Reducing greenhouse gas (GHG) emissions is among the momentous and most urgent challenges humanity is facing1,2. The problem is that mitigating climate change is essentially a global social dilemma, where mitigation efforts are very costly for individual countries (or firms), but the benefits are shared globally and would be realized in the future3,4,5,6,7,8,9. An additional barrier for successful mitigation stems from the considerable heterogeneities between countries regarding wealth, projected loss from climate change, historical emissions, etc1,10,11,12,13,14,15,16.

We are in the midst of a digital revolution in production17. With the Internet of Things and an increasing number of connected devices, raw data are often created as by-products of firms’ daily operations at very low costs; for example, logistic data from a company (such as volumes of truckloads, GPS truck positions, truck routes, etc.). Following Varian’s18 definition, data are raw alphanumeric values that need to be organized and contextualized to be turned into information. Knowledge is the conclusions and insights learned from that information. Data-based knowledge allows firms to gain better control over their production and supply chains, and the opportunity to reduce their costs and emissions. While firms are hesitant to share confidential information and knowledge with other firms, they are more willing to share encrypted data19.

This paper deals with data sharing among firms along a supply chain for the calculation of product-level Scope 3 emissions; e.g., data on all material and energy inputs, purchased product components, transportation, and their direct emissions. It answers the following questions: What are the benefits of primary data sharing for Scope 3 emissions? What are the barriers for successful emission data sharing along the supply chain? And what are the current developments in overcoming these obstacles? Before answering these questions, we provide some background on Scope 3 emissions and their importance.

Effective data sharing became possible due to recent technological developments (e.g., digitalization in production and processes, increase in computer power, new encryption technologies, etc.). As mitigating climate change is a social dilemma, data sharing can increase the transparency of the situation and the accountability of the actors20, and can even transfer the social dilemma to a coordination game21. Thus, the overarching motivation in sharing primary data for Scope 3 emissions is to optimally mitigate GHG emissions. In particular, by sharing company data with other firms, data are contextualized and turned into information. Latest innovations in cryptography allow the computation or analysis of shared data without requiring the parties to reveal their private inputs. The encryption reduces the risk for data owners to reveal sensitive data or lose control over it, and thus increases the willingness of firms to engage in data sharing19. With information derived from their data-sharing activities, firms can conclude (i.e., gain knowledge) about parts of the supply chain (e.g., emission reduction potentials).

We focus on Scope 3 emissions along the upstream supply chain (“cradle-to-gate”). In this regard, an important distinction is between “cradle-to-gate” and “cradle-to-grave”. The former includes all upstream emissions in the production process (i.e., along the supply chain up to the reporting firm), whereas the latter also includes downstream emissions from the consumption of the good. Currently, much of Scope 3 data estimates are calculated based on industry averages, financial information, and approximations20,22. Such approximations sometimes yield close estimates to the actual emissions of a product, but other times the derived values are much less accurate. For instance, Downie & Stubbs23 find that when conversion information is not available, there are wide discrepancies in reported emissions for the same activities among Australian firms. In this respect, corporate environmental performance (CEP) rating schemes do not yet provide a sufficiently good measure for emissions, and current CEP rating databases lack consistency24. The problem of using industry averages is that they may lead to insufficient incentives for firms to adopt cleaner-than-average technologies. By contrast, primary emission data that is shared along the supply chain can provide exact emissions on product- and firm-level. Thus, it allows firms to tackle emissions reduction at the most effective point on the supply chain, and also provides incentives for the adoption of cleaner technologies beyond the industry average.

The structure of the paper is as follows: In the next section, we introduce the notion of Scope 3 emissions. Section 3 reviews the main obstacles for sharing primary data for Scope 3 emission reporting. Section 4 explains what is required to overcome the main data sharing obstacles and discusses the current development in overcoming these obstacles. Finally, Section 5 concludes.

