Assessing Corporate Decarbonization

assessing corporate decarbonization in a world of volatile emissions growth

*Post was originally written on March 11, 2022. You can find a twitter thread summary here.

This past week saw a few notable updates on global and corporate emissions efforts that spurred some dialogue amongst some US team members on the importance of assessing decarbonization efforts with context.

Microsoft, which is viewed by many as a sustainability leader, reported their 2021 Sustainability Report[1] and Update[2] on Thursday which revealed that absolute emissions across Scopes 1, 2, and 3 went up. Microsoft has aggressive ambitions to be carbon negative on a Scope 1+2 basis by 2030 and reduce Scope 3 emissions by 2030 and has been reported to be the world’s largest corporate purchaser of carbon removal today (~1.4M tons in 2021).

The NYT ran a piece[3] which criticized Microsoft for the annual rise in emissions relative to their goal of being carbon negative. I’d expect this type of rhetoric to increase as the world nears irreversible tipping points as a result of climate change and understand them but it’s important to understand these figures and criticisms with appropriate context. I firmly believe Microsoft is taking aggressive steps to reduce their emissions, setting proper incentives for their teams to executive on that mission, and demonstrating their products are designed to help avoid future emissions relative to alternatives. I also think YoY volatility in emissions figures – especially those out of their control – should not always be taken at face value. This does not mean any company gets a ‘pass’ on emissions growing but it shows the required nuance for a lot of these discussions/criticisms.

As you can see below, Microsoft reported updates to their annual emissions from 2017-2021. On a total emissions basis, we see that emissions rose, primarily as a function of Scope 3 emissions increasing +23% YoY. (*Please note: Carbon Table 2 above uses market-based emissions vs. location-based emissions you’ll see in Table 1 below. A location-based method reveals what the company is physically putting into the air and the market-based method shows emissions the company is responsible for through its purchasing decisions – e.g., renewable energy and carbon removal.)

As you’ll see below, Scope 3 represents ~98% of Microsoft’s total emissions. The bulk of these emissions (up to ~93%) come from 3 categories 1) Purchased Goods & Services (~35%), 2) Capital Goods (~30%), and 3) Use of Sold Products (~28%). These emissions include, among other things, the carbon emitted from electricity generation to power devices at home or at work and for producing the concrete and steel Microsoft uses in construction. Microsoft works to limit these emissions when they design and manufacture products, and work with suppliers to reduce their emissions—but still saw an increase. This serves as an important reminder that Scope 3 emissions are the most difficult to control and reduce. This year has highlighted the challenges of reducing Scope 3 emissions. Microsoft called out a few of the reasons for the increase below:

• One Scope 3 category that saw an annual increase was capital equipment. This category includes the emissions to produce equipment we bought this year (and will use for a long time) as well as the carbon emissions associated with the construction of new buildings including both datacenters and office space. For example, as the Microsoft Cloud business expands, construction of new datacenters and new server equipment is needed to support growing customer demand. The emissions associated with the capital equipment category accounted for about 30 percent of our Scope 3 total this year.”
• “Another Scope 3 category that saw an increase was the use of our products, specifically the estimated emissions associated with powering products we sold including Xbox and Surface devices. This category accounted for about 29 percent of our Scope 3 total this year. Around the world, the electricity used to power devices today largely comes from grids reliant on carbon intensive sources such as coal or gas, so Microsoft must account for that carbon for the lifetime of our devices to accurately track our Scope 3 emissions. And, because every part of the world has a different mix of electricity generation sources, we calculate a device’s climate impact based on where the device was sold.”

They connect all of these ideas very well with the following quote: “The rate at which we can implement emissions reductions is dependent on many factors that can fluctuate over time, ranging from our own business growth and supplier mix to the rate of growth of green infrastructure, such as the supply of and transition speed to renewable energy. Additionally, different sources of emissions will be addressed on different time frames as climate technologies and renewable alternatives advance. And as measurement methodologies improve and new standards emerge, companies like ours may see emissions rise or fall year-over-year”

Further, I think some important context is required around use of sold products for a company like Microsoft’s. For example, Microsoft’s cloud services can be up to ~93% more energy efficient and up to ~98% more emissions efficient than traditional enterprise datacenters[6]. As datacenters proliferate around the world, one could make the argument that Microsoft’s growth in that space is preferred versus alternatives. This brings up the topic around ‘Scope 4’ or ‘Avoided Emissions’ that requires a separate conversation but wanted to highlight the evidence around the integration of sustainability into products and services. Further, outside of energy efficiency targets on products like Surface or Xbox, until all grids are decarbonized they will continue to have emissions from use of those sold products and will likely need to offset those with carbon removals.

I think that is a really fundamental takeaway that while our models and estimates may show linear/smooth lines of progression towards 2030 and 2050 targets, in reality, they will be a lot more choppy and may have volatility in any given year. I’d hope that companies continue to ultimately push towards decreasing emissions as fast as possible but also need to be realistic about where the sources of emissions are from, where they have the most ability to influence those emissions, and the role of external factors on their own emissions (e.g. global power mix).

Speaking of that last point, the IEA released an update on global energy-related emissions this past week and Global CO2 emissions from energy combustion and industrial processes rebounded in 2021 to reach their highest ever annual level[4]. Outside of just Microsoft, this has implications for how emissions are calculated up (energy used to manufacture products or process raw materials) and down the supply chain (energy used to power products like Microsoft’s).

I would note, that investors are paying attention to emissions intensity (Emissions / Revenue). Ultimately, this metric allows us to look at emissions growth relative to the underlying growth of the business (as measured by revenue). As an example, you can see Microsoft’s carbon intensity for their wholly controllable inputs (Scopes 1 + Scope 2 + Scope 3 Business Travel) trended downward.

On a global level, the average emissions intensity of global economic output stayed constant at 0.26 tonnes of CO2 per $1,000 but it actually ticked up slightly in advanced economies. This is a worrisome trend as prior to that, it increasingly appeared that GDP Growth and Emissions Growth had been increasingly decoupling[5]. Given everything going on with the world currently, it’s reasonable to assume this figure might display volatility in years to come, especially in the EU as a result of energy security/energy independence policies.

All of this is relevant to address the relative speed and progress of corporate decarbonization efforts. A company like Microsoft outlines specific programs around how they plan to manage emissions down across their business in robust detail but may still run into blips where total emissions figures trend the wrong way for a short period of time. Some of the recent updates in their annual report include: connecting performance against emissions targets into compensation for senior management, increasing their internal carbon fee on Scope 3 travel to $100/ton and boosting efforts on sustainable aviation fuel procurement, investing ~$570M invested into climate initiatives/investments, ramping up their carbon removal purchases (~$20-30M total spend range but increasing), and releasing a new policy that they will not do specialized work for energy businesses without a 2050 net zero target (they can still access commercially available software and cloud services).

I’d note that Microsoft is a small emitter in the grand scheme of things (a tiny fraction of material industries like Energy/Steel/Concrete/Commodities) and the dynamic between Scopes 1, 2, 3 emissions and control over their supply chain varies drastically between sectors, I thought it was an interesting case study of why the climate considerations for companies require nuance and context.

Sources:

1. Microsoft 2021 Environmental Sustainability Report
2. An update on Microsoft’s sustainability commitments: Building a foundation for 2030
3. Microsoft’s Pursuit of Climate Goals Runs Into Headwinds
4. IEA Global Energy Review: CO2 Emissions in 2021
5. Absolute Decoupling of Economic Growth and Emissions in 32 Countries
6. The Carbon Benefits of Cloud Computing: a Study of the Microsoft Cloud