Numlock Sunday: James Temple on the accounting bug behind millions of tons of carbon emissions
By Walt Hickey
Welcome to the Numlock Sunday edition.
This week, I spoke to James Temple of MIT Technology Review who wrote “The Climate Solution Actually Adding Millions of Tons of CO2 Into the Atmosphere” with ProPublica’s Lisa Song. Here's what I wrote about it:
California’s Air Resources Board has rolled out a statewide carbon credit policy, but the issue is that it oversimplifies how much a given tract of forest actually mitigates carbon emissions. As a result, a new analysis found that of the 130 million credits generated as of last fall — worth $1.8 billion at current prices — an estimated 20 million to 39 million carbon credits don’t actually achieve any benefits for the climate. The “ghost credits” didn’t preserve any additional carbon in forests, and empowered additional pollution commensurate with annual emissions of up to 8.5 million cars. The Air Resources Board disputes the study, which looked at the specific pricing lines within forests to identify areas where the value of an acre appeared to be inaccurately priced compared to adjacent areas, allowing landowners to game the system.
This story is fascinating because it takes a concept that gets discussed a lot in the abstract and drills down what the heck is exactly being measured, and in doing so discovers ways that the system can be improved and finds things it’s failing to capture.
James, you wrote a story with Lisa Song of ProPublica all about California's carbon credit program, and you found a really fascinating issue with it. But before we get into what you found, let's kick it off with what's a carbon credit and why is there a marketplace for these in some states?
Yeah, that's a really good question and a good place to start. A carbon credit is something that is developed within an offsets program. I think probably the offsets programs that people are most familiar with involve forests. Basically the idea is that a landowner can earn carbon credits through various offset systems by altering their land management practices in ways that either prevent carbon emissions — by, say, logging less than they otherwise would have, changing their timber harvesting cycle, just committing not to log at all — or by trimming trees or certain other practice that leaves that undergrowth and certain other practices that can promote additional growth over time, that would remove more carbon than otherwise, year by year.
The idea is that if you have these landowners doing this in certain parts of the world, companies can buy those credits either through a voluntary offset system (and say that we bought enough credits to offset X tons of emissions that are our corporation is doing) or within a regulated carbon offset market like the one that California operates on, which is by far the biggest in the U.S., where it operates within a cap and trade program. It offers the state's polluters that have to operate within this cap one way of finding credit that would allow them to emit a little more than they would otherwise be able to.
A carbon credit itself is, is what's the unit there, is it like a ton of CO2?
That's exactly right. It represents a ton of CO2.
Cool. Your analysis was of the 130 million credits that had been generated in the state of California. As of last fall, you wrote that at current prices that's worth $1.8 billion. This is a lot of money changing hands, but when you actually went into the weeds here and the details, you found out that anywhere from 20 to 39 million of these carbon credits don't actually achieve any benefit for the climate. What did you find?
I should say here that the starting point for all of our work was a study by a San Francisco non-profit known as Carbon Plan, which has a mission of evaluating the scientific integrity of carbon removal efforts. That can include forests, but it also includes direct air capture machines that suck CO2 out of the air and various other ways that this can be done. They were in the midst of doing this kind of system-wide analysis, looking at all the paperwork for all the projects in the system that they could dig up enough data onto to analyze in a systematic way. What they ultimately found was that the system is issuing, according to their conclusions, tens of millions of carbon credits that don't represent actual carbon and climate benefits.
Because the state polluters in a regulated system can buy those credits for the right to emit more than they otherwise would have been able to, that suggests that a system that should be operating in a carbon neutral way could actually be increasing the net emissions, at a point when states and nations all need to be making as rapid progress as possible in addressing climate change.
This does stack up to a lot, that would be up to the existence of 8.5 million extra cars on the road in the United States. You wrote about how granular you had to get with some of these maps, so I guess I'm wondering where's the accounting bug?
That's where the whole thing gets complicated pretty quickly, and that's where we faced the biggest challenge in writing the story. Just starting to get to the point where we understood the issue that they were raising. And to be honest, that took several calls before it really clicked, at least in my own mind, to the point where I could actually go to an editor and say, “this is the issue they're identifying, this is why it's a problem and this is why I think we should free up some resources to give me the time to look into this in a deep way," and later in coordination with the Lisa Song at ProPublica.
Here's the basic issue. That basic level, the program determines the average carbon levels for forest types across broad regions. Landowners who have forests that contain more than that can earn credits for the difference, so long as they commit to keep those forests at those levels, around those levels, at least on average, over the next 100 years. And Carbon Plan identified two related problems that basically stem from that. The first is that they contend that the average is too coarse to represent actual typical forestry practices in many areas, which is what that number is supposed to be a proxy for. So, many projects are earning more credits than they really should as far as atmospheric CO2 is concerned.
The second is that, at least in the aggregate, some project developers have come to recognize that you can maximize the credits you earn by specifically looking for projects that, for various reasons, stand out significantly above those averages. Carbon Plan was careful in a way they talked about that to say that they weren't pinpointing specific people or specific projects, but they were saying in the whole, it seemed to be the case that the system had created this adverse selection air issue, or what they have come to refer to as, at least in conversations with us, as a cherry picking practice.
