Regulatory design, institutional pressures and the coupling of human rights due diligence

Modern slavery, including forced labour, human trafficking and other exploitative labour practices, remains a persistent and often hidden feature of supply chains (Gold et al., 2015). In particular, global supply chains are especially prone to violations of worker-related human rights, such as decent and safe work, or more perversely, the right to liberty and security of a person (LeBaron, 2021). Consumers, in light of increasingly frequent media scandals, have begun to pay more attention to the working conditions under which their goods and services are produced, and have exerted pressure on policymakers and businesses to act (Islam and Van Staden, 2022).

One such intervention by policymakers has been the introduction of mandatory reporting regulations on businesses. Earlier examples of this are the 2015 UK Modern Slavery (MSA, 2015) and the 2018 Australian Modern Slavery (MSA, 2018) Acts (collectively referred to as MSAs). These required businesses to disclose their detection and intervention measures for modern slavery in their operations and supply chains, although the effectiveness of these efforts has been questioned (Christ et al., 2019; Rogerson et al., 2020; Schaper and Pollach, 2021).

Despite these efforts, the quality of disclosures remains poor, with scholars citing vagueness and a general lack of substantive reporting as key issues (Meehan and Pinnington, 2021; Schaper and Pollach, 2021; Pinnington et al., 2023). This is not because businesses failed to comply, but rather because the minimum mandatory reporting requirements are superficial (Christ et al., 2019; Islam and Van Staden, 2022). In the literature, this has resulted in a divergence between those critical of business responses (Christ et al., 2019; Mai et al., 2023), regulatory design (Islam and Van Staden, 2022) and transparency regulations entirely (Gold and Heikkurinen, 2018; Schaper and Pollach, 2021). However, the literature focusing on regulatory design is limited, mainly due to the number of regulations (Nolan and Pryde, 2024), and presents an understudied area.

This study builds on the existing debate by evaluating the Corporate Sustainability Reporting Directive (CSRD, 2022), the EU’s newly implemented mandatory environmental, social and governance (ESG) disclosure framework. Transposed into Irish Law in 2024 as the Corporate Sustainability Reporting Regulations (CSRR, 2024), these regulations require large organisations to report in accordance with the European Sustainability Reporting Standards (ESRS). In particular, Social Standard 2 (S2) specifically addresses “workers in the value chain” and obliges organisations to disclose, in essence, the extent of their human rights due diligence (HRDD) process related to supply chain labour risks, including modern slavery. Unlike earlier legislation, the CSRD introduces legal enforcement, limited third-party assurance and prescriptive reporting categories, which may potentially result in improved effectiveness.

Our study frames these shortcomings in modern slavery regulation as a problem of regulatory design, resulting in a weak neo-institutional pressure to address human rights issues in supply chains (Meyer and Rowan, 1977; DiMaggio and Powell, 1983). Weak or ambiguous regulatory structures allow firms to engage in symbolic or loosely coupled compliance, where the disclosures of firms are not accurate representations of their actions (Weick, 1976; Bromley and Powell, 2012). In contrast, the CSRD’s prescriptive reporting guidance, mandatory assurance requirements, potential penalties and double materiality assessment process are expected to tighten the relationship between reporting and practice. This study, therefore, examines how CSRD design features shape the degree of coupling between disclosure and practice in HRDD. Our research questions (RQs) ask:

The CSRD and the forthcoming Corporate Sustainability Due Diligence Directive (CS3D, 2024) are creating a substantial amount of uncertainty for businesses (JARO and YouGov, 2025), who have reported confusion over reporting expectations and a lack of best-practice guidance (ECCJ, 2025). In Ireland, the policy implementation of the United Nations Guiding Principles on Business and Human Rights (particularly Pillar 2 on corporate responsibility of businesses to respect human rights) has been limited to non-binding national action plans. This context makes Ireland a compelling case for exploring how companies respond to newly mandated human rights due diligence obligations. Additionally, as managers struggle to make sense of how to comply with the novelty of the CSRD, early evidence will be helpful in informing emerging practices. This study aims to contribute to emerging regulatory policy, inform business practice and advance the theoretical understanding of regulatory design and compliance dynamics.

