Carbon Accounting Measurement with Digital Non-Financial Corporate Reporting and a Comparison to European Automotive Companies Statements

Abstract

The regulatory environment for both sustainability and financial reporting is changing as standardisation and digital reporting (e.g., XBRL) are gaining traction within regulators. The measurement methodology and mandatory information content of disclosures are yet to be decided for corporate CO₂ reporting by EU regulators and standard-setting organisations. In our study, we reviewed the sustainability reports of three leading German automotive groups by revenue for the period 2016–2020 as a case study. The research methodology was carried out with text-mining-aided content analysis to provide a collection of sustainability standards (GRI and SASB) in the evaluation of emissions reporting. As an addition to prior literature, conditions of relevance and clarity regarding published information were introduced in the evaluation process of compliance to CO₂ disclosures. Companies by reporting practice were assigned to different stages of carbon management and actual emissions were evaluated. In the conclusions, discussion of the reliability of reported sustainability information, the applicability of digital reporting is provided through regional perspectives. We found that although analytical methods are available to assess the level of corporate carbon management, their usefulness is limited if the data are not reliable. Significant progress can be expected from analyses using standardised, comparable corporate carbon data.

1. Introduction

Carbon accounting is an emerging field of business economics and covers a wide range of activities, including measurement, calculation, monitoring, reporting, and auditing [1]. Industrial companies use large amounts of energy and materials in their activities and are therefore particularly responsible for greenhouse gas emissions. On the other hand, their technological innovation potential makes them major players in the climate change challenge [2]. According to the recent study of Spreafico, the environmental impact of cars makes up the largest share of the total global impact of passenger transport, due to their ubiquity and high impact per kilometre travelled [3]. To reduce the environmental impact of the car industry, many governments have introduced significant mitigation measures. As a result, all car manufacturers have embarked on wide-ranging interventions to increase sustainability [2]. The pursuit of sustainable development leads companies to react in different ways. Some choose proactive and innovative strategies, while others opt for reactive environmental strategies. A key feature of proactive environmental strategies is the use of modern methods integrated into management systems [4]. Companies should integrate their sustainability goals across the business platform, rather than treating them as separate from their business strategy [2]. Carbon accounting issues need to be integrated into functional areas to support the achievement of corporate and climate policy objectives [1]. Based on the literature review to be presented in Section 2, a research gap was identified in carbon accounting, specifically in its qualitative measurement in non-financial corporate reporting. Theoretical grounding relies on the Resource-based View (RBV) theory, which explains corporate competitive advantage as a mix of non-substitutable resources [5]. The RBV theory itself raises the need to use resources efficiently to achieve sustainable performance [6], yet companies do not see their use of resources as the most efficient, according to business reports. Corporate annual sustainability reports discuss the efficiency of the use of these resources from several perspectives, but the qualitative analysis of these resources has been little discussed in areas of corporate reporting and environmental economics. Digitisation has not previously been utilised directly to measure sustainability qualitatively in this area; in addition to using digitization, we looked at the standardisation guidelines proposed by the current European regulatory environment.

In the current study, we connected prior models of carbon accounting with the European legal obligations of quality (standards for sustainability information content) and format (digital), highlighting cross-connectivity between the two. We conducted exploratory research through case studies, comparing theoretical studies and standard requirements with published sustainability reports of the three largest German automotive OEMs. Our research objective was to generate a validation methodology scale because the actual compliance with the required standards is hard to measure while considering many qualitative factors, therefore this contribution improves the traceability of sustainability attributes based on corporate reporting practices.

2. Literature Review

2.1. Carbon Accounting in Corporate Practice and Regulation

There is a massive body of literature discussing the conceptual framework of corporate carbon accounting [7,8]. Several guides, standards, and norms are available to help organisations disclose information about environmental performance and carbon performance [4]. The most widely used accounting tool is the Greenhouse Gas (GHG) Protocol, serving as a standard for measuring and reporting organisational level direct and indirect greenhouse gas emissions [9]. The GHG Protocol Initiative is a multi-stakeholder partnership of businesses and NGOs, governments, and others convened by the World Resources Institute (WRI), a global research non-profit organisation, and the World Business Council for Sustainable Development (WBCSD), a Geneva-based coalition of 170 international companies. The GHG Protocol Initiative launched in 1998, and the mission is to develop internationally accepted greenhouse gas accounting and reporting standards for business and to promote their broad adoption [10,11].

