Revisiting Corporate Digital Responsibility in the Digital Era: Lessons from Management and Accounting Theory

1. INTRODUCTION

Algorithm screening of job seekers’ resumes, determination of credit prices, and management of insurance claims; Distributed ledgers track the location of global goods, and the data centers supporting these systems have consumed the energy equivalent to a small city. Digital technology has moved from the edge of business operations to the center. Therefore, a question that combines management and ethical significance arises: in the context of the increasing emphasis on computing to create value in the economy, are organizational leadership, monitoring, and control systems still applicable?

In practice, the gap between technological feasibility and ethical application is constantly widening. When retailers use customer data or manufacturers’ automated carbon emission reports, the responsibility for failure is dispersed and it is often difficult to determine which department is responsible for misclassification, data breaches, or biased recommendations. Management scholars have long been exploring who businesses truly serve (Donaldson & Preston, 1995); The accounting field has surpassed the historical cost category, but the value of data assets has not been reliably evaluated; Although ESG reports have become mainstream, their credibility remains questionable (Berg et al., 2022; Christensen et al., 2022). A system should be established to connect digital functions with accountability mechanisms for daily work, rather than just publishing sustainability reports.

There have been three types of literature exploring this issue, but each study has left some unresolved questions. Firstly, empirical studies linking digitalization with ESG performance have shown a positive correlation (Zhang & Li, 2025; Zhao et al., 2024; Wang & Esperança, 2023), but digitalization is often seen as a capability that can automatically bring sustainability benefits. However, these studies have not explained why the same ability can lead to algorithm bias, privacy breaches, and increased energy consumption, and have almost remained silent on the harm that only occurs in the digital field. Secondly, despite the emerging literature on Carbon Data Disclosure (CDD) (Lobschat et al., 2021; Mihale-Wilson et al., 2022; Mueller, 2022; Jones & Comfort, 2022; Trier et al., 2023; Wirtz et al., 2023) proposed principles of privacy, fairness, inclusivity, and accountability, but relatively little attention was paid to the organizational mechanisms required to implement these principles, such as budgets, indicators, controls, and safeguards. Finally, although mature sustainable accounting tools have been developed, such as the Sustainable Balanced Scorecard (SBSC), carbon accounting, and safeguard practices, these tools are rarely associated with CDR. Therefore, the specific research gap lies in the lack of normative rules and standardized operational processes at the organizational level.

This article explores the following two issues. RQ1: How to integrate CDR, ESG goals, and green innovation into a single governance system for digital transformation? RQ2: How will management accounting be applied to this system?

Therefore, this article draws the following four conclusions. Firstly, to our knowledge, this is the first framework that clarifies the governance hierarchy between carbon reduction as a digital capability and ESG outcomes, with green innovation as the conversion mechanism and management accounting as the operational infrastructure; This positioning is the main innovation of this article. Secondly, the framework combines the perspectives of stakeholders, agency, and management control, and clarifies the division of labor to explain why each perspective is indispensable and insufficient to be used alone. Thirdly, this article derives three propositions (P1-P3), consciously distinguishing between the content supported by existing evidence and the speculative part, thereby providing a verifiable research agenda. Fourthly, this article mapped the CDR field to the ESG pillar at the tool level, corresponding it to specific management accounting tools, providing practical and feasible entry points for managers and accountants.

The subsequent structure of the article is arranged as follows: the second part provides a theoretical basis, explaining the differences and complementary relationships between CDR, CSR, and ESG. The third part presents empirical results, identifies unexplainable issues, and proposes a comprehensive framework and related propositions. The fourth part introduces the management accounting tools provided by the system. The fifth part links the literature review with the research findings of this article. The sixth part lists the contributions, applications, shortcomings, and future research directions of this study.

2. THEORETICAL FOUNDATIONS

2.1. Stakeholder Theory Identifies the Concerned Parties

(Freeman’s, 2010) stakeholder theory suggests that managers should consider all groups influenced or capable of influencing the organization, such as employees, customers, suppliers, communities, shareholders, etc., rather than just focusing on shareholders. Although this theory is sometimes considered overly idealistic, neglecting these stakeholders can result in practical harm such as talent loss, brand damage, litigation, and regulatory tightening. Digitization not only increases the quantity of these costs, but also expands their scope. Data leakage will harm customers who expose their data and investors who suffer losses; Screening algorithms with systemic bias can expose businesses to legal liability and damage their reputation among candidates and the public; Data centers bring climate costs to those who will not engage in business with the company in the future. With the continuous development of enterprise digital systems, more and more people are starting to participate. (Mueller, 2022) further proposes that in the digital age, as decision-making power is increasingly delegated to algorithmic systems, the stakeholder category may need to include both human and technological entities. Regardless of how this viewpoint is extended, its fundamental principle remains consistent: responsibility begins with determining who will bear the consequences of digital behavior. Therefore, stakeholder theory answers the first question in this framework which should be considered, but this alone cannot provide a specific action path to implement this conclusion.