This section reviews two widely studied disclosure-based regulatory instruments: the UK (2015) and Australian (2018) Modern Slavery Acts. It identifies the key structural weaknesses of these initiatives, focusing on their (in)ability to drive meaningful corporate action on modern slavery. The CSRD is then introduced, providing an overview of the procedural and regulatory burdens it places on firms. To develop the regulatory design component as a theoretical anchoring of this study, the CSRD is then contrasted to the aforementioned instruments to establish the proposed key drivers for why this regulation would be expected to produce alternative findings to the literature. This relationship between regulation and corporate action is then theorised through neo-institutional theory, within which the problems of loose coupling in response to neo-institutional pressures have played prominent roles in the literature.

The UK Modern Slavery Act came into effect in 2015 and required organisations to disclose their efforts to identify and remediate any instances of Modern Slavery in their supply chains (Stevenson and Cole, 2018). Section 54, the Transparency in Supply Chain (TISC) disclosure, requires businesses to report on modern slavery in their supply chain. The Australian MSA was approved in 2018 and is similar in requirements to the UK MSA (Nolan and Pryde, 2024).

Although there are minor variations in the reporting requirements, at a general level they require disclosures on the company’s business structure, modern slavery policies, supply chain due diligence policies for modern slavery, identified modern slavery risks and actions taken, how they ensure modern slavery is not occurring in the businesses’ supply chains and the training it provides to staff on modern slavery (Sinclair and Nolan, 2020; Flynn and Walker, 2021).

The MSAs have been noted as problematic by the academic community for several reasons. Previous research into the UK MSA has highlighted a lack of reporting coverage (Schaper and Pollach, 2021) and quality (Pinnington et al., 2023) with evidence of organisations deploying “strategic ambiguity” to avoid accountability for their lack of modern slavery management systems (Meehan and Pinnington, 2021). Both MSAs have also been criticised for lacking financial penalties, having limited guidance on how to adhere to their requirements (Sinclair and Nolan, 2020) and for facilitating a box-ticking approach to compliance (Rogerson et al., 2020; Nolan and Pryde, 2024). Several studies have highlighted that organisations are able to meet disclosure requirements with little or no evidence of substantive due diligence in their supply chains (Schaper and Pollach, 2021; Mai et al., 2023; Pinnington et al., 2023).

Although several studies have assessed voluntary disclosures in the Australian context (Christ et al., 2019), data is currently limited on how the MSA has impacted the modern slavery disclosures of Australian firms (McMillan, 2023). The early evidence on the topic suggests that organisations are either struggling to meaningfully comply or are intentionally using selective reporting to avoid having to engage in addressing modern slavery in their supply chains (Christ and Burritt, 2023; McMillan, 2023). Additionally, Christ and Burritt (2023) found that over 20% of firms failed to report on how they assess the effectiveness of their actions, a core HRDD expectation, often due to the lack of clear metrics or standards.

In diagnosing why a singular reliance on corporate reporting regulations is problematic for tackling modern slavery, Nolan and Pryde (2024) put forward three key components: the quality of the statements, the lack of enforcement and the limited focus on remedy. Although organisations are technically compliant, disclosures often lack the explanation, depth or specificity needed to meaningfully evaluate modern slavery risk management in supply chains (Schaper and Pollach, 2021; Christ and Burritt, 2023; Pinnington et al., 2023). Here, quality refers to the depth and specificity of explanations provided (Pinnington et al., 2023), making comparability and accountability challenging, and as such does not incentivise organisations to “race to the top” in addressing modern slavery (Rogerson et al., 2020; Nolan and Pryde, 2024). The lack of enforcement also reduces the regulatory pressures exerted on organisations to comply.

Moreover, none of the above Acts require organisations to implement remediation systems for victims of modern slavery in their supply chains, despite being a core pillar in the UN Guiding Principles on Business and Human Rights (UNGPs). Remediation in this context refers to the grievance mechanisms and/or the provisions in place for adversely impacted supply chain workers to raise concerns and obtain redress (Pryde et al., 2024). Altogether, these findings highlight the limitations of voluntary and minimally enforced disclosure frameworks in addressing modern slavery. The next section examines whether the CSRD’s more robust regulatory design can address these weaknesses.

