Limitations

The GHG Protocol has three scopes: direct emissions (Scope 1), indirect emissions (Scope 2) and upstream and downstream emissions (Scope 3). It allows some flexibility in reporting, leading to variations in granularity across areas like the reporting of fluorinated gases in direct emissions, renewable energy, indirect emissions (market- vs. location-based) and upstream and downstream emissions (Scope 3). Reporting the latter is currently optional, only requiring companies that report on Scope 3 to follow the Scope 3 standard (Greenhouse Gas Protocol Corporate Value Chain (Scope 3) Accounting and Reporting Standard).

It is important to note that the GHG Protocol itself does not have the goal of comparing companies with each other. Thus, as companies often alter their reporting annually, making consistent comparisons across companies, as they are depicted in the interactive visualisation, is challenging. To address these data gaps, we included a table in the next section and an interactive company CSR database to guide the careful interpretation of our findings.

According to the GHG Protocol, companies need to account for 100% of emissions from operations over which they have control, so for owned subsidiaries, they need to report data. If it is a joint operation, they can either apply an equity share or control approach. The equity share approach means that if a company owns 50% of a joint venture, it accounts for 50% of total emissions. The control approach, by contrast, only obligates it to account for emissions it has (financial or operational) control over (The Greenhouse Gas Protocol Revised Edition). One approach must be selected and applied consistently.

In most cases, this level of granularity is not provided. Often, companies make the inclusion of subsidiaries in their CSR reporting transparent, noting this practice in footnotes, unless stated otherwise. Some companies also provide more transparency by assigning separate values to their subsidiaries. However, in certain cases, it is unclear which method they chose and whether subsidiaries were included in the reported data. Importantly, when companies acquire or divest subsidiaries, the GHG Protocol framework allows the revision of historical data retrospectively to reflect these changes. For consistency, we assumed that subsidiaries were included in the reported figures and used the latest available data for the analysed years.

As previously mentioned, a key challenge when analysing CSR reports is that companies often revise their data. These revisions may be due to mergers and acquisitions, updates to global warming potential (GWP) values, the inclusion of additional GHGs and changes. In addition, indications regarding these changes are often hard to find. This often goes hand in hand with the recalculation of base year emissions as recommended by the GHG Protocol (The Greenhouse Gas Protocol Revised Edition). Our analysis always attempts to use the latest available data to maintain consistency.

Companies do not use a common definition of energy. This leads to the challenge that some companies use energy and electricity interchangeably, while others use energy purchased rather than energy consumed. The interpretation of "renewable energy" or "green energy" also varies among manufacturers, with some including RECs and PPAs (reported in Scope 2), while others focus on local grids or on-site generation, such as wind and solar power (reported in Scope 1).

This can be explained by the GHG Protocol guidance to include a variety of instruments—RECs and PPAs, among others—as "green power programs" in market-based emissions in Scope 2 without defining what should constitute "green energy" (GHG Protocol Scope 2 Guidance). Thus, aggregating very different definitions into a single figure often obscures specific contributions and inflates renewable energy shares. As a result, we chose not to display the renewable energy usage of individual companies in the charts.

Companies vary in how granularly they differentiate emissions from manufacturing (front- and back-end) versus non-manufacturing parts, such as offices. Since non-manufacturing sites typically have much lower emissions, this limitation likely has a minimal overall impact.

Similarly, it is often unclear whether companies focus exclusively on front-end manufacturing in their reporting. A clear differentiation between the ecological footprints of these two stages is rarely provided, although strong evidence suggests that front-end manufacturing has a much greater impact (Hess 2024).

Analysing data from CSR reports is challenging due to varying metrics and primary energy factors (PEFs) across countries. For instance, while the basic conversion from gigajoule (GJ) to megawatt-hours (MWh) is fixed (1 MWh = 3,6 GJ), the amount of primary energy required to generate electricity varies by country. This is due to differences in energy sources, power plant efficiency and national standards for calculating PEFs. For example, Korea applies a different PEF (2.67) compared to the EU (1.9), reflecting its energy mix and grid efficiency. These inconsistencies complicate direct comparisons, so we carefully adjusted the data to ensure consistency and accuracy.