Scope 3 emissions: Introduction and current estimation

Emissions generated in the production process can be classified as direct or indirect emissions. Direct emissions are those created by the reporting entity itself, whereas indirect emissions are embodied in products as raw materials are transformed along the supply chain25. In this respect, the GHG Corporate Protocol26 provides a standardized definition for direct and indirect emissions in the production process. It classifies emissions as follows: Scope 1 emissions are direct emissions from owned or controlled sources, while Scope 2 and 3 emissions are indirect emissions. Scope 2 are emissions from the generation of acquired and consumed electricity, steam, heat, or cooling27. Scope 3 emissions include all other emissions along the supply chain, both upstream emissions related to the purchased goods and services, and downstream emissions related to sold goods and services (ref. 28, pp. 31–32). In this paper, we focus on upstream Scope 3 emissions consisting of emissions from purchased goods and services, capital goods, transportation of goods and services, waste generated in operations, business travel, employees commuting, and leased assets – see Fig. 1.

Fig. 1
figure 1

Scope 1, 2 and 3 emissions (Source: ref. 28, Fig. 1.1, p. 5).

Under the corporate standard, firms are required to quantify and report Scope 1 and 2 emissions. However, to date, the reporting of Scope 3 emissions is recommended, but optional22,29. This is especially problematic because according to Matthews et al.30, Scope 3 emissions represent, on average, 84% of a firm’s total carbon emission footprint (whereas the remaining 16% are Scope 1 and 2). Moreover, in nearly two-thirds of all economic sectors, Scope 3 emissions represent more than 75% of their carbon emission footprint30. Similarly, Huang et al.31 find that Scope 3 emissions are responsible for 70–80% of firms’ total analyzed footprint for most manufacturing industries. Moreover, Hertwich & Wood32 find that while direct (Scope 1) emissions increased by 47% between 1995 and 2015, indirect emissions grew considerably more: Scope 2 emissions increased by 78% and Scope 3 emissions by 84%.

To assess its Scope 3 emissions, a firm must map its total value chain (i.e., cradle-to-gate) and identify all emission sources. Often, a firm does not have access to the emission data from other firms along the supply chain (especially when firms are further away in the supply chain)33. The data used to calculate Scope 3 emissions should ideally be primary data. Primary data are data on actual emissions that are directly collected by firms (e.g., meter readings, purchase records, utility bills, engineering models, direct monitoring, etc.). With digital production processes, such data are produced as by-products or at very low costs. When we refer to data sharing, we refer to sharing of such primary data. When primary data are not available or shared along the supply chain, it becomes necessary to use secondary data to calculate Scope 3 emissions. Secondary data are not directly collected by the firms and are, e.g., industry average data, financial data, proxy data, and other generic data (ref. 34, p. 21). Such data could be used with so-called “environmentally extended input output (EEIO) models”30,31,32,35,36. EEIO models attribute emissions to final demands, providing a method of accounting for the embodied emissions in different sectors and regions for different products.

The potential benefits of accurately calculating and reporting Scope 3 emissions

To overcome climate change, governments, firms, and consumers need to reduce GHG emissions. Reporting of Scope 3 emissions increases transparency as it reveals the emissions in production processes, i.e., it allows firms to identify emissions as well as the most promising intervention points. Moreover, reporting leads to (at least some) accountability. Increasing accountability could increase cooperation through the actions of “conditional cooperators”37,38,39. Additionally, reporting opens the possibility for various types of sanctioning that have been shown to overcome social dilemmas40,41 and could facilitate coordination between firms along the supply chain. However, when firms report their Scope 3 emissions, the overall emissions along the supply chain may become a liability for the firms, as they may be forced to reduce those emissions (e.g., by adopting low-pollution technologies, switching to “cleaner” suppliers, and improving material efficiency25,42). Yet, for the individual firms, emission reporting helps in identifying potential cost savings, potential and future climate change-related risk factors, and in seeking new business opportunities43.

Scope 3 reporting based on primary data (i.e., representing the actual production technologies of the individual firms instead of industry averages) is also useful for investors seeking to assess a firm’s strengths and reduce uncertainties. In fact, Matsumura et al.44 find “that the markets penalize all firms for their carbon emissions, but a further penalty is imposed on firms that do not disclose emissions information”. Moreover, if consumers are aware of the embodied emissions of competing products, they may consider this information in their purchase decisions45,46. This would put pressure on firms to reduce emissions along the supply chain that could result in switching to low-emission suppliers, provide incentives for adopting low-pollution technologies, increasing material efficiency, etc. However, if Scope 3 emissions are reported based on secondary data of industry averages, then firms do not have an incentive to improve their processes beyond the industry average.