The CSRD represents an effort to standardise sustainability reporting across the EU and enhance corporate accountability. Building on the earlier Non-Financial Reporting Directive (NFRD), it introduces several new sustainability reporting requirements for EU companies. These include thorough mandatory and highly prescriptive reporting requirements through the European Sustainability Reporting Standards (ESRS), a double materiality assessment (DMA) process that must be followed when identifying sustainability-related impacts, risks and opportunities, and finally, a required limited third-party assurance on the information disclosed and the DMA. There are in total 10 ESRS “sustainability topics” that the scope of the CSRD covers. Within each topic, there are subtopics and sub-subtopics. Companies are not required to report on all of these, and are instead required to identify which topics they have the highest impact on (impact materiality) and are most impacted by (financial materiality) through the DMA (Baumüller and Sopp, 2021). A simplified overview of the ESRS topics and the DMA process is presented below in Figure 1.

This study focuses specifically on Social Standard 2 (S2) of the ESRS using the 2024 Sector Agnostic standards set by the European Financial Reporting Advisory Group (EFRAG), which mandates disclosure on workers in the organisation’s value chain. This essentially mirrors the supply chain due diligence guidelines for responsible business conduct set out by the OECD Due Diligence Guidance for Responsible Business Conduct and the key elements of the forthcoming CS3D requirements. They also provide a more granular and structural alternative to the UK and Australian MSAs’ TISC disclosure requirements, through detail-rich provisions under each reporting sub-criterion. The key differences between these regulations are summarised in Table 1.

The CSRD can be significantly distinguished from the prior MSAs for several reasons. Firstly, focusing on the points raised by Nolan and Pryde (2024) as to why the MSAs failed to adequately address modern slavery. The CSRD contains penalties for non-compliance. It is also far more prescriptive in what must be reported, with a rich amount of guidance already having been provided by EFRAG. Furthermore, remediation mechanisms, previously absent from the UK and Australian MSAs, are explicitly addressed under S2-3 requirements, and companies are expected to disclose their policy framework for remedy under S2-1.

Secondly, a mandatory DMA is novel for organisations. Financial and impact materiality are not necessarily novel; however, as a mandatory requirement subject to limited assurance, the CSRD introduces a conduct-based requirement rather than purely a disclosure-based requirement. This makes it a theoretically unique regulation to examine in the supply chain human rights and modern slavery domain.

Finally, the CSRD also mandates limited third-party assurance, requiring external auditors to verify the data that produces the ESRS disclosures and that the DMA was conducted appropriately. In theory, this should reduce any intentionally ambiguous or superficial reporting practices, as noted in the assessment of the MSAs above (Meehan and Pinnington, 2021; Pinnington et al., 2023).

In summary, from a regulatory design perspective, the CSRD addresses several design failures identified in earlier disclosure frameworks, including vagueness, non-enforceability and the absence of an accountability mechanism for labour abuses. To make sense of the relationship between this regulation and the way businesses disclose their human rights due diligence efforts, a neo-institutional theory framework, focusing on the isomorphic pressures and the degree of coupling between policy and practice, has been adopted.

This paper adopts an institutional theory approach, specifically neo-institutional theory, focusing on the way institutional environments shape organisational behaviours (Meyer and Rowan, 1977). This branch of institutional theory posits that similarities between organisations’ practices can be partly explained by the institutional environments they operate within, and the “institutional pressures” these exert on businesses (Meyer and Rowan, 1977; DiMaggio and Powell, 1983).

In particular, DiMaggio and Powell (1983) posited three key sources of “isomorphic” neo-institutional pressures. They observed that coercive, mimetic and normative pressures could explain why businesses operating in the same institutional environment adopted similar practices over time. Coercive pressures emerge from regulations and societal expectations (Buttke et al., 2024), and are common in the MSA literature linking the regulations to reporting behaviours (Flynn and Walker, 2021) or in linking buyer needs with due diligence adoption (Samal and Jena, 2024). Normative pressures arise from the professionalisation of a field, as a certain standard of conduct becomes normalised, and is commonly observed in the sustainability reporting field more broadly (de Villiers et al., 2014). Finally, a mimetic pressure to adopt similar business practices to competitors can be observed when firms face uncertainty, such as what to include in a supplier code of conduct (Brockhaus et al., 2019) or a modern slavery statement (Christ et al., 2019). In short, neo-institutional theory provides a lens to understand how firms behave in response to institutional expectations, and what pressures drive them to do so.