Although the issue of carbon accounting is widely addressed in the literature, its level of integration in corporate practice varies greatly between businesses and regions. In general, all Environmental, Social, and Governance (ESG) information is gaining significance amongst capital markets and other stakeholder groups, which leads to a financially material effect on company valuations [12]. EU directives set in 2015–2020 are putting pressure on the car industry to meet CO₂ emission targets, with potentially significant economic and financial consequences due to the fines that can be imposed for non-compliance [13]. The findings of Szász et al. suggest that even if strategy-driven sustainability practices lead to improved environmental and social performance, these improvements are not reflected in the short-term financial performance of automotive companies [14]. Besides governance and internal assessment (financial materiality or risk) of sustainability-related operations, corporate reporting is also an issue. It means a great effort is required by organisations to ensure that ESG disclosures are meeting quality standards that offer a systematic approach to presenting sustainability information [15] and to mitigate concerns regarding the presentation of non-existent sustainability-related performance [16,17]. Accounting and reporting for GHG emissions can be considered a critical factor in the valuation of automotive companies due to their environmental impacts, such as global warming potential, stratospheric ozone depletion, ecotoxicity and human toxicity, and air quality [18].

2.2. Development of Digital Reporting for Sustainability

Among other internal corporate procedures, sustainability measurement and reporting were affected by the emergence of big data solutions by new taxonomy concepts, new sensory measurement technologies, and the need for transparent disclosures and open discourse [19].

With the traditional unstructured sustainability information format, the measurement of indicators is challenging due to their inconsistent nature between reporting years and entities. Quality assessment of emission-related disclosure can be based on composite indices; however, their reliance is limited due to information loss during the aggregation step, dependencies among criteria, and scoring of qualitative indicators [20]. Due to the textual format of the information, terminology for basic sustainability criteria needs to be standardised as well. Many metrics, indicators, and criteria are thematically overlapping and formulated in different wordings in disclosures [20]. The eXtensible Business Reporting Language (XBRL), as an emerging international business reporting standard, serves as a technical framework for the integration process based on electronic tags for each item of data. Additionally, these specific XBRL tags should be embedded in an interoperable way between the financial and sustainability reports, or comprehensively in an Integrated Report (IR). The disclosed data’s design will enable sustainability Key Performance Indicators (KPIs) to be automatically retrieved and reviewed by computer software [21]. Entity-relevant facts are tied to concepts in one or more relevant taxonomies, therefore the complexity of reported information can be enhanced with the embedded narrative explanation of quantitative items within the text [22].

Efimova et al. consider XBRL as the formation of various reporting sources of financial and accounting nature (IFRS, stock exchange listing requirements, managerial accounting systems, Integrated Reporting frameworks), big data, and other disclosures (i.e., sustainability) [23]. The history of existing sustainability taxonomies and methodological questions of currently developed taxonomies were addressed by Beerbaum and Puaschunder, where disclosures of carbon emissions serve as one of the main transition risks [24].

The integration of sustainability standardisation into management accounting control needs to take a strategic focus, as previously companies mainly collected data for lagging indicators [25]. Digitally unified reporting (e.g., XBRL) provides a digital enhancement both to data management and standardised reporting as a twin-track approach to sustainability management [26]. Reporting in such a digital format enables big data that can be used as part of managerial decision-making in the form of predictive tools for reducing environmental pressure [27]. Souza and da Silva presented an integration framework adding the Global Reporting Initiative (GRI) disclosure recommendations into sustainability information reported in the XBRL format usable in analysis and audit processes [28]. As an example for technical usage, the United States Environmental Protection Agency (EPA) published a public database of 2019 Greenhouse Gas (GHG) emissions from large facilities of the U.S., requiring reporters to use its XML-based reporting platform [29].