The current situation, where only Scope 1 and 2 emissions must be reported, may lead to a re-allocation of emissions to industries with stronger lobbies (i.e., with a lower chance to be regulated), or to “carbon leakage” to regions with less strict environmental standards. Since GHG emissions matter globally, this may impair the fight against climate change. In this context, Mytton29 reports that when firms deploy information technology to the cloud, they switch their emissions from Scope 1 and 2 to Scope 3 emissions. Given the current reporting policy, firms may avoid reporting their Scope 3 emissions and thus brand themselves as “cleaner” (as they have now lower Scope 1 and 2 emissions), although they did not reduce emissions along the supply chain. Another point is that for emission trading institutions (e.g., EU ETS) to work efficiently, firms need to be aware of their actual emissions. Precise reporting (which can only be attained by sharing primary emission data along the supply chain) will also increase the efficiency of the emission permit trading systems.

Therefore, the overall consequences of not reporting Scope 3 emissions based on primary data (i.e., using secondary data that may or may not be close to the true values) may lead to a distorted climate policy32, in particular to a less-than-optimal emission mitigation. Finally, firms that share data with each other may establish partnerships and also generate new business models, more efficient supply chains, and revenue from the monetization of their data47.

Obstacles to primary data sharing

Today’s value chains are complex networks that include multiple tiers of different suppliers across the globe48,49. To (precisely) calculate Scope 3 emissions based on primary data, firms need access to data that are not under their control; all suppliers (in all tiers) need to measure and share relevant data with each other. However, some firms lack the in-house knowledge and personnel to prepare the required data for Scope 3 in a sharable way and would need to make significant up-front investments50. This is especially problematic for small and medium-sized firms with smaller operations, less personnel, and smaller budgets. And since sharing such data could also lead to a competitive disadvantage, as it might reveal confidential information on the production processes (e.g., costs, technology), firms are often hesitant to make the required investments51.

In our analysis, we do not reflect on questions of data measurement. In fact, FERF52 has estimated that the internal costs of accounting constitute between 1 (for large firms) and 3 percent (for smaller firms) of the revenues (see ref. 49, p. 949). Instead, we focus on the obstacles to sharing available emission data and review the following issues: lack of legal clarity and regulatory concerns, lack of data and action interoperability (allowing two or more IT systems to exchange data which is understood and can be re-used by the other system), and high risk in sharing sensitive data22,47,50,53,54,55.

Legal and regulatory challenges

The rapid growth of global data volumes raises multiple legal and regulatory questions, for example with regard to data ownership, access to, and re-use of data56,57,58,59. For Scope 3 emissions, currently, firms do not have access to all the data they need and face (contractual) limitations when they re-share received data as part of their own emission footprint to the next tier in the supply chain. In addition, individual countries have started to protect data for local value creation by implementing an increasing amount of data localization measures60; i.e., measures that limit data flows across borders to not lose control over the data. An example for such a data localization measure is article 27 of the Data Act by the European Commission: It requires cloud computing providers to prevent the international transfer of non-personal data where such transfer might create a conflict with Union law. (Retrieved November 8, 2022 from A review of different regulations and their impact on non-personal data sharing in general is offered in ref. 61,62. For example, data sharing between firms may infringe anti-competition law, e.g., by fixing prices50. In particular, Scope 3 emissions data could include competition-relevant information such as production capacity, truck routes, etc22. Contrary to information and knowledge, data and their specific use regarding Scope 3 calculations are not covered by existing laws. For example, the Copyright DSM Directive by the European Commission grants copyright protection to information and knowledge representing a creative expression, but not to non-creative factual data61. Therefore, firms must decide on a case-by-case basis which data can be shared for what purpose47. Furthermore, “[u]nclear messages from governments on future regulations or major changes in reporting methodology may discourage early movers from investments in a specific reporting [(and data sharing)] infrastructure and generate a wait-and-see attitude [among firms]” (ref. 43, p. 25). In particular, the regulatory body should commit to a clear strategy for future emission reporting requirements (e.g., regarding precision and timing), as without clear legal commitments firms may be hesitant to report Scope 3 emissions and rather opt to “let sleeping dogs lie”, fearing that reporting might lead to legal implication without urgency (see ref. 43, p. 29). In sum, these legal and regulatory challenges lead to high uncertainty and costs for supply-chain data sharing.


