What makes a credible corporate climate pledge?

The British government likes to advertise its contribution to cutting carbon emissions. And the figures do look good at first glance. The UK was apparently doing a good job of meeting commitments agreed at the end of 2015 well before ministers even turned up in Paris for a signing session.

UK emissions peaked 50 years ago, a time when coal was the predominant fuel for both heating and electricity generation. Though investment in its planned replacement, nuclear, petered out, the once proud ‘dash for gas’ since the end of the 1980s helped cut overall emissions. The UK’s estimated greenhouse-gas (GHG) output almost halved in the three decades since 1990, according to figures from the Office of National Statistics (ONS) published in February 2022. It’s a remarkable reduction and the kind of shift that is possible for any nation. As long as you omit some important information.

As activist Greta Thunberg pointed out in a briefing for Unicef in response to an earlier set of ONS statistics in August 2021: “They are very creative at carbon accounting, I will give them that.”

To achieve such reductions does not really take a whole lot of creativity: the metrics traditionally used for assessing emissions make it all too easy to do. One of the biggest contributions to lowering GHG emissions is to cut demand and the UK has, apparently, done particularly well on that front. ONS figures show industrial demand for energy dropped by two-thirds from 1970 to 2019, by exporting it.

The government is not unusual in shifting its emissions to third parties. A number of companies, such as PwC, claim to be ‘carbon-neutral’. Based on their direct use of energy, this is a justifiable claim. Much of the management consultancy’s carbon neutrality is met by powering its own offices with electricity from suppliers with guaranteed output from renewables. On top of that, since 2007, it has bought carbon credits to offset the emissions from staff activities such as business travel. In combination, PwC can claim to generate no greenhouse gases on a net basis.

Attention is now turning to the bits that routinely get omitted from carbon accounts – not just at intergovernmental level, such as the UK’s offshoring of its industrial base and the associated emissions, but even in corporate finance as companies respond to pressure to report much more clearly on their environmental impact.

The upstream contribution, for the most part, lies in embodied carbon: the amount of greenhouse gas emitted in the production of the stuff these organisations, and ultimately the rest of us, buy from other suppliers.

Apple’s disclosure to the climate-change reporting charity CDP shows how much outsourcing can knock off the carbon bill if you choose not to quantify those numbers. The company accounted for emissions of around 60 kilotonnes of CO2e for its own operations and electricity and heating bill in its August 2022 CDP response, which covers the financial year ending September 2021. That output was dwarfed by the emissions the company estimated were produced by the network of subcontractors responsible for making and shipping its products, as well as the estimated emissions by the use of its products and those that come from the staff travelling to and from work and for business. Apple’s total, as reported to CDP, came to just over 23MtCO2e.

In fairness to the consumer and computer giant, Apple has provided metrics informally on the emissions generated during the manufacture of its machines for well over a decade. This makes it comparatively easy to see how things have changed in its products and how front-loaded emissions are for some classes of product. The vintage-2008 15-inch MacBook Pro was accompanied by an environmental report that showed for the average user, the emissions generated during production would most likely account for a little over half its lifetime total of 560kgCO2e.

Now fast-forward to the slightly smaller 13-inch MacBook Pro launched last year. Here, the ratio of production emissions has soared to 71 per cent even though, overall, it’s a cleaner design. It is responsible for less than half the lifetime emissions of its 15-year-old predecessor.

Apple’s approach to disclosure is becoming commonplace for larger companies as they seek to benefit from claims to be advancing to net-zero status faster than their competitors. This trend is also driven by pressure from environmentalist groups such as Make My Money Matter and As You Sow, which are independently trying to push banks into confining their lending to green industries.

Activities like these are influencing the way in which investment portfolios are assembled by pension funds and other large sources of money to the stock markets, though they have more room to manoeuvre. A 2022 paper by analysts working at Blackrock found greenwashing a stocks portfolio under traditionally accepted practices is an easy process. Choosing the ACWI mid- and large-cap index managed by MSCI for part of their analysis, they found energy, utilities and materials producers accounted for more than 80 per cent of direct emissions but less than 15 per cent of total market capitalisation. “This property makes it easy to ‘decarbonise’ a portfolio with some select underweights,” they wrote.

As external pressure mounts to make net-zero plans clearer, more companies are signing up to transparency plans, though there are plenty of loopholes that they can exploit. If you look at a completed CDP questionnaire, the reporting methodology has a lot in common with financial reporting: to make sense of the headline figures, you need to look carefully at the footnotes. Though CDP provides a list of approved methods for various materials and supply-chain components, respondents get to choose which of them they report on.

The Science Based Initiatives (SBTi) project has gone further by bringing in an approvals process for company-set targets. However, groups such as the Germany-based NewClimate Institute have criticised corporate plans approved by the SBTi for leaving out key contributors to emissions in their supply chains.

So far, several thousand businesses have committed to introduce plans to meet net-zero. Cisco, for example, aims to reach that status by 2040.

The timescales for Cisco demonstrate the difficulty of going to full net-zero operation that takes everything into account. The networking-hardware supplier expects to be able to cut emissions from its own operations and the heating and electrical energy it sources by 90 per cent by the middle of this decade. Management expects the rest of the emissions to fall by only 30 per cent by 2030.

The easier fraction falls into what are known in carbon accounting circles as Scopes 1 and 2. Scope 1 covers direct carbon emissions: mostly this will be the gas emitted by burning methane for heating and fuel for a fleet of trucks. This is reasonably easy to assess and so features heavily in the carbon-neutral claims made so far. Scope 2 emissions are the first level of indirect usage, which makes them harder for an end user to assess. Typically, this covers electricity generation. Although companies that buy electricity generated by nuclear or renewables can make reasonably confident claims, much depends on what the suppliers use versus what they report to their clients. 