2.3. Theoretical Models for Corporate Carbon Management

With a comprehensive approach to understanding its role in business models, Suk distinguished 5 stages of carbon management [2]. Stage 1, the passive strategy, is the basic level at which companies may recognise or become aware of the need for energy management, which is limited to energy conservation and management. Stage 2 is the reactive strategy, which involves measuring an organisation’s energy and carbon footprint and implementing an emissions reduction strategy based on international standards. Stage 3, the defensive strategy, is the stage of the process where companies assess their internal options to reduce GHG emissions and respond to government regulations by setting up units or departments. Stage 4 is the adaptative strategy, in which companies seek ways to optimise their carbon emissions potential. They are implementing more innovative carbon management, e.g., by publishing carbon emissions reports. The rationale for corporate carbon reporting can be financial, as well as to differentiate from competitors and gain recognition. Strategies at this stage are more concerned with stakeholders, investors, and consumers. Stage 5 is the proactive strategy, characterised by optimising the trade-offs between costs, time, and carbon emissions, based on accumulated information and solutions, integrating carbon and financial data to deliver financially optimised sustainable business improvements. This phase also involves companies identifying opportunities with the highest return on investment through strategy development and implementation that goes beyond regulations.

According to Ibrahim et al., the number of publications on CO₂ management decision making has increased especially in the last decade (2010–2020), with a significant increase in 2017 [30], especially in the automotive sector during the introduction of the Worldwide harmonised Light-duty Test Procedure (WLTP) [31], that consists of several customised models, tools, methods, and frameworks that have been developed to support CO₂ management decisions. Similarly, Multi-Criteria Decision-Making (MCDM) methodologies can be used by automotive OEMs to reconcile interrelated decision criteria to make Pareto-optimal decisions [30,32].

Wahyuni and Ratnatunga studied that companies, regardless of the industry type, make choices on allocating scarce resources in their internal corporate carbon management practices [33]. To provide theoretical grounding, the Resource-based View (RBV) was adapted that by prior literature understands companies as first, a bundle of productive resources, and second, an administrative framework that coordinates resource deployment. Their results suggest that traditional organisations, which see business and carbon economics as separate domains, adapt passively, whereas proactive firms are characterised by a holistic approach. They effectively integrate carbon-aware thinking into their business practices for higher and greener profits. The theory of RBV was applied in more studies sustainability-related studies to determine new factors influencing competitive advantage. Zhang et al. analysed the transformation of unique internal and external resources in electric battery supply chain processes [34]. The management of resources was mainly discussed from the perspective of green human resource management practices, and hence was connected to the Social and Governance pillars of ESG reporting [35,36,37]. Rotjanakorn et al. warned however that the theory focuses on the management of resources that are available under normal conditions in the automotive industry [38], therefore dynamic technological development might trigger unexpected resource management.

RBV theory’s application in carbon management research was also contributed to by Shibin et al. with the inclusion of legislated standards for carbon emission, and threats associated with environmentally irresponsible behavior as coercive pressures on positive environmental performance [39]. Methodological research of internal resource measurement was extended by environmental constraints (pollution prevention, product stewardship, and clean technology) by De Stefano et al. to support CO₂ emissions measurement [40]. At the same time, recent findings of Bocken and Short address the issues concerning unsustainable business models of the transportation industries [41]. Business models of automotive companies build on the necessity of fossil-fuel-powered transportation, profit from global value chains, and make use of technological investments without correctly pricing environmental impact and stimulation of further fossil-fuel use. Micek et al. analysed the market opportunities and regional competitiveness of automotive firms [42] assuming the aim of German development of Central European automotive manufacturers being to mostly acquire better political and organisational capabilities, which implies the lack of integration of sustainability in business models regarding regional development. In summary, based on the underlying principle of the RBV theory concerning the allocation of resources, carbon accounting, including the release of regular reports on emission data [2] among other factors influence corporate competitiveness. Based on the literature, the RBV theory can be used to distinguish between companies in terms of their level of carbon management.

3. Materials and Methods

For this study, we analysed the biggest three automotive Original Equipment Manufacturers (OEMs) by 2018 net sales revenues [43]. The three selected company groups are the Bayerische Motoren Werke (BMW) AG, Daimler AG, and Volkswagen AG, located in Germany and operating in the passenger cars and other motor vehicle markets. In the European automotive sector, Germany holds a significant leading position, with 579.9 EUR billion of revenues and 46.3% of the total European automotive manufacturing revenues. In 2018, Germany also accounted for 32.2% of the total value of EU exports in automotive products [43]. In

Table 1. Consolidated financial and economic data of sample companies, 2020.

