Is Additionality More Important than Comparability?
It is commonly assumed that we read company reports to compare one company against another, but does this notion stand up to scrutiny?
The comparability assumption undoubtedly has some truth, especially when two or more companies have analogous business models and can be evaluated using similar benchmarks. Potential employees, advocates, and investors (especially active rather than passive investors) all have good reasons to compare company risks and opportunities and assess which company has the most effective strategy. This assumption is typically accompanied by the premise that we should develop more standardized metrics and disclosure requirements to improve like-for-like comparisons.
However, this focus on comparability can lead us to overlook another important reason for reading company reports: additionality. In other words, understanding the collective whole of company reports—i.e., what a compilation of reports tells us in combination—may be as or more important than comparing companies against each other.
The recent publication of systemic risk assessment reports by the world’s largest social media companies (as required by the EU Digital Services Act) provides a helpful test case. The purpose of these reports was to better understand systemic risks in the EU arising from the operations of social media companies, such as risks to human rights, civic discourse, and public health. While the term “systemic risk” was not defined by the EU, it is commonly taken to imply that a failure in one part of a system can cascade and cause damage elsewhere in a system and that broader societal harm may result.
In this context, report readers likely gained as much from considering what the reports told us about overall social media risk in the EU as they did from directly benchmarking companies against each other. The most interesting analysis was to understand how risks arise from social media platforms in combination, such as the respective roles of search, app stores, and user-generated content services in spreading and addressing hate speech, misinformation, and terrorist content. We read the reports to understand each company's leverage to address these risks and how the different social media platforms interact—not just to compare them against each other.
In other words, the most helpful task was not to compare companies A, B, and C (or X) and ask, “Who did it best?” but to consider the combined risks of A, B, and C and ask, “What do they add up to?” It was about additionality, not just comparability.
Additionality is not a free-for-all and, like comparability, requires standardization. Companies should apply the same principles, standards, and methods for selecting and presenting information and ensuring information quality and integrity. Common report structures, taxonomies, and reference points also help readers. However, the importance of additionality does raise the question of whether efforts to achieve perfectly comparable metrics and disclosures may have diminishing returns—perhaps we should spend more time reflecting on the combination of information we desire, even if it is not directly comparable.
Direct comparability is essential for some topics, especially those that lend themselves to standardized measurements, like greenhouse gas emissions, diversity, and privacy breaches. However, many other issues do not lend themselves to standardized measurements, and in these cases, differentiation in how companies report can be more instructive than their similarities.
It is also important to remember that many report users have a stake in systemwide impacts rather than whether one company outperforms another on industry benchmarks. For example, investors with stakes across the whole economy (i.e., most passive funds and index trackers) cannot diversify away from climate risk and are only successful if every company plays its part in combination. A company may “improve” its climate resilience by selling its fossil fuel subsidiary, but the investor gains nothing if they also own the buyer's shares—and even if they don’t also own the buyer’s shares, the fossil fuel subsidiary still increases climate risk for the remainder of their portfolio. Similarly, in my systemic risk assessment example, one company’s excellent work to remove terrorist content can be undermined if that content is easily cross-posted from a more negligent platform.
The faulty complaint that we can't compare apples to oranges perfectly illustrates the dominance of comparability over additionality in our narrative about company reports.
My problem with this complaint has always been that we can compare apples and oranges. Both are edible fruits, but one is orange while the other is green or red. Both are good sources of vitamins and fiber, but apples are higher in fiber, and oranges are higher in vitamin C. We can bite an apple, but an orange must be peeled or cut.
However, I have realized that the underlying question about comparing apples and oranges is wrong, not just the answer. We shouldn’t just be interested in comparing apples and oranges but also consider how they combine with other fruits to form an overall fruit salad and what impact different combinations of fruits have on our dietary health and taste buds. We eat the whole fruit salad, not just parts of it. The same is true with risk.