, you should pitch that more! There are more cost-effective ways to just reduce emissions :)
That report is super valuable, thank you for sharing it. I thought there were no better climate actions hehe. Please let me know what’s on your mind.
The point of my post was to clarify what retiring allowances accomplishes under what circumstances, and it is my fiduciary duty to address the criticisms ;) I’ll keep Ultra Civic on the sidelines, as I’d prefer to discuss the mechanics of allowance retirements. I’ll bring up inflationary taxation via crypto issuance as an alternative to finance public goods in another post.
I’m arguing that skepticism is not warranted in certain programs (like RGGI).
The idea that additionality requires polluters not to have banked allowances is wrong. Granted: if every year they used all issued allowances, then retiring an allowance causes a one-ton reduction in emissions this year. If they were to use all issued allowances over a longer time period, the retirement would cause a one-ton reduction in emissions over that longer period. We can never know if polluters will use all issued allowances, but we can know if polluters plan to use all available allowances, and the answer is positive if the current market price of allowances is bounded away from zero.
Retiring allowances in a small cap-and-trade program like NZ is not super scalable—at best you’d correct the negative externality of emissions in tiny NZ. But the programs I listed cover 2GtCO2/year. I’d say there’s potential for scale?
(RGGI-specific) It’s good that RGGI imports power, not bad—imported power has a strong Canadian hydro component and is 40% cleaner than in-RGGI power. It would be bad if RGGI’s power imports were mostly coal-fueled.
Thanks again for bringing all this up (and sorry for taking so long to respond—I’ve been travelling a lot)
Hmm. I think once we get into the territory of “plan to use”, though, you end up with the same types of criticisms that apply to carbon credits and RECs, no?I really think that a nonzero price is not a good enough signal here! Polluters don’t have perfect information about future climate regulations, future technologies, future market demand for their products, or even necessarily the future of the cap per Ville’s point below. I read Making Climate Policy Work recently, which is a really thorough critique of the compliance markets; could be useful to see which of their criticisms apply to retiring allowances.
Giving Green hat on: if you want relatively certain and near-term emissions reductions at this price point, we usually recommend Tradewater. However, generally we would argue that unconstrained donors (i.e. not meeting some kind of net-zero reporting requirement) should donate to nonprofits working on policy or other systems change; GG’s cost-effectiveness analyses land about an order of magnitude lower in $/ton than most allowances or credits, and I think also Founders Pledge’s climate fund’s analyses do as well.
Hi ethai! Thanks again. I had the chance to read Making Climate Policy Work and found it super insightful.
Regarding perfect foresight
You’re right: I glossed over the extent of imperfect foresight when discussing uncertainties. For RGGI specifically, here are reasons why polluters likely have accurate expectations about their future allowance usage:
Actual RGGI caps have gone according to plan, with the exceptions of three adjustments explicitly made to reduce excess banked allowances (see my response to Ville’s comment below, adjustments 1-2, and adjustment 3).
RGGI covers power generation, arguably the sector with the clearest and most predictable pathways toward decarbonization (Cullenward and Victor, 2021, Fig. 1.2).
Recent forecasts indicate higher power demand in 2026--2030 than previously anticipated. If anything, this suggests that power plants may have underestimated their allowance needs, making full usage more likely.
There are also valid reasons for uncertainty:
RGGI is inherently political and dependent on voter preferences, which can change unpredictably over a 5-year horizon.
Low-carbon tech exists (solar, wind, nuclear), but breakthroughs in fusion or storage could unexpectedly render fossil fuels obsolete.
Whatever the foresight of polluters is right now, it should improve as third-parties retire allowances. Right now, a power plant’s cost of purchasing a CO2 allowance it won’t use is $20. The higher the price of allowances, the greater the cost of making mistaken purchases, so you’d expect their foresight to improve as more allowances are retired and prices increase.
Regarding carbon credits and Tradewater
I think it’s useful to distinguish between carbon credits and the climate actions they incentivize. Any climate action has to answer: “How much does it reduce atmospheric CO2?” But carbon credits must also answer: “How much future climate action does purchasing this credit incentivize?” That’s another hard question, and the impact of carbon credits is often unclear.
For example, Tradewater burns refrigerants and issues carbon credits, which they later sell. Even if burning refrigerants were perfectly additional, it’s unclear how much more refrigerant they will burn because you paid them $21 for a carbon credit. Similarly, someone might retire CO2 allowances and issue carbon credits to sell later. Even if retiring allowances were perfectly additional, you can’t know how many more allowances that person will retire because you bought a carbon credit from them.
Regarding policy, systems change, and Making Climate Policy Work
Cullenward and Victor emphasize the critical role of political feasibility. Carbon pricing programs (CO2 taxes + cap-and-trade) are politically costly and consequently haven’t been very useful, whereas targeted regulations (e.g. car emissions standards) have quietly done the most to decarbonize society.
They do acknowledge that a global cap-and-trade program with allowance prices matching the social cost of carbon would efficiently eliminate harmful carbon emissions. It’s just that it’s infeasible.
There is a role for cap-and-trade programs, only smaller than previously thought. To be effective in driving decarbonization, they need higher allowance prices. Their policy recommendation is to stop overallocating allowances.
You can view retiring allowances as implementing this policy recommendation, only without the politics.
Thanks again for engaging. I will really appreciate hearing your further thoughts.
Gotcha. Since you’re already in the UChicago network I think it’s definitely worth getting in touch with someone about Climate Vault, as in the early days they were aiming at permanent retirement—I don’t know the specifics of why they pivoted away from that model but I imagine you’ll learn something useful.
