Carbon offsets aren't working, and probably can't

A conversation with Cullenward & Victor, part two.

All right, Voltaires, today we have the second part of my conversation with Danny Cullenward and David Victor, authors of the new book Making Climate Policy Work. (See part one for their credentials and background.)

The book, you will recall, is an extended argument that carbon pricing schemes, despite their attractiveness in economic theory, run afoul of fundamental forces in political economy.

In part one we talked about why people love carbon pricing so much and why, despite promises of expansion, most carbon pricing system still only cover a narrow band of economic sectors.

Today we tackle two more of the book’s big topics.

The first is carbon offsets. A carbon offset is basically a payment to a third party to take some carbon-reducing action — plant trees, prevent deforestation, capture and bury carbon, what have you. The idea is that entities for whom reducing direct emissions is expensive or impossible can still pay for reductions.

In the context of carbon pricing systems, offsets are designed to play two roles. First, they are supposed to hold prices down, since they are often cheaper than direct reductions. And second, they are supposed to allow carbon pricing systems to reduce emissions even in sectors that are not covered by the system and are otherwise difficult to reach.

Like carbon pricing itself, carbon offsets are an appealing idea. Pay less, for more emission reductions, and help save forests while you do it! More and more cities, states, and companies are making net-zero pledges without really knowing how to get there, so lots of them are looking around for cheap reductions, and carbon offsets are sitting right there.

But the model only works in practice if the carbon reductions represented by carbon offsets are real — verifiable and verified. Cullenward and Victor argue that they rarely are, and furthermore, that the incentives created by the programs themselves work against reliable verification. Large-scale offsets programs inevitably spark a race to the bottom.

Second, we talk about linkages. Just as carbon pricing programs were supposed to expand to cover more economic sectors, they were supposed to expand to bring in more jurisdictions, more states and countries. That hasn’t happened either, and the explanation turns out to be more or less the same.

Let’s jump in!

Carbon offsets haven’t worked, don’t work, and won’t work

David Roberts

Let’s talk about carbon offsets, a perpetually hot topic. Offsets do a lot of work in carbon pricing systems, easing the political difficulties that the pricing systems themselves create. That would be fine if they were reliable and verifiable. But you guys say they're not, and what’s more, never will be. Explain.

Danny Cullenward

There's two parts. First is the difficulty in saying what a good offset is. The second is the politics — what offsets are for and how the money gets used.

The fundamental problem with offsets is that we can't observe what doesn't happen. The offset project is making a claim: if you didn't pay us, we wouldn't do this thing. It's a counterfactual claim that's factually unobservable. You can only infer, by evidence, reasoning, and detailed sector-specific knowledge, what might or might not happen. That's a challenging problem in the best of circumstances.

Everyone imagines there's this vast army of people working on this problem, ensuring that there's a good outcome, but in practice, you see understaffed regulators — very few people paying close attention to this who have no financial interests. And you see special-interest nonprofit organizations figuring out if they line up behind these offset programs, they can direct the revenues to particular causes.

An offset is basically a way of channeling a compliance obligation on a polluter to some third-party activity. So if you can direct the flow of those funds, that's powerful. Just a quick data point: in California, there's over 200 million offset credits, easily worth 10 bucks apiece. That's $2 billion in private financial flows that have moved through a regulatory program that has fewer than a dozen full-time staff members.

That problem is massive — you lose that fight every time, because of the fundamental political economy of offsets. Almost everywhere they've been used in compliance programs, they’ve been used for cost containment.

The story is, I can deliver a climate pollution reduction somewhere else at lower cost. If the quality of that offset credit is identical to a reduction inside the program, you have no change in emissions, you just have lower costs.

David Roberts

Again, a very appealing theory.

Danny Cullenward

It's very elegant.

It's important to point out some of the environmental justice [EJ] concerns. If you don't reduce your pollution here, the people who experience the local air pollution associated with, say, a refinery have sincere and legitimate concerns. But assuming we solved the local air-quality problem and the EJ problem, it's an appealing theory, for sure.