The contribution of biomass to the balance of GHGs is a thicket of conflicting estimates. Much of that is because biomass as a category covers a wide variety of sources, all the way from agricultural slurry and household food waste processed locally to wood from old forests sliced and diced into pellets that are then shipped to a different country to be loaded into a furnace. Generating companies such as UK-based Drax argue their use of wood pellets represents not just a carbon-neutral approach to electricity generation but one which is potentially carbon-negative on the basis the generator will in the future apply carbon capture at the turbine.

The UK government agrees, enabling Drax to claim payments under the Renewables Obligation scheme; rising gas prices have seen these arrangements become hugely profitable for generators. Burning pellets for power is regarded as clean because it’s assumed that more trees will replace those felled, absorbing CO2 as they grow.

One point on the side of the biomass burners is that using wood pellets for fuel is only one way in which the embodied carbon from trees will be released as GHGs. For example, left to decay or burned in fields, agricultural biomass can release methane, which carries a higher GHG intensity. But circumstances vary dramatically. Studies of carbon flows have found that some kinds of grass grown on land that is otherwise of little agricultural value offers a good trade-off, whereas old forests being felled and replaced may not become effective carbon sinks for decades after the originals were felled and their embodied CO2 released.

This variance leads to what a group of non-governmental organisations called an accounting loophole in their evidence to BEIS’s summer 2021 call for evidence on bioenergy, arguing Drax and similar generators should not receive clean-energy subsidies. Another issue with the use of wood pellets in the UK is that many of them are imported.

“Bioenergy is treated as zero-emitting in the UK energy sector and elsewhere only because UNFCCC accounting rules assume carbon emissions associated with biomass removals will be properly accounted for in the land use sector of the country where it is sourced. However, many countries do not account fully, or at all, for their land use emissions,” the NGO group claimed. “As a result, large quantities of emissions associated with bioenergy are simply ‘missing’ from international ledgers.”

Despite the complexities of Scope-2 calculations that quickly emerge as you dig into the fine print, Scope 3 is even more difficult to assess and equates to the UK’s own exporting of emissions.

Corporations such as Apple and Schneider Electric have formed extensive programmes that are designed to try and capture this data from suppliers. But these programmes are far from commonplace, and even businesses that have embraced the idea of including Scope 3 are still feeling their way. Submissions to CDP indicate that companies can take quite different views on what they need to include in their Scope 3 reports. Some include the emissions generated by the user base. Others focus on the upstream supply chain: how much greenhouse gas is generated by mining, refining, production, assembly, and shipping from subcontractors. Some, such as Apple, do both.

The overall contribution of the user base differs substantially based on the class of equipment. Whereas Apple’s consumer products are clearly dominated by embodied carbon, industrial-equipment supplier Schneider Electric estimated in 2022 that more than 80 per cent of the carbon emissions come from its customer base, though many of its control devices may themselves reduce the GHGs generated by heavy equipment.

On top of the vagaries of estimation, a potential issue with the reporting of Scope 3 numbers lies in double counting. Should Apple or Schneider need to report the emissions of its business users when those other organisations include them as part of their own Scope 2 calculations? The suppliers have limited visibility other than national averages into what levels of renewables are supplying the electricity that powers their products in the field or indeed for how long those products are in service.

Similarly, though Apple, Schneider and others say they are looking more closely at what suppliers use, they currently need to make broad estimates of the emissions from materials and fuel production because, so far, comparatively few of the resource-extraction companies file any reports on their own emissions. As with biomass, the estimates could be overly optimistic or pessimistic. As more accurate reporting comes in, the prospect of double counting could present another issue because Scope 3 numbers begin to overlap with Scope 1 and Scope 2 figures both upstream and downstream.

Yet some researchers believe double counting may have benefits as a mechanism for encouraging more efficient use of fuels and materials. Work with economic models by Professor Felipe Caro and colleagues at University of California at Los Angeles indicated a decade ago that double counting provides a stronger incentive to change than trying to come up with a mechanism that allocates each tonne of CO2e to a specific organisation.

There is a potential answer to the problems raised by self-reporting: use economics to push suppliers and customers in the right direction. For most economists in this sector, the solution would be to give carbon a price that recognises its impact.

“Carbon pricing would support a market-​led approach to reducing emissions if it was applied to all three scopes of emissions,” says Kelly Widdicks, lecturer in computer science at Lancaster University and co-author of a paper calling on the ICT industry to account for its full carbon emissions. “However, we will still need transparent evidence on ICT’s emissions and the reductions in these as we align the digital sector, and global sectors generally, with the Paris Agreement, ensuring consistent approaches across organisations to this evidencing and action.”

The big practical problem with using carbon pricing as a policy lever is that the price set has to be global. No one currently believes the international community is anywhere close to agreeing a realistic carbon price, which would likely be on the order of $100 per tonne of CO2e.

Caro and colleagues envisaged that the prospect of double counting would lead large corporations to set their own internal or shadow carbon price in negotiations with suppliers as an incentive to drive down emissions across the supply chain.

So far, it seems that corporations currently favour engagement, though some including Drax and the European Bank for Reconstruction & Development have decided to use a shadow carbon price for at least some operations and investments. Drax’s internal carbon price follows the UK Emissions Trading Scheme, which fluctuated between £70 and £100 over the past year.

Even for those using engagement, the implication is that suppliers which do not meet targets will ultimately stop being suppliers.