Giving Green briefly looked into some brokers with similar models here as well; ultimately we’re pretty skeptical. Even if your marginal additionality is 1 for small purchases, it’s inherently not a scalable model. If the core of what you’re doing is funding public goods and not selling carbon allowances, you should pitch that more! There are more cost-effective ways to just reduce emissions :)
That report is super valuable, thank you for sharing it. I thought there were no better climate actions hehe. Please let me know what’s on your mind.
The point of my post was to clarify what retiring allowances accomplishes under what circumstances, and it is my fiduciary duty to address the criticisms ;) I’ll keep Ultra Civic on the sidelines, as I’d prefer to discuss the mechanics of allowance retirements. I’ll bring up inflationary taxation via crypto issuance as an alternative to finance public goods in another post.
I’m arguing that skepticism is not warranted in certain programs (like RGGI).
The idea that additionality requires polluters not to have banked allowances is wrong. Granted: if every year they used all issued allowances, then retiring an allowance causes a one-ton reduction in emissions this year. If they were to use all issued allowances over a longer time period, the retirement would cause a one-ton reduction in emissions over that longer period. We can never know if polluters will use all issued allowances, but we can know if polluters plan to use all available allowances, and the answer is positive if the current market price of allowances is bounded away from zero.
Retiring allowances in a small cap-and-trade program like NZ is not super scalable—at best you’d correct the negative externality of emissions in tiny NZ. But the programs I listed cover 2GtCO2/year. I’d say there’s potential for scale?
(RGGI-specific) It’s good that RGGI imports power, not bad—imported power has a strong Canadian hydro component and is 40% cleaner than in-RGGI power. It would be bad if RGGI’s power imports were mostly coal-fueled.
Thanks again for bringing all this up (and sorry for taking so long to respond—I’ve been travelling a lot)
Hmm. I think once we get into the territory of “plan to use”, though, you end up with the same types of criticisms that apply to carbon credits and RECs, no?I really think that a nonzero price is not a good enough signal here! Polluters don’t have perfect information about future climate regulations, future technologies, future market demand for their products, or even necessarily the future of the cap per Ville’s point below. I read Making Climate Policy Work recently, which is a really thorough critique of the compliance markets; could be useful to see which of their criticisms apply to retiring allowances.
Giving Green hat on: if you want relatively certain and near-term emissions reductions at this price point, we usually recommend Tradewater. However, generally we would argue that unconstrained donors (i.e. not meeting some kind of net-zero reporting requirement) should donate to nonprofits working on policy or other systems change; GG’s cost-effectiveness analyses land about an order of magnitude lower in $/ton than most allowances or credits, and I think also Founders Pledge’s climate fund’s analyses do as well.
Hi ethai! Thanks again. I had the chance to read Making Climate Policy Work and found it super insightful.
Regarding perfect foresight
You’re right: I glossed over the extent of imperfect foresight when discussing uncertainties. For RGGI specifically, here are reasons why polluters likely have accurate expectations about their future allowance usage:
Actual RGGI caps have gone according to plan, with the exceptions of three adjustments explicitly made to reduce excess banked allowances (see my response to Ville’s comment below, adjustments 1-2, and adjustment 3).
RGGI covers power generation, arguably the sector with the clearest and most predictable pathways toward decarbonization (Cullenward and Victor, 2021, Fig. 1.2).
Recent forecasts indicate higher power demand in 2026--2030 than previously anticipated. If anything, this suggests that power plants may have underestimated their allowance needs, making full usage more likely.
There are also valid reasons for uncertainty:
RGGI is inherently political and dependent on voter preferences, which can change unpredictably over a 5-year horizon.
Low-carbon tech exists (solar, wind, nuclear), but breakthroughs in fusion or storage could unexpectedly render fossil fuels obsolete.
Whatever the foresight of polluters is right now, it should improve as third-parties retire allowances. Right now, a power plant’s cost of purchasing a CO2 allowance it won’t use is $20. The higher the price of allowances, the greater the cost of making mistaken purchases, so you’d expect their foresight to improve as more allowances are retired and prices increase.
Regarding carbon credits and Tradewater
I think it’s useful to distinguish between carbon credits and the climate actions they incentivize. Any climate action has to answer: “How much does it reduce atmospheric CO2?” But carbon credits must also answer: “How much future climate action does purchasing this credit incentivize?” That’s another hard question, and the impact of carbon credits is often unclear.
For example, Tradewater burns refrigerants and issues carbon credits, which they later sell. Even if burning refrigerants were perfectly additional, it’s unclear how much more refrigerant they will burn because you paid them $21 for a carbon credit. Similarly, someone might retire CO2 allowances and issue carbon credits to sell later. Even if retiring allowances were perfectly additional, you can’t know how many more allowances that person will retire because you bought a carbon credit from them.
Regarding policy, systems change, and Making Climate Policy Work
Cullenward and Victor emphasize the critical role of political feasibility. Carbon pricing programs (CO2 taxes + cap-and-trade) are politically costly and consequently haven’t been very useful, whereas targeted regulations (e.g. car emissions standards) have quietly done the most to decarbonize society.
They do acknowledge that a global cap-and-trade program with allowance prices matching the social cost of carbon would efficiently eliminate harmful carbon emissions. It’s just that it’s infeasible.
There is a role for cap-and-trade programs, only smaller than previously thought. To be effective in driving decarbonization, they need higher allowance prices. Their policy recommendation is to stop overallocating allowances.
You can view retiring allowances as implementing this policy recommendation, only without the politics.
Thanks again for engaging. I will really appreciate hearing your further thoughts.
Paco