2020 Data	BMW	Daimler	Volkswagen
Vehicles produced (million)	2.256	2.795	8.900
Net sales revenues (M €)	98,990	154,309	222,884
EBIT (M €)	5222	6603	11,667
Total assets (M €)	216,658	285,737	497,114
Shareholders’ equity (M €)	60,891	60,691	127,049
Net cash flow (M €)	1501	4165	9103
Number of employees	122,874	99,640	665,445
Countries of operations	15	16	30
Number of vehicle manufacturing units/total units	16/31	8/28	50/118

Includes the 79,000 employees of Chinese joint ventures. Source: official corporate reports and websites of companies.

, the most important financial and economic data were listed of the selected three company groups. Due to their high production volume, the selected OEMs have a major impact on the sustainability of automotive manufacturing processes on a global level, along with their extensive supply chains.

A qualitative content analysis based on the Global Reporting Initiative (GRI) and Sustainability Accounting Standards Board (SASB) standards was carried out on the reported sustainability information of the companies on a 5 year (2016–2020) scale with the sources presented in

Table 2. Sustainability standards and disclosure used in the content analysis.

Referenced Document	Years	Compliance to Standards (2020)
BMW Sustainable Value Report BMW Integrated Group Report [44]	2016–2019 2020	GRI, TCFD
Daimler Sustainability Report [45]	2016–2020	GRI, SASB, TCFD
Volkswagen Group Sustainability Report [46]	2016–2020	GRI, UNGC, SDG, GSC

. Sustainability reports were primarily used, supplemented with additional information and datasets available on official company websites. In general, the data held an unstructured textual format incorporating external links to online and downloadable data. Based on the samples, the efforts towards compliance to multiple standard systems resulted in an overwhelming stream of information mixed between management, operational, and sustainability facts, decreasing comparability. Notably, ongoing standard-setting discussions and regulatory activity are aiming to solve the current issue with the use of XBRL.

Based on the results of Wahyuni and Ratnatunga [33] regarding contextual factors of carbon strategy adoption by the use of the contingent RBV theory, information disclosed in the sample of reports serves as an official presentation of resource allocation towards different carbon management activities. We investigated whether the automotive companies under study are at the same or different levels of adoption of a corporate carbon strategy based on the proposed framework of Suk [2]. The following questions were examined: Did the automotive companies studied reach at least Stage 4 during the observed period? What motivations led them to enter Stage 4? As the analysed automotive companies carry out their production processes in many countries, we reviewed them through the prior assumption that regional differences in emissions exist and are reflected in sustainability reporting.

As an additional factor, we added relevance and clarity of information parameters to the corporate carbon management model. Based on our hypothesis, the higher stage of the carbon management model presents more relevant and clear information. In other words, the relevance and clarity of information are in direct proportion to the carbon management stages. Based on the findings of Appendix A, we developed a checklist for the testing of the conditions of relevance (R) and clarity (C) of sustainability information. Each of the disclosures was reviewed following the reported references to standards to either the main text or external links. Condition R was considered fulfilled (marked with “X”) when disclosure (main text or intra-document/external reference) was linked to the relevant sustainability (GRI or SASB) standard. Notably, due to the change in standard identifier codes in 2016, some companies used the older version in that year. Condition C was considered fulfilled (marked with “X”) when disclosure contained the clearly definable numeric or textual value (that provides comparability and a level of standardisation according to the relevant standards) and unit of measure. Notably, the condition was not fulfilled, with percentage values of yearly changes displayed without the absolute base value of the comparison. The applied conditions ensure the representation of reality in terms of the quality of reported information. In the analysis of both Conditions R and C, the related textual data were classified in either category of containing (1) verifiable and measurable data, (2) not verifiable data but fitting reporting practice, or (3) data that is not validated nor fitting reporting practice.

As part of the study, we applied text-mining aided methods to attempt to standardise textual data of sustainability standards and reports. A combination of manual content analysis and Natural Language Processing (NLP) tasks were carried out, of which the latter used the Provalis WordStat 8.0.33 software [47]. As a methodological guideline, the prior literature of [48,49,50] was used in the topic of text-mining-based evaluation of sustainability reports. The content analysis steps included (1) textual data retrieval from sustainability reports and supplementary material (additional reference tables, standards) (2) processing of textual data of standards to identify search criteria in reports, (3) quantitative content analysis based on the defined standard categorisation, (4) observation of sustainability disclosure occurrence in main report body text, (5) comparison between years and companies following research questions that were previously discussed.