However, complementary theory and research demonstrate the existence of gaps between what may be intended by a regulation and how the firms actually respond (Meehan and Pinnington, 2021). This symbolic compliance allows companies to maintain legitimacy in front of these institutional expectations, while not necessarily having to significantly change how they operate (Jamali et al., 2017). In the context of modern slavery disclosures, this is commonly observed; firms meet reporting obligations by disclosing generic statements while failing to implement substantive supply chain due diligence (Stevenson and Cole, 2018; Christ et al., 2019). That being said, whether this is intentionally strategic (Meehan and Pinnington, 2021) or a limitation of internal and supply chain capabilities (Rogerson et al., 2020) is unclear in the literature.

To theorise this, the gap between policy and practice is framed as a coupling problem (Weick, 1976). Coupling as a lens examines the degree of connectedness and determinacy of elements in an organisation. When these elements are “loosely” coupled, they retain higher degrees of independence and indeterminancy, while the opposite holds true in “tightly” coupled situations (Orton and Weick, 1990). Although many studies in the SCM domain do not distinguish between policy, practice and outcomes in examining coupling problems, in light of Bromley and Powell (2012), the authors specify that the focus is on the coupling of policy and practice, as opposed to outcomes. These previously reported coupling problems emerging from the MSAs are presented as a consequence of weak regulatory design, which, through the CSRD design-features discussed above (prescriptiveness, assurance, DMA), are speculated to “tighten” the coupling between policy and practice such that one element is more reflective of the other. This is visualised in Figure 2 below, as the research framework for this study.

The underlying assumptions behind this theorisation are that the design characteristics of regulations influence whether firms comply substantively or symbolically. The existing literature indicates that policies and aspirational sentiments, potentially in the form of targets or goals, are symbolic forms of engagement (Meehan and Pinnington, 2021; Mai et al., 2023). For a “tighter” coupling between policy and practices, companies would be expected to undertake substantive actions, reflective of their policy stances, and could include worker engagement (Benstead et al., 2021), remediation mechanisms (Stevenson and Cole, 2018) and the implementation of mitigating and preventative measures through supply chain management (Trautrims et al., 2021; Jiang et al., 2023). This offers a relatively novel contribution to the SCM literature, bridging neo-institutional theory with regulation and governance in the context of sustainable supply chain management.

To understand how the CSRD has impacted HRDD and modern slavery-related disclosures, a deductive content analysis was conducted on corporate sustainability reports, using the S2 “workers in the value chain” code from the ESRS as the deductive criteria for analysis, as set out in the background section. To add contextual depth and interpretive insight, 32 interviews were conducted with sustainability and supply chain managers in CSRD-applicable companies, as well as consultants serving as industry experts (see  Appendix 1 for details).

Disclosure analysis is a widely used method for assessing modern slavery disclosures and sustainability reporting (Hahn and Kühnen, 2013; Ishaya et al., 2024). This study adopted a deductive content analysis against the pre-established S2 ESRS categories. Content analysis involves systematically reducing textual data into pre-defined categories, enabling a quantitative assessment of the data (Krippendorff, 2018). This is especially suitable for analysing the presence of key elements in a textual document. This approach is commonly used in the analysis of sustainability reports (Hahn and Kühnen, 2013) and modern slavery disclosures (Schaper and Pollach, 2021; Mai et al., 2023).

Disclosures were analysed across two dimensions. Firstly, the presence of a disclosure sub-criterion was assessed, i.e. was coded as 1 if present, and 0 if absent (Schaper and Pollach, 2021). It should be noted that although CSRD reporting is mandatory, some companies may, through the DMA, not find HRDD a material topic, and will not be obliged to report on it. Alternatively, in cases where S2 is material, they may have no policies, practices, metrics or targets in place and are permitted to produce a temporary minimum disclosure requirement (MDR). Where present, the details provided under the disclosure heading were inductively extracted and are discussed in the findings under the S2 disclosure heading. This coding and extraction process was subject to a second reviewer for consistency.