The political problem is that these programs want offsets to keep costs down. To be honest, they don't particularly care about quality. At the end of the day, their bottom line is about increasing the volume at the lowest possible cost.

So the fundamental political economy is structured around massive volumes of low-priced offsets — the most problematic ones. It becomes a race to the bottom, where people compete on price and not quality.

And nobody really cares about quality in these public markets. That's why in every single major compliance offsets program, you see this downward spiral in quality. They are not about achieving comparable environmental aims. They are about reducing compliance costs.


David Roberts

Is there not a single offset program you could point to that has fought off these trends and maintained high quality?

Danny Cullenward

I'll be honest, I spend a lot of time working on this and it's hard for me to point to a lot that's working well.

The counter-examples are, you're starting to see some companies in the private sector say that they don't care what the price is, they’re only concerned with quality. Those private sector buyers are going to help create markets for quality offsets.

There are some small-scale things you can point to, but when you look to large volumes of offset credits at low prices, I don't think there's a single program that's delivered that successfully.

David Victor

I agree. The place to watch is voluntary markets amongst folks who understand the quality problem, understand the logic of why offsets, if they can be done with integrity, are important.

One of the reasons offsets persist, keep coming back like bell-bottoms, is that you now see a bunch of companies and governments under enormous pressure to make net-zero pledges.

It’s coming from the right place. But when they go do the numbers, they have a heart attack. They don't know what to do, especially if you go all the way up to scope-three emissions. [Scope-three emissions — indirect emissions embedded in a company’s or jurisdiction’s supply chains — are explained here.]

So they all say the right things in public: we want to make all the reductions ourselves to the extent possible, yada yada. Meanwhile, the analysts say, you've made this public pledge, how are you going to do this? The only backstop answer is offsets.

So they almost have to imagine that these problems of quality don't exist, or can be solved — “Mark Carney is working on it, so it'll be fine.” Once that magic wand is waved and offsets are available, net zero is possible. The sad irony here is that a lot of companies and governments are trying to do the right thing, but net-zero pledges are propagating the belief that offsets are working fine.

David Roberts

There’s an argument to be made that if these companies and governments didn’t have access to offsets, and just had to reduce their direct and indirect emissions on their own, it would appear so daunting and terrifying that you'd get less support. Perhaps having a release valve in offsets is enabling these corporations and jurisdictions to feel confident enough to step up at all. Do you worry about cutting that off?

Danny Cullenward

That's the political trap of starting with a flawed paradigm.

I'm trying to make a slightly different case to them for what to do. Take a company that has made a bold pledge and set aside a budget for this. I don't want them to reduce the funds they’re putting into climate, I want them to grow that over time, and I think a lot of companies get that and want to do that. But is that money going to good things?

When you're competing on price and nobody's looking over quality, difficult things that are real can't compete. So we need to break the current relationship between price and effort in these systems. People need to stop pretending that we can cheaply offset our way to net-zero goals.

To a company that has been blindly purchasing generic offsets and is becoming aware of these problems — don't stop. Take the money and make the most of it. That means not fetishizing tonnage claims and the implementation of net zero.

That's a difficult thing to do, because people have made all these bold pledges. But we have to figure out a way to search for quality. We don't see that working in offsets.

David Roberts

What about the fact that there are big emitting sectors that are problematic and don't fit well under a carbon pricing system? Are there ways to get at them other than offsets?

Danny Cullenward

When people say offsets can help us get to sectors we can't otherwise get to, you need to remember two things.

First, what does it take to do a good offsets program? It means you know a great deal about what's going on in that sector, what's possible, what's not possible, who's doing it — you need all of this detailed information. To set your additionality standards and baselines and all the technical details about non-observable phenomena, you have to have rich and intricate knowledge. Well, that's also what you need to do to set up a regulatory program.

The second thing is, people talk about offsets as a way of getting a toehold in these tough sectors, getting them to play with carbon pricing, maybe get comfortable and expand the reach of carbon pricing programs over time.

In practice, offsets do the exact opposite. You cannot earn an offset credit if what you are doing is legally required. So offsets create an incentive for industry to lobby against new regulations. New regulations blow the whole thing up!