4. Results

We presented the results in three subtopics. Firstly, we aimed to develop a standard system for the evaluation of Carbon Accounting using corporate sustainability reports. Secondly, we carried out a content analysis on the sample of 3 companies and 5 years of sustainability reports focusing on carbon disclosures and presented the additional parameters of relevance and clarity of data. Thirdly, we reviewed the regional factor of the reported information.

4.1. CO₂ and Financial Disclosures’ Compliance to Legal Obligations and Standard (GRI, SASB) Directives

In

Table 3. Text mining keywords for Carbon Accounting in sustainability reports.

Definition	Description	Topic Keywords
CO2 emission	Emission of CO2 from the combustion or biodegradation of biomass	carbon, carbon accounting, CO2, carbon dioxide, emission
CO2 equivalent	A measure used to compare the emissions from various types of greenhouse gas (GHG) based on their global warming potential (GWP) or relative ozone depletion potential (ODP)	* equivalent, CFC, GWP, ODP
Greenhouse gases	Gas that contributes to the greenhouse effect by absorbing infrared radiation	greenhouse gas, GHG, GHG *, * emission
direct (Scope 1) GHG emissions: GHG emissions from sources that are owned or controlled by an organisation
energy indirect (Scope 2) GHG emissions: GHG emissions that result from the generation of purchased or acquired electricity, heating, cooling, and steam consumed by an organisation
other indirect (Scope 3) GHG emissions: Indirect GHG emissions not included in energy indirect (Scope 2) GHG emissions that occur outside of the organisation, including both upstream and downstream emissions
GHG trade	purchase, sale, or transfer of GHG emission offsets or allowances	GHG * trade
Global Warming Potential	value describing the radiative forcing impact of one unit of a given GHG relative to one unit of CO2 over a given period	global warming potential, GWP
Significant air emission	air emission regulated under international conventions and/or national laws or regulations	significant air emission

Note: asterisks indicate arbitrary characters. Source: GRI standards package (2016–2018).

, keywords connected to Carbon Accounting are presented based on the GRI Standards Glossary 2020 and SASB Automotive (identified as TR-AU) disclosures. In the next step, the frequencies of the words and phrases were reviewed in every GRI standard individually to determine the relevant standards and disclosures discussing the emission of CO₂ and other GHG gases, and propose methodologies of measurement.

We reviewed all standards proposed by the GRI and SASB focusing on their information content on CO₂ emission for automotive companies. As opposed to GRI standards, SASB provides a package specifically for the automotive industry. All textual data of individual standards were reviewed with text-mining aided frequency analysis using snowball-stemming as the text preprocessing methodology. Those standards indicated by the high frequencies of keywords mentioned in Table 3 underwent a final manual selection process. The standards containing valuable guidelines for reporting disclosures for CO₂ emissions are listed in

Table 4. Selected non-financial reporting standards relevant in reporting CO₂ emissions.

Standard Name	Carbon	CO2	GHG	GWP	Emission
GRI 305: Emissions 2016	5	34	150	26	11
GRI 201: Economic Performance 2016	4	2	2	0	0
SASB TR-AU: Automobiles	3	2	1	0	4
GRI 302: Energy 2016	0	0	6	0	4
GRI 307: Environmental Compliance 2016	0	0	0	0	0

, including some indicative examples for keywords frequencies. Notably, not all standards included keywords directly; GRI 307 however was selected due to the growing significance of legally required compliance to standards.

Each organisation is responsible for the quality of the information published in corporate reports. The information content of annual reports and financial statements are often limited to the most necessary supplementary information due to the obligation to pass financial audit processes, but this will only be the case for sustainability statements after the future EU legislation is implemented [51]. For instance, the GRI 201-2 disclosure (Financial implications and other risks and opportunities due to climate change) as an ideal set of indicators proving traceability to financial results was under-represented in all 2020 sustainability reports and supplementary GRI index tables. In

Table 5. Financial traceability by GRI 201-2 disclosures on CO₂-related risks, 2020.