A study by the European Commission47 identifies missing interoperability as one of the main obstacles for more data sharing across firms in general. Interoperability can be defined as “a measure of the degree to which diverse systems, organizations, and/or individuals are able to work together to achieve a common goal” (ref. 63, p. 2). In particular, the lack of interoperability regarding Scope 3 emissions is twofold: First, a lack of harmonized standards to measure GHG emissions, and second, the absence of a common infrastructure to technically exchange emission data across IT systems. Existing standards (e.g., ISO standards, GHG Protocol standards) and sector guidelines (such as the Product Category Rules or Product Environmental Footprint method) leave room for interpretation22,53,64. Scope 3 emission footprints build up on top of the emission data input of different suppliers, but lack an overarching harmonization and integration33. With missing data interoperability, i.e., when suppliers within a supply chain use different calculation standards or interpret certain inputs differently, the resulting Scope 3 emission footprint is inconsistent and insufficient. But even if all firms along a supply chain agree to a certain calculation standard, they still face a lack of action interoperability, i.e., most IT systems cannot exchange data with other systems across company boundaries. In a global value chain including very small producers, a variety of IT systems are used. Current data exchange is realized with high manual efforts and surveys or spreadsheets leading to high costs22.

Data privacy

Another major obstacle for data sharing is the risk of sharing sensitive data. Product-level emission data can enable reverse engineering to conclude insights into production processes (e.g., information about product composition and supplier networks) and is therefore considered to be competitively relevant33. Firms are very hesitant to share such data. For example, a survey among producing firms revealed that 42% fear losing innovative or competitively relevant knowledge if data are shared or locked into one data platform65. Moreover, Pauer et al.66 conducted a survey with executives from large enterprises as well as small and medium-sized firms throughout Germany: 57% of the respondents stated that “[a]nxiety about core data and business secrets being exposed” is a “very big [or] rather big obstacle” for data and information sharing, 51% expressed that “[s]haring competitive information is dangerous because it might create advantages for competitors”. Furthermore, 59% of the respondents stated that it is a “very big [or] rather big obstacle” that “we cannot check who is going to read our data if they are uploaded to the platform” (ref. 66, p. 25). Those data privacy concerns sum up to a significant obstacle to sharing primary data and thus enabling other firms to calculate their emission footprint.

Overcoming the obstacles to primary data sharing

There is already a large (and constantly increasing) number of initiatives seeking to address the obstacles reported in the previous section and thus promote successful sharing of emission data along the supply chain. In this section, we discuss different initiatives and their approaches (focusing on the obstacles as categorized above rather than on individual initiatives).

Overcoming legal and regulatory challenges

Private firms have only limited possibilities to influence legal and regulatory environments. It is the task of policy makers and regulators (e.g., international bodies such as the United Nations and its subsidies, European Commission, national governments) to remove legal and regulatory barriers to data sharing. First, they could make Scope 3 emission reporting obligatory (regional attempts have been made, e.g., the Corporate Sustainability Reporting Directive of the European Union, and the Enhancement and Standardization of Climate-Related Disclosures of the U.S. Securities and Exchange Commission). Second, policy makers and regulators have the task to ensure “that information provided by [firms] is timely, reliable and relevant” (ref. 43, p. 24). The more the precision requirement is emphasized, the more important it becomes to share primary emission data (as it is the only means to provide actual numbers rather than estimates). For example, under the UK Carbon Reduction Commitment Program, the Environment Agency conducts third-party audits of 20% of the participants every year. In Japan, entities that submit false reports or fail to meet the reporting criteria can be fined (ref. 43, p. 38). In May 2022, the U.S. Securities and Exchange Commission fined the Bank of New York Mellon USD 1.5 million for misstatements in its sustainability reporting. (Retrieved November 7, 2022 from Third, when governments commit to a future date where Scope 3 emission reporting would be mandatory, they could provide incentives for first movers. Fourth, to meet the global dimension of supply chains, regulatory bodies could create regulations to enable free movement of data across borders (e.g., regulation on the free flow of non-personal data by the European Commission).

