In a recent post, I argued that we have lost sight of what carbon accounting is actually for, and that the discipline has become too focused on methodological complexity rather than the decisions which the numbers are supposed to enable. Since then, I have sailed this thought over the horizon and into uncharted territory, and have arrived at a new island on which to drink rum and plant my flag.
Even if we make carbon accounting better, does it ever produce emissions reductions?
I think it’s fair to say that over the past 20 years, a measurement-industrial complex has emerged. Companies have their hands in their pockets more than clothes mannequins trying to pay for all the manpower and tools required to measure and report on greenhouse gas emissions. They pay consultants to help them develop their inventories, software companies to provide a shiny new wrapper for the same uncertain data, assurance providers to raise a few cursory issues before signing off on numbers which are extremely difficult to independently verify, SBTi to validate targets that they don’t have to meet, and countless internal man hours for downtrodden sustainability managers to go around the internal and external houses looking for data that invariably arrives a day late and a dollar short.
I should point out here, if I’ve not already lost you, that I have spent most of my career so far working across that system – and by no means do I argue that anyone doing any of those jobs is somehow to blame. What I am saying is that there is a disconnect between the goals and outcomes. A complex and costly global ecosystem has been built around simply calculating and reporting organisational emissions, largely on the assumption that doing so will eventually contribute to reducing them, with very little of the total cost directly going towards decarbonisation. It’s surprisingly difficult to work out exactly how much bees and ‘oney is spent on the ecosystem globally, but ERM estimates that an average large US company spends about $237,000 a year on GHG analysis and disclosure. CDP’s Corporate Environmental Action Tracker contains around 12,300 companies, so multiplying one by the other gives a very crude annual bill of about $2.9 billion, before we have included companies outside that dataset or much of the wider ecosystem of software, assurance, target validation and disclosure. For the purposes of this post, I think that puts us in the order of billions of dollars per year, against a backdrop of a world where emissions continue to rise. Obviously, I’m not suggesting that carbon accounting has caused emissions to increase, or necessarily that it has had no impact but, in that context, I think it is fair to ask what have those many billions bought?
I am not aware of any evidence for the impact of carbon accounting on global emissions, so I’d like you to bear with me for a moment while I do an irresponsible estimation to illustrate the problem. Those same 12,300 companies report around 9 billion tonnes of Scope 1 emissions annually through CDP. If their emissions fell at the 4.2% annual rate historically required by SBTi for a 1.5°C-aligned target, that would represent around 378 million tonnes of annual reductions. What proportion of those emissions reductions do you think are actually attributable to the measurement-industrial complex? 10%? 5%? 1%? I don’t have a scooby, and nor do you, and that is the whole point. If we generously assumed that measuring, reporting and target-setting were responsible for 10% of these hypothetical reductions, that would put the abatement cost associated with the measurement-industrial complex at around $80 per tonne. At 5% it would be around $150, and at 1% nearly $800. More recent CDP analysis found that only the top 15% of companies were reducing emissions at around 4% a year, while the rest were averaging closer to 1%. Using 1% rather than 4.2%, which is more indicative of the real reductions reported for the rest of the companies in that analysis, increases those abatement costs to around $320, $640 and $3,200 per tonne respectively. Even the act of putting some numbers on it makes me question whether this is the best way we could be spending money on climate.
The difficulty is that the decarbonisation effects of measurement and reporting are largely second order. Calculating an inventory can inform a target, which may influence a decision, which might lead to an investment, which could eventually reduce emissions. Where are those emissions reductions actually coming from, though? Typically, simple interventions like paying money to upgrade your fleet to electric vehicles, switching to renewable electricity, getting lower carbon materials, all of which can be informed by simple, product level analysis. I am completely unconvinced that you need the level of complexity which we have built around organisational carbon accounting. In fact, I’ve got a business pitch for those 12,300 companies – I’m 50% cheaper than your current providers. If you pool together and give me $1.45 billy a year, I too can tell you to stop burning fossil fuels in your operations and stop buying things from factories that burn fossil fuels. You can use the other $1.45 billion to spend on replacing your fleet and boilers (or bonuses).
The basic problem, which I hope I’ve illustrated, is that emissions are systemic and simply reporting them doesn’t equal reducing them. The measurement-industrial complex costs billions a year, all for the purpose of assigning responsibility for emissions to individual organisations with ever growing complexity. A company reports emissions associated with the parts it buys, the part manufacturer reports emissions from the steel that it buys, the steel company reports the emissions from making it, and a bank might report some proportion of the parts factory in its financed emissions. Each one of those organisations spends lots of cash working out what moral responsibility they have for those emissions so they can slap it into a sustainability report which nobody is going to read. What we completely ignore, though, is that having emissions attributed to you does not mean that you have the agency to eliminate them. A supermarket can calculate the emissions from fertiliser used to produce the food that it sells, but it cannot independently commercialise green ammonia. A construction company can calculate the embodied emissions from cement, but it cannot independently develop low-carbon cement, so what happens is that we end up with ever more complex emissions inventories without necessarily getting any closer to the solutions.