	Provided Information Content	Reference (SR)
BMW	n/a	n/a
Daimler	Assessment of the probability of occurrence (probability threshold in %) and possible impact (financial threshold in €) of Low, Medium, and High levels of risks	External link
Volkswagen	Qualitative description of the Risk management system and internal control system, Sustainability and compliance risks, and Nonfinancial risks 2021 project costs targeted on CO2 reduction: €13 million	pp. 39–40, 47–53

Source: Sustainability Reports (SRs), 2020.

, information content is presented where reference to standards was unavailable, while some qualitative details were available in the statements of Daimler and Volkswagen. However, the actual financial impacts in the reporting period were not represented in any text parts or supplementary external material.

4.2. Emission Measurement in Corporate Reporting

After reviewing the carbon emissions model of Suk, an adaptation to the reporting scheme was developed to compare compliance or deviation from the model [2]. Upon the manual review of the corporate sustainability reviews information by the proposed indicators of Carbon Management Activities (CMAs) as presented in Appendix B, it can be implied that the observed companies are at stages 4 and 5 of carbon management. Their disclosed business models reach a stage 4 (accommodative) strategy, as stand-alone or integrated carbon reports were published. The time of adaptation precedes the currently relevant regulatory context [52], i.e., the first carbon report publication date (that should not be taken into account) is before 2016. Companies may disclose the information either because it is financially significant (so they have a legal obligation to disclose it) or because it differentiates them from competitors in a positive way for stakeholders. The companies with the same geographical scope of operation (i.e., operating in Germany) started disclosing certain information before the analysed period, which suggests a lack of external pressure. However, certain disclosures vary in information content, as presented in the sections below.

The highest stage strategy (stage 5) is proactive, where compliance with a public expectation or standard not being the reason for disclosure. Another important element is that the company integrates carbon emissions data into its financial reporting, i.e., creates a link between financial and non-financial data. At this stage, it can also be observed that the company is adapting to the internal market, which it can exploit to gain a competitive advantage. The fulfillment of requirements of the strategic stage is not possible with current reporting systems, as there are no common measurements of such internal processes. To gain more insight into the nature of emissions, the links between vehicle production and different scopes of CO₂ emissions were reviewed. In Figure 1 the negative impact of the ongoing pandemic on production volumes can be seen, along with the changes in the direct emission values.

From 2016 to 2018, steady growth was reported in production volumes with an average 1.35% drawback in 2019. In 2020, Q4 productions were in many cases disrupted or held back which resulted in an average 16.59% decrease in vehicles produced for the year. In contrast, Scope 1 emissions directly controlled by companies stagnated for BMW and moved conically for Daimler during the years, but showed a 15.12% increase for Volkswagen. As the information on the emissions by scope per vehicle was reported in “custom” formats—therefore low-quality conditions—we included the calculations of

Table 6. Reported emissions by scope, per vehicle produced, tons of CO₂ equivalents.

		2016	2017	2018	2019	2020
BMW	Scope 1	0.238	0.249	0.229	0.250	0.285
	Scope 2	0.368	0.204	0.212	0.118	0.037
	Scope 3	29.405	28.620	28.759	29.267	28.861
Daimler	Scope 1	0.347	0.359	0.367	0.371	0.367
	Scope 2	0.619	0.531	0.497	0.382	0.370
	Scope 3	23.509	23.725	23.360	35.173	36.605
Volkswagen	Scope 1	0.407	0.368	0.355	0.348	0.488
	Scope 2	0.507	0.463	0.390	0.351	0.315
	Scope 3	31.563	31.282	38.803	41.340	41.453

Note: Emission values were calculated using reported data on total vehicles produced (annual statements) and CO₂ emissions by type (sustainability reports).

to review lifecycle emissions of produced vehicles as well.

It can be seen that while Scope 2 emissions have seen a significant reduction (an average of 20.42% per year), Scope 1 and Scope 3 emissions per vehicle varied in trends. According to the definition of the GRI 305: Emissions standard, other indirect emissions are a consequence of an organisation’s activities but occur from sources not owned or controlled by the organisation [53]. Both upstream (e.g., value-chain operations, business travel, or waste generated in operations) and downstream (e.g., distribution, use of products sold, or end-of-like treatment) emissions are included by the indicator.