2.2. Principal-Agent Theory Explains the Demand for Credible Information

The first issue is information asymmetry: the principal hires an agent to manage the company, and the agent has a much higher level of understanding of the company’s daily operations than the principal (Jensen & Meckling, 1976). Accountants have started to address this issue to some extent. Digital operations have changed the nature of the problem, but have not completely solved it. They generate forward-looking non-financial data in almost real-time, such as carbon emissions, supply chain risks, customer sentiment, and algorithmic bias indicators; Although these data are often more decision relevant than last year’s profits, their standardization is lower and therefore more susceptible to manipulation. People often say that the measured content will be managed; Otherwise, the measurement would be meaningless. Recent research has confirmed this. (Zhang & Li, 2025) conducted research on Chinese listed industrial enterprises and found that digitization enhances ESG performance by promoting green innovation, transparency, and internal control; (Zhao et al., 2024) tracked similar broad effects through large-scale cross industry panel data, namely improved operational efficiency, reduced agency costs, and increased innovation. Both sets of data indicate that assuming the data obtained by these tools is accurate, digital tools can reduce information asymmetry between managers and external personnel. Therefore, agency theory suggests that the foundation of this system lies in precise measurement of responsible digitization. The blind spot is that the relationship between the principal and the agent does not cover all affected parties.

2.3. Management Control Theory Explains the Organisation of Behaviour

Management accounting is the application of strategy in practice. Its measurement and reward are not neutral; If performance indicators only show short-term financial returns, then investments that cannot immediately recoup costs including many sustainable development and responsibility initiatives will be reduced. The sustainable development path in this field provides a direct tool to address bias. Environmental management accounting tools help businesses reduce costs and minimize environmental damage (Schaltegger & Burritt, 2010), such as carbon footprint accounting, lifecycle cost analysis, and material flow analysis; SBSC (Sustainability and Business Strategy) coordinates environmental and social goals with strategic maps, making sustainability a part of the definition of success rather than an attachment to it (Figge et al., 2002; Hansen & Schaltegger, 2016). Control theory answers the third question in this framework – how to align behavior with the goals of digital responsibility, but it cannot clarify what these goals should be. These contents must be provided from external sources outside the control system.

2.4. Combining Lenses Through Optimised Division of Labour

These three theories are usually mentioned separately in previous research; Here, we will combine them to answer three questions that any feasible digital responsibility system must address: whose interests constitute responsibility? Stakeholder theory. Why does responsibility require specific data? Principal agent theory. How is responsibility reflected in daily life? Management Control Theory. Each of these theories has its own shortcomings when viewed individually. If there is no control mechanism, stakeholder theory will only remain empty talk; If the moral circle is not expanded in the agency theory, only the parties bound by the contract need to bear responsibility; And without normative guidance for control theory, its goals may not be achieved, as evidenced by financial crises and common cases of “green bleaching”. The combination of the three also reveals the practical logic of digital liability issues: damages are borne by stakeholders, hidden due to information asymmetry, and can be prevented through control design. This diversified theoretical foundation meets the demand of the sustainable accounting community for theoretical triangulation of single perspective explanations.

2.5. The Difference Among CDR, CSR and ESG

The second conceptual task is to clarify what content CDR adds to the vocabulary system that already includes CSR and ESG, as these three concepts are often mixed up but should not be confused. CSR is an earlier and broader value-oriented category that encompasses voluntary responsibilities of businesses in the economic, legal, ethical, and philanthropic fields (Carroll, 1991). ESG is an investment risk assessment system that has developed standardized indicators to measure a company’s environmental friendliness, social responsibility, and governance level. CDR is the youngest of these three concepts (Lobschat et al., 2021): it refers to the values and norms shared by organizations in creating, operating, evaluating, and adjusting digital technologies and data (Mueller, 2022; Jones & Comfort, 2022). Table 1 shows its specific content.