Overcoming missing interoperability

Interoperability in emission data sharing brings a twofold challenge: First, to create a harmonized standard to measure GHG emissions, and second, to create a common infrastructure to technically exchange emission data across IT systems. To address the first challenge of interoperability, standard setters (e.g., Greenhouse Gas Protocol, ISO, European Commission) strive to establish one harmonized standard on how to measure Scope 3 emissions within supply chains in a consistent way. The ISO developed and is still working on various standards to measure emission data (e.g., ISO 14067:2018 defines requirements and guideline for measuring the carbon footprint of products). Additionally, the European Commission published the Product Environmental Footprint method (PEF), quantifying the environmental impact of products, and the Product Environmental Footprint Category Rules (PEFCR), defining guidance at the level of specific product categories. Private coalitions or public-private partnerships (e.g., World Business Council for Sustainable Development (WBCSD), SME Climate Hub, CEO Alliance, We Mean Business Coalition) can use their market reach to facilitate overarching conversations and to establish a harmonized standard across industries. For example, the WBCSD has initiated the Partnership for Carbon Transparency (PACT), that brings together leading firms from a range of industries, technology players, industry-focused initiatives, standard-setting organizations, reporting bodies and regulators seeking to define an overarching standard for Scope 3 emissions. PACT launched a coordinated and open approach to calculate product carbon footprints (PCFs) along the whole value chain. The partnership published a methodology of standards for the calculation of PCFs that leverages and aligns with existing methods and standards, e.g., GHG Protocol, PEF and PEFCR, and removes the room for interpretation and inconsistency24,26,27,28,34.

To solve the issue of action interoperability, i.e., the ability of different IT systems to exchange information, PACT, together with the non-profit SINE Foundation, launched a technical infrastructure for data sharing based on open-source technology. This infrastructure defines rules for data sharing such as data formats and standards describing how to connect the different IT systems used by firms across the value chain22. PACT initiated a collaborative process in which large corporations and software providers developed technical core components of the network together. With the standards defined in the methodology, the infrastructure is open and interoperable, connecting different IT systems. Another example for an open-source infrastructure for sustainability data sharing is OS-Climate, an open-source project backed by the Linux Foundation, building technology and data platforms focusing on data availability, comparability, and reliability. The goal is to provide data about climate change impact and incorporate that data into global financial decision-making and risk management.

As a non-profit organization consisting of 122 firms from the automotive industry, the Catena-X Automotive Network e.V. created a uniform standard for data exchange along the entire value chain. It connects all members in the automotive supply chain without lock-in effects, provides a sustainable solution for the digitalization of supply chains, especially for small and medium-sized firms, and supports the cooperation and collaboration of market participants and competitors. Catena-X allows for different kinds of data sharing including emission data. Finally, the Smart Freight Center was established in 2013 as a non-profit organization and works with the global logistics community to drive transparency and industry action. To harmonize the calculation and reporting of the logistics emission footprints, the Smart Freight Center published the Global Logistics Emissions Council Framework which can be implemented by shippers, carriers and logistics service providers.

To overcome the obstacle of interoperability, the different initiatives need to remain open and collaborative to establish one harmonized standard across countries and industries for consistent data measurement and the exchange of emission data.

Overcoming data privacy concerns

Ensuring data privacy is a necessary condition for successful data sharing between firms. One promising possibility to enable data sharing while protecting the firms’ privacy are data trusts; neutral stewards that manages a firm’s or person’s data on their behalf. A data trust “works within the law to provide ethical, architectural and governance support for trustworthy data processing” (ref. 67, p. 6). A data trust has two key functions: First, providing data holders the possibility to execute their rights and to set limitations on who can do what with their data. Second, providing a technical environment in which data processing and analysis can take place while holding everyone accountable for his or her actions. Organizations such as the SINE Foundation, OS-Climate or Catena-X were established as non-profit organizations to be able to function as data trusts for the exchange of sustainability data. Besides a data governance, data trusts also need to provide technical tools such as encryption technologies to protect the data.