A total of 20 disclosures were examined from the checklist, of which a summary table was created consolidating CO₂ disclosures by companies and years, presented in

Table 7. Compliance of CO₂ disclosures with the relevance and clarity conditions in %, 2016–2020.

	Relevance (%)					Clarity (%)
	2016	2017	2018	2019	2020	2016	2017	2018	2019	2020
BMW	65	65	65	65	70	65	65	65	65	65
Daimler	55	55	60	75	70	45	45	45	55	55
Volkswagen	60	70	70	50	50	40	45	45	40	40

. Typically an advancement in compliance with the conditions was seen other than in the case of Volkswagen, where earlier years (2017–2018) provided compliance with a higher number of standards (GRI, SASB). This trend of decreasing relevance and clarity over the years of Volkswagen regarding the CO₂ emission in sustainability reports might be related and can be a logical consequence to the fact of the legal proceedings happening in these years. The most accurately reported disclosures were the GRI 305-1/2/3 emission standards fulfilling both relevance and clarity criteria.

4.3. Reporting on a Regional Scale: Do Companies Deliver Regional Distinction in Emission Disclosures?

As the GRI 305: Emissions standard recommends: where it aids transparency or comparability over time, provide a breakdown of the GHG (Scope 1, 2 and 3) emissions by [53]:

• business unit or facility;
• country;
• type of source (stationary combustion, process, fugitive);
• type of activity.

Automotive OEMs typically operate with a wide range of suppliers ranging from Tier 1–3. Germany as the European centre of the automotive industry, and especially the analysed three biggest OEMs by realised sales revenues, can be considered as a centre of supplier chains involving various manufacturing processes before the finished product. However, regional differentiation in emission data was less commonly discussed. As an example, Volkswagen provided differentiation in the number of vehicles produced in the last 5 years between Germany and other countries (in an aggregated format); however, this was provided separately from the sustainability data in the annual report. All of the observed years’ production complete abroad was above 87%, therefore such a differentiation would be also recommended to comply with future sustainability regulations and the transition towards a transparent carbon management system.

The introductory study of Thun and Müller reviewed German automotive manufacturers from the viewpoint of green supply chains [54]. Their findings revealed that at the time, OEMs were aware of environmental issues but were moving slowly towards realise the full potential of green supply chains, with the goal mainly of complying with legal regulations. Green supply chains were not viewed as increasing a comparative advantage. The more common ecological programs are mostly internally focused on corporate strategic planning, not covering environmental effects in different regions of operations. According to Böttcher and Müller, for German automotive suppliers, firm size played a significant role in the perceived importance of CO₂ emissions reduction [55], based on which the analysed companies are in a favourable position.

5. Conclusions and Discussion

The current study serves as a contribution to the literature review on the non-financial reporting disclosures of automotive companies, especially on the topic of CO₂ emissions. Prior work has focused mainly on the industrial side of the topic in the form of questionnaires on the perceived importance of sustainability disclosures from the corporate side. Our study attempted to cover the area from the viewpoint of reviewing sustainability reports and connected material that acts as legal obligation for the analysed German automotive manufacturers. We provided an assessment of currently available sustainability standards that play a role in the corporate reporting of CO₂ emissions, along with their proposals for measurement methodologies. In contrast to that, we also carried out a content analysis on the sample’s reports in the last 5 year period.

Our research objective was to generate a validation methodology scale of carbon accounting information disclosed in corporate sustainability reports. To this end, based on the results presented in Section 4.2, Suk’s (2018) methodology does not show significant differences between the German car manufacturers studied. Problems in recording sustainability information are not apparent based on the methodology for internal management processes, showing that despite minor differences, the companies are at the same sustainability stage. Almost all of the activities required by Stage 4 are fulfilled and reported, but most of the requirements of Stage 5 are not presented. Disclosing textual information that differs in quality and clarity is, in the RBV theory’s view, a way of communicating the provision of resources and its role in internal carbon management that includes its CO₂ emissions performance. For these high-level CO₂ abatement activities, companies do not allocate resources based on the descriptions in their annual reports.

In addition to the quality and clarity of the sustainability information reported, there is a relatively clear difference in the absolute Scope 1–2–3 CO₂ emissions between the companies analysed. While Scope 1–2 is a well-measured and manageable category due to its direct and indirect link to the manufacturing process, Scope 3 is outside the scope of manufacturing activities. The communication of information on resources is the weakest in this case, and although the criterion for reporting figures is met, there is no improvement in quality between years. Based on the conclusions, we provide a discussion of three topics as follows.