The three contents of Table 1 are explained as follows. Firstly, these three have synergy; CDR can be seen as a manifestation of the specific obligations of CSR arising from dataization and algorithmic decision-making in the digital age (Jones and Comfort, 2022). Secondly, CDR has not filled the gap left by ESG indicators; That is to say, algorithmic accountability and data ethics have not been included in the current rating system, but they belong to governance (G) issues; Digital inclusion is a social (S) issue; The footprint of digital infrastructure is an environmental (E) issue, but it is not fully reflected in most ESG data. Thirdly, this dependence is bidirectional; That is to say, CDR itself does not provide the basic conditions for generating ESG indicators, but rather the ESG measurement mechanism and its guarantee ecosystem make CDR visible to the outside world. Section 3 expands on the above content as a foundation.

3. FROM EMPIRICAL EVIDENCE TO AN INTEGRATED FRAMEWORK

3.1. What the Evidence Shows Now

As this paper belongs to theoretical research, it is necessary to clarify which connections are supported by data and which are not. At present, three research results have received relatively sufficient literature support, although most of these studies focus on Chinese listed companies. Firstly, digitalization is positively correlated with ESG performance (Zhang & Li, 2025; Zhao et al., 2024), and has been applied to small and medium-sized enterprises (Wang & Esperança, 2023). Secondly, digitalization is increasing the quantity and quality of green innovation (Wang & Zhong, 2024). Thirdly, green innovation has to some extent mediated the impact of digitalization on ESG outcomes (Zhang & Li, 2025). However, two additional points need to be noted: firstly, the measurement method is relatively fragile (digital strength is usually represented by text mining indicators, while ESG performance depends on third-party ratings provided by different suppliers, Berg et al., 2022); Secondly, the identification of causal relationships has not yet been fully realized.

3.2. What The Evidence Does Not Explain

Even from a literal perspective, these results do not provide sufficient justification. If digital capabilities can automatically bring sustainable development benefits, then companies with similar levels of digital investment should follow similar ESG development trajectories; But the fact is clearly not so. The recorded correlation coexists with the equally proven digital hazard. Residual differences indicate that certain factors, such as choice, direction, and observation, have not been taken into account in the empirical model between ability and outcome. The following section proposes two conceptual supplements (green innovation as a transformation mechanism; CDR as a governance level) and an infrastructure supplement (management accounting), and views them as propositions rather than outcomes.

3.3. Green Innovation as the Conversion Mechanism

Green innovation refers to the new era of activities that pursue economic and environmental coordination in products, processes, services, and business models. It will demonstrate how digital functions in the system contribute to improving the environment. This model focuses more on the organizational level rather than the technical level: sensor networks visualize waste, data analysis locates its source, distributed ledgers validate green claims but only when these capabilities are combined to create new products, processes, and models will environmental performance change. The above content indicates that the same set of digital tools can generate different ESG outcomes due to the different spaces provided by technology, and innovation can transform some of these spaces into tangible benefits. The intermediary evidence provided by (Zhang & Li, 2025), as well as the quantitative quality research results of (Wang & Zhong, 2024), are consistent with this report.

3.4. CDR as the Governance Layer

Enterprise capability answers what enterprises can do; Green innovation explains how to achieve benefits; But neither of them involves what the enterprise should do. This guidance function is provided by CDR. The record of ethical practices in artificial intelligence is disappointing: despite frequent references to transparency, fairness, and non-harm in guidelines, they are difficult to implement in practice (Jobin et al., 2019; Hagendorff, 2020). Service research reveals the reasons behind this: businesses repeatedly face CDR trade-offs between the commercial value of data usage and privacy and fairness risks (Wirtz et al., 2023), while the broader concept of digital responsibility extends this logic to the social, ethical, and ecological consequences of digitization (Trier et al., 2023). Therefore, CDR should not be seen as another form of reporting, but rather as a design philosophy that actively embeds responsibility into the system itself to prevent problems from occurring, which ESG indicators or green innovation cannot provide independently. CDR guides decision-making by selecting and shaping digital initiatives that align with ESG goals, and establishing review mechanisms to exclude projects whose benefits depend on externalizing digital harm.

3.5. Management Accounting as the Operating Foundation

Management still needs to follow regular management procedures. Section 4 points out that the three major categories of management accounting tools; SBSC, carbon management accounting, and ESG measurement and certification, should be used as the infrastructure for achieving CDR and ESG goals at the practical level, rather than just as ideal goals.

3.6. Integrated Framework and Its Concept

Fig. (1) shows the components of the framework. Digitization (premise) is associated with ESG performance (outcome) through two paths: one is the indirect path, namely green innovation; The other is the direct path, which involves transparency and governance effects. Carbon reduction (governance level) controls these two paths, while management accounting infrastructure (infrastructure level) provides measurement, target setting, control, and disclosure functions for all other parts. The solid arrow represents the relationship supported by the experiment; The dashed arrow represents the association relationship proposed in this article.