Academic research on so-called ‘homomorphic encryption technologies’ has now become possible in practice due to the increase in computing power68,69. Without homomorphic encryption technology, data need to be revealed to be utilized (leading to a decrease in privacy of the sharing party, i.e., the privacy-personalization paradox70). Examples for homomorphic encryption are “zero-knowledge proofs” and “secure multi-party computing”. With “zero-knowledge proofs”71, a party can prove that it meets certain criteria (e.g., proving that a product carbon footprint is below a certain threshold) without revealing any other information. “Secure multi-party computing”72 is a protocol that allows computation or analysis of combined data without the different parties revealing their private input73. A large pool of data is always a security risk and will be luring cyber attackers. One solution (used e.g., by PACT) is to allow the firms to keep their data decentralized and share it peer to peer which decreases the attractiveness for hackers. Data trusts and homomorphic encryption technologies in combination with decentralized data storage enable firms to overcome the privacy-personalization paradox and thus solve the data privacy concerns as an obstacle for emission data sharing.

Finally, whenever a firm holds confidential data or information, there are issues of cybersecurity involved. Cybersecurity refers to “the set of technologies, processes, and practices designed to protect networks, computers, programs, and data from attack, damage, or unauthorized access, in accordance with the common information security goals: the protection of confidentiality, integrity, and availability of information” (ref. 74, p. 105). The global cybersecurity market is constantly increasing. Fortune Business projected it “to grow from USD 155.83 billion in 2022 to USD 376.32 billion by 2029” (Retrieved from and private cybersecurity firms will be part of securing the data and information.


Table 1 presents a short summary of the three main obstacles to primary data sharing, and current approaches to overcome them.

Table 1 Obstacles to data sharing and current approaches.

Concluding remarks

This article underlines the benefits of and the obstacles to sharing primary data for the calculation of Scope 3 emissions along the supply chain. We started by defining Scope 3 emissions, their relevance in sustainability reporting, and why their precise calculation (beyond industry averages) requires data sharing among firms. While the ongoing digitization of production is reducing the costs of data measurement, data sharing among firms requires overcoming three main obstacles: legal and regulatory challenges, missing interoperability, and data privacy concerns. The paper reviewed these obstacles and discussed the current progress in overcoming them.

Nowadays, the production of goods is a global and complex process involving several suppliers. The number of sustainability start-ups focusing on the measurement of Scope 3 emissions is increasing every day. Many of them provide estimates based on industry averages. While such data cannot provide precise emission values, they may serve as starting points. However, only primary data sharing provides actual emission numbers and thus has the potential to accelerate the decarbonization of the economy. Primary data sharing may also provide additional benefits to individual firms and enable new business models within digitized supply chains. For example, to account for transportation emissions, a firm needs to share data on the mode of transportation (e.g., road, rail), distance covered, and load specifications22. If these data are shared in real-time, firms could use them to coordinate collective logistic optimization. For example, Procter & Gamble (P&G) and Tupperware shared their logistics data and identified significant potential for load consolidation. They used similar routes, and their truckloads were complementary and thus offered substantial potential gains from coordination: Tupperware’s trucks were filled to around 80% of the maximum volume of the vehicle but only 30% of its weight—while P&G’s trucks were at 50% of the maximum volume of the vehicle but 95% of its weight. With real-time data sharing, they coordinated a consolidation of their loads, saving 17% of the transportation costs and more than 200 Mt CO2. (Retrieved March 27, 2022, from In addition, Pedreira & Melo75 suggest a possible reduction of CO2 emissions of up to 23% if all firms sought similar arrangements as described above. Irannezhad et al.76 conducted a study at the Port of Brisbane showing an emission reduction potential of up to 40% if the logistic firms operating at the port would share their data to coordinate their activities (e.g., to reduce the number of empty trucks).

Thus, the potential benefits of data sharing among firms (e.g., logistic consolidation) are even larger than the mere benefits of calculating Scope 3 emissions. In order to utilize the sharing of primary data for the calculation of Scope 3 emissions and beyond, different stakeholders—including regulatory bodies, standard setters, climate initiatives, research institutes, and private firms—need to take a systemic and global perspective to overcome the three major obstacles for data sharing along supply chains.