The GHG Protocol, which recently announced that its consolidated corporate standard will not be published until Q4 2028, has now gone further into the weeds than Ash Ketchum. Furthermore, they’re so exposed to competing industry interests that it’s hard to even be sure that the updates will be in the interest of better accuracy. It appears to me that we just have hundreds of good people, voluntarily squandering their precious endowment of minutes alive on Earth, talking about hourly reporting of electricity emissions, market instruments and multi-statement reporting, without any way of quantifying their impact on global emissions. How much more money are we going to pour down the drain trying to get better at allocating these things? I know that there are good intentions behind much of this complexity, and I know many of the people involved are genuinely trying to fix real problems with the existing standards but it’s all so myopic. Ask yourself this: if companies comply perfectly with whatever new accounting system emerges, what exactly do we expect to happen to global emissions as a result? How do these new complexities genuinely drive reductions, and can we clearly quantify the expected reductions? I feel the same about the science-based targets initiative. Setting a target is obviously better than having no intention of reducing emissions at all, and there is some limited evidence that companies with targets subsequently reduce certain emissions. However, targets as they are currently set are just a statement about something that you intend to happen in the future. Nobody is bound to achieve them and there is still limited evidence on how consistently organisations actually reduce emissions in line with the pathways, pathways which they spent untold Benjamins assessing and having validated.
If you’ve stayed with me this far, thank you, I’m about to give you your lump of sugar (a half-baked solution to this problem) and pat you on the nose. What if, instead of asking companies to spend ever increasing amounts of money determining and reporting which emissions can be attributed to them, we asked them to redirect that money to solving the hardest decarbonisation problems?
Imagine that every company which does emissions reporting instead paid either the money it currently spends on those activities, or a tiny proportion of its profits, into a global decarbonisation fund. That money could then be directed towards the most cost-effective technological decarbonisation opportunities globally, or towards areas where we already know enormous technological and financial barriers remain, like low-carbon steel, cement, aluminium, fertiliser, shipping, aviation, long-duration energy storage and so on. Rather than thousands of companies independently spending money producing Scope 3 strategies which eventually arrive at the same fairly obvious conclusion that their steel suppliers need to decarbonise, while having very little agency to make that happen independently, those companies could collectively put the same money into a pot which, once aggregated, might actually be large enough to help make low-carbon steel commercially viable. I think using the carbon accounting money in this way would have a better chance of enabling the 4.2% annual reductions we need. Rather than spending billions repeatedly identifying the same problems across thousands of individual companies, we could redirect those resources towards solving the difficult technological challenges which, once solved, would enable deep and rapid reductions across all of the companies whose emissions are fundamentally driven by the same legacy processes.
There is also another fairly obvious benefit in comparability. Rather than trying to compare two Scope 3 inventories built using completely different data and methodologies, you get one number which everyone can understand. How much money did you actually cough up to address climate change relative to how much money you made? It’s crude, but completely verifiable, and gives investors, consumers or anyone else a simple way of working out whether a company is actually putting its money where its mouth is.
I know there are enormous problems with this idea, and that I’m possibly straying into “We send the EU £350 million a week, let’s fund our NHS instead” territory. Firstly, I don’t know if I believe the contribution should be mandatory or voluntary. I don’t think there’s any viable mechanism for making thousands of multinational corporations pay into a global decarbonisation pot. It is essentially a tax, and people might rightly ask why that money isn’t being collected and distributed at a state level. If it’s voluntary, only some companies might pay in while the rest free-ride, although this isn’t dissimilar to the current situation with voluntary reporting and decarbonisation action. There’s also the very difficult question of how much everyone should pay. Should it be based on profits, the emissions impact of your organisation, or simply what you already spend on carbon accounting and related services? Once you figure all of those things out, you then get to the real issues, like who manages the money and which decarbonisation technologies qualify for investment. Directing billions towards specific technologies would invariably create winners and losers, but not before an enormous amount of lobbying over what actually counts as a decarbonisation technology. Perhaps funds like this could instead be set up at a sector level. Every company in the construction sector could pay in, for example, with the money then directed towards solving shared problems like low-carbon cement and steel. This would at least create a clearer link between the companies paying into the fund and the emissions problems the money is being used to solve.
Photo: Getty Images via The Independent
The question I am really asking, though, is much simpler. Which do we think would reduce emissions by more, putting $2.9 billion directly into physical decarbonisation, or spending $2.9 billion independently measuring and reporting emissions across thousands of companies? I strongly suspect it is not the latter, and even if my fund is a completely terrible idea, I am sure somebody who knows more about economics than me could think of a better way of spending that money.
Maybe I’ve got it all wrong. We do clearly need some measurement to understand where emissions come from, identify important sources and determine whether reductions are actually happening. That said, I think that the GHG Protocol, SBTi, software platforms, consultants and all other members of the measurement-industrial complex need to have a long look in the mirror and ask themselves how much measurement is enough, and what we are giving up when we continually demand more and more complex versions of it. If a company has £500,000 available to spend on climate change next year, would you rather that it used that money to produce a misleadingly precise Scope 3 inventory (that will be useless in 2028 when the rules change) and have its non-binding targets validated, or put £500,000 into a low-carbon cement plant? As much as I respect my colleagues and friends in the industry, I don’t think we have enough evidence that the money we spend on the former is producing the emissions reductions we need.