There's this really illustrative map in the piece that basically there's a coastal region that has a certain average metric ton of CO2 per acre, and then an inland region. But that's a fantasy. That's not how forests works. There's no line in the sand saying this is the good forest, and this is the less good forest —- it's a gradation. Basically, that there's a region within here that you can make more money off the forest hypothetically than you otherwise should be able to, based on the actual carbon content of it. Is that right?
That's right. The way that we tried to get this really complicated issue across early in the story is by laying out the example that you're talking about: if you were to go to this part of Northern California, up near Oregon, where you have these really dense Redwood, Douglas fir forests, really close to the National and State Redwoods Parks, you have this line, that's just sort of arbitrarily — I mean, maybe not totally arbitrarily, but to the degree that the system is going to work the way it does, it has to set bounds somewhere — and these bounds run through forests that if you're directly on either side of it, they're basically the same types of forest filled with lots of really carbon rich Douglas fir trees.
But one side of it falls into an area where it's almost exclusively that, and the other side of it falls into an area where that's just sort of an anomalous zone. Inside an area where, as you get deeper and deeper inland California, and it gets drier and hillier, the overall average ends up being dragged down by Ponderosa Pines and other sorts of trees, that are more sparse, aren't as thick and tall as a coastal Douglas Fir and are more sparsely spaced out. So, the average in that area, it get gets dragged down, even for like a mixed conifer forest. If you were having basically the same type of land on one side of it, you can earn millions of dollars worth of credits, and just on the other side you would earn none.
So, to us, that was just the clearest way of underscoring that these boundaries end up mattering a lot. Then you can see by just where projects are clustering, that other people seem to recognize that this issue exists and presents opportunities, potentially, to earn a lot of credits, if you were to locate projects on those sorts of land.
Like you wrote, ecology is messy, but I suppose where, from a policy standpoint, does this take us next? I know that Carbon Plan has kind of said, listen, the paper is criticizing the design, not the actual landowners, many of whom are Native American tribes often, but I guess where do we go from here? What lessons are we taking forward from it?
Yeah, this is an incredibly important question that I don't have a great answer to as a reporter, without a PhD in ecology. I'm not a carbon market expert, I'm not a policy expert. I feel like all I can say is that this has been a subject that I've been covering for a couple of years now, offsets and carbon removal, this sort of broad space of negative emissions and this tension between knowing as a scientific and climate matter that we have to do a lot of negative emissions on the one hand, and knowing on the other hand from the reporting I've done and the numerous studies that have come out that doing it well is really hard. There are just some fundamental challenges, particularly with forest offsets.
And just to name three: It's close to impossible to know what landowners really would have done in the absence of this program. Whatever carbon we saved today through a program like this could still go up in smoke in a wildfire, a decade or two from now. And it's really hard to understand how global markets for timber respond to offsets.
Offset projects will reduce timber supplies, right? But if some landowner on the other side of an ocean doubles their timber output because they're seeing changing market prices that incentivize them to change their behavior as a result of these offsets program, then we actually haven't gotten anywhere from a climate perspective.
I guess the most I'll say is that I'm sure there are better ways to do this and better ways of addressing this to some degree. But I think you could also find people that think that these are things you just can't fix. These are just such deep problems that as soon as you take this obvious clear goal of, "we need to be able to plant more trees and preserve more trees and there are carbon benefits from that," taking that as a broad goal, and then putting it into the human political system that requires it to be done in a very precise ton for ton balanced out way, that there are just some really deep, deep challenges to doing that in an accurate and reliable permanent way.
The other thing I'd say is to the degree that we could do this better, and I'm sure that we can, doing so better requires careful evaluating how it's working so far and making good faith efforts to address the issues that arise.
I really enjoy your work here at the MIT Technology Review. You're the Senior Editor for Energy, so you cover a lot of fairly bleeding edge stuff. What kind of stuff do you typically cover on your beat in general on this, and where can folks find you?
I've been kind of specializing in negative emissions issues for the last couple of years, just because it's become a much more important and lively debate as companies and governments take climate change more seriously and start to think through, well, how realistically are we going to do all this? So, that's a big area. That's been a big area of my focus for the last two years.
We'll see, I'm going to have some hard thoughtful conversations with my editors now that I'm on the other side of this project about is this something we have exhausted. In general, I've always covered a lot of technical, clean energy issues and try to in the same way assess those sorts of things to some scrutiny in terms of what looks most promising, what sorts of challenges do various potential tools face in terms of actually getting them out into the market and in ways that they meaningfully address climate and emissions goals.
I do a lot of pretty nerdy, technical battery coverage as well. I've done a lot of carbon removal, also done some geoengineering stuff. But what I cover next, I guess we'll see.
I would just be remiss to not say that it was an honor to get to work with ProPublica on this. I learned a ton working with Lisa who was just fantastic, working with T. Christian Miller, the main editor on the project and just having just these fascinating conversations with their design and art folks who think really carefully and thoughtfully about how you find art that represents what the story you're trying to tell and helps clarify the concepts. I mean, I think my favorite part about this project ended up being the comic drawn by Taylor Dow.
I appreciate the support of my editors and colleagues who sort of filled in the gaps on me being largely gone from our site for a long time there.
If you have anything you’d like to see in this Sunday special, shoot me an email. Comment below! Thanks for reading, and thanks so much for supporting Numlock.