Every single beneficiary of an offsets program has a direct and specific incentive to lobby against any rule that would mandate what they are doing for voluntary incentives.

David Victor

The solution to those so-called “hard-to-abate” sectors [think airlines, shipping, heavy industry] is, you work sector by sector on a plan and a bunch of experiments that push the frontier for that sector. The cost is going to vary. It has to be bespoke to sectoral conditions.

When the UK government hosts COP 26, they're going to have a bunch of these challenge areas — decarbonizing the grid, some stuff on hydrogen. The logic is to get small clubs of countries and firms together and just do stuff in those sectors, as opposed to pretending you're solving the problem with a bunch of pledges and then offsets.

The great linkage that wasn’t

David Roberts

The other great hope for carbon pricing systems, one I confess I have always harbored, is that they will grow and link to one another — that there's something inevitable about their spread. But in practice, linkages just haven’t happened. Why doesn’t it work?

David Victor

Creating a carbon market is a little bit like inventing a new form of money. The quality of the money is only as good as the integrity of the underlying institutions. If you link a bunch of countries together, then the quality of your market is only as good as the oversight and integrity of the least upright member.

And there’s a political issue: if you put countries together, your politics become tethered to the sea anchor of the least ambitious country in your system.

Economists can tell you a lot of fancy ways to give extra allocations to one sector or jurisdiction or another, but then you make the political problem even worse, because if there are misallocations of credits in the beginning, then when the markets are linked together and equilibrate, you have these big capital flows that move across countries that are politically very toxic.

The economic logic here tells you that what you should do is link together the most unlike markets — that's the biggest gains from trade. But what we see in the real world is either no linkages at all, or linkages of the most like markets, because those are the markets where the political willingness is similar or easier to manage, and where the administrative capabilities are the most comparable and highest.

Danny Cullenward

If you look at linking in the real world, there's very little of it. There's basically Europe and the peripheral states [the EU plus Iceland, Liechtenstein, and Norway], which have all cooperated — that is extraordinary. Then you have the Western Climate Initiative [WCI], which is now just California and Quebec, but was once envisioned to be most of Canada, half of the western United States, and parts of Mexico. And you have the RGGI [Regional Greenhouse Gas Initiative] states on the East Coast.

One reason you see linking between almost identical sectors is that most of the linking is happening at the subnational level [in the WCI and RGGI], and subnational governments can't write treaties. Policymakers that are running these [subnational] systems know that they have to make them easy to enter and exit, because they cannot legally bind other governments to stay within the system.

When a party comes or goes [from the system], if that party is dissimilar to the rest of the market, entry and exit can be very disruptive. Whereas if all of the players are roughly similar in their ambitions, coming and going is an administrative detail.

That is exactly what we saw when Ontario pulled out of the WCI program. I don’t celebrate it. But because Ontario was so similar to the rest of the program, it didn't fundamentally destabilize the market.

Linking like markets doesn't really change a whole lot, and if linking like markets is the only thing that can be done, it doesn't really help us proliferate [climate policy] to places that have different levels of ambition, or different qualities of institutions.

David Roberts

One possible counter-argument is that if you create these mutually beneficial systems, low-performing jurisdictions will want to gain entry and thus will have an incentive to improve their institutions. Has that proven true at all in carbon markets?

Danny Cullenward

You do see some of it. RGGI is a great example. It’s a fairly low-ambition but very transparent and clear program. It makes it easy for states that are on the periphery of climate action to join.

You see [linkage] when there are transparent programs, relatively modest in their ambition, that have turnkey operations. The broader the coverage of the program, the more you're asking that new entrant or lower-performing entity to take on, in terms of fighting all industries at once. When you do it sector by sector, and you do it carefully and transparently, you offer an on-ramp to people who want to do a little climate action now.

That's really important, but let’s not oversell the contribution.

Stay tuned tomorrow for an explanation of “Potemkin carbon markets” like the one in California — markets that aren’t nearly as effective on the inside as they look on the outside.