5.1. Reliability of Sustainability Information

External and intra-document references were not consistent in all cases and could not be followed from the viewpoints of relevance to international sustainability standards, and clarity in proposed information content. As such, multiple pages of text were assigned to individual standards, as well as in reverse, i.e., multiple standards were assigned to a multiple page text without assigning a specific “fact” value to them. This can be explained by internal sustainability assessments of German automakers mainly being used for reporting, instead of retrospectively evaluating development, and using findings to improve marketed products [56]. Reported units of KPIs are rarely consistent within an industry, especially for metrics that are harder to quantify [57]. The possibility of customised emission indicators (changes between direct/indirect and Scope 1/2/3 emission approaches) decreases comparability. Gallego–Álvarez identified certain environmental indicators that are included in the GRI 305 (Emissions) standard and compared their presence in the analysed sample [58]. Certain sectors such as raw materials pay more attention overall to the GRI 300 environmental indicators compared to other industries. In addition, Ordonez–Ponce and Khare attempted to harmonise GRI 300 disclosures and Sustainable Development Goals (SDGs) established by the UN, with findings of Germany being among the best performers with a consensus value of 90% of total emission standards reported [59]. In our study, the most commonly referenced standards and disclosures were highlighted in the carbon reporting of the biggest European automotive companies, with the addition of the relevance and clarity factors as part of the evaluation concept. Both factors can be strengthened by the steady development towards XBRL format reporting that supports the removal of replicated or backup accounting data—both financial and non-financial—to build a unified system of collecting and processing information [60]. We propose the further research area of integrating the five stage model of carbon management activities in text-mining based evaluation. Based on the methodological experience gained, disclosures on carbon accounting could be analysed on a larger scale, involving the entire population of automotive manufacturers on a certain regional level. Such research would provide essential contribution in the form to the development of the ongoing regulatory process of sustainability reporting in the EU.

5.2. Role of Integrated Reporting and Digitalalisation of Reports

Integrated reporting is gaining traction as BMW has published its first integrated annual report in 2020, containing both financial and non-financial information. Considering the expected arrival of digitally standardised reporting (XBRL, or ESEF as called by the European Commission) this implies corporate opening to clearly defined linkages between the two kinds of information with the double materiality concept [51]. Currently, there is an ongoing regulatory discussion regarding draft standards by the responsible organisation of the European Union, the European Financial Reporting Advisory Group (EFRAG). In April 2021, the technological recommendations were published along with a timetable for the development of current sustainability reporting standards [61]. The regulation process aims to review and develop the current European Non-Financial Reporting Directive (NFRD) active since 2018 [52]. In Appendix A, a detailed description was provided of the relevant sustainability standards and disclosures to be used and adopted by future legislation. The result serves as the basis for further evaluation of corporate sustainability information and can be used as a validator for similar content analyses carried out in the current study.

5.3. Regional Assessment of Emissions in Reporting

The regional questions of emissions reporting are generally not addressed in scientific literature; however, some proposals are given that can be reflected in regionalisation. Mitigation of emissions could be carried out by evaluating which improvements to car components can most reduce their environmental impacts [3]. The issue relies on internal processes; however, manufacturing of car parts is carried out in different specialised manufacturing units globally. Carbon accounting, as part of corporate reporting, is capable of recognising aggregate emission of manufacturing processes by regions, therefore the mitigation of certain environmentally overloaded regions is possible, but its economic benefits are questionable and case-dependent.

The conceptual framework of Böttcher and Müller categorises the drivers of low-carbon operations to (1) low carbon products, (2) low carbon production, and (3) low carbon logistics [55]. These categories are available to link existing internal managerial accounting systems to sustainability costs from a regional perspective. Additionally, XBRL has the potential to embed land cover data, e.g., GIS data on ecosystem extent and condition to offer a feasible solution to regional linkages [62].

As a future research area, we aim to extend the methodology of the current study to a wider international sample of automotive OEMs and to develop the hereby presented relevance and clarity conditions of sustainability disclosures to a multivariate model. With an extended data source, differences in carbon accounting practices and their reporting could be detected on the level of countries or continents.