To triangulate and contextualise these findings, exploratory interviews were conducted using semi-structured interviews to understand how organisations are complying with the CSRD requirements and the incoming CS3D. A semi-structured approach was adopted here as the literature on the CSRD or CS3D is limited, and the wider topic of HRDD and reporting in supply chains is nascent (Schilling-Vacaflor and Gustafsson, 2023). Given the exploratory nature of the topic, a semi-structured format enabled novel insights to emerge (Edmondson and McManus, 2007). Interviews were analysed using the Gioia method (Gioia et al., 2013), a variation of the grounded theory approach (Glaser and Strauss, 2017). Although predominantly an inductive approach, Gioia et al. (2013) acknowledge that when the prior literature has been consulted in advance, the underlying approach is more akin to abductive research. The Gioia method follows a process of extracting first-order codes in the data, preserving the natural participants’ language (Villena and Gioia, 2018). These are then grouped through an axial coding process, developing more abstract, theoretical categories. Axial coding is a qualitative data analysis technique that progresses from open, first-order codes to theoretical categories of codes (Strauss and Corbin, 1998; Gioia, 2020). As per the Gioia process, these categories then inform third-order themes or “data structures” that ultimately inform a final model (Magnani and Gioia, 2023).

This study was a component of a larger research project, so not all questions in the interview schedule were relevant to the findings. As the study was semi-structured, and the questions were somewhat interrelated, discussions sometimes contributed to several questions at once. For this study, conversations stemming from QB2, QB3, QD1 and QD2 in the schedule were particularly relevant for the analysis. The Gioia analysis was based on the full interviews, capturing responses to all questions. This analysis was conducted by two of the authors who initially coded together to establish a common understanding of the data in the first-order coding stage. Following this, the coders operated independently and periodically met to compare and contrast first- and second-order codes. From there, the third-order themes and final data structure were arrived at collaboratively. These findings were retroactively compared with neo-institutional theory and coupling to answer the research questions.

The study consisted of two samples: CSRD-compliant sustainability reports and interviews. The CSRD has a phase-in process, where the initial financial year 2024 reporting requirements only apply to Wave 1 Companies (see Table 1 in Section 2). The exact number of Wave 1 companies is not documented, but these companies would also be subject to the CS3D, for which there are publicly available registries detailing which companies are in-scope for that incoming regulation. Using CS3D-applicable companies as a proxy for Wave 1 CSRD-applicable companies, 76 potential companies were identified as potentially being a Wave 1 company from the SOMO CSDDD Database (Leth, 2025). All listed companies’ sustainability reports were reviewed, and 21 were identified as aligning with the ESRS requirements, either on a preparatory or mandatory basis. Additional information on them is available in  Appendix 2. Our analysis screened 100% of the population’s reports and included 28%. While this list is NGO-generated and may not be exhaustive, it currently represents the most comprehensive public source of Irish CS3D/CSRD-applicable firms.

For interviews, Wave 1 companies, which reported on CSRD this year, and Wave 2 companies, which were preparing to report on CSRD for the financial year 2025, were approached and interviewed. In particular, through piloting with a subsample, the researchers found that CSRD implementation was a highly centralised process, with an implementation lead being assigned for the process. These individuals oversaw CSRD implementation across functions and had the most holistic insights of implementation within their organisations, often liaising with the C-Suite. During interviews, consultants were mentioned as being heavily involved in the interpretation and implementation processes, and were also approached to offer sector-level perspectives. These interviewee insights were used to interpret the findings of the content analysis, allowing the researchers to understand the background processes behind the reports. Interviews were approximately 30–90 min, held online and were conducted over Microsoft Teams. The interview schedule is available in  Appendix 3.

Ireland was chosen as the region to collect data from for several reasons. Firstly, it is one of the EU Member States which has transposed the CSRD. Secondly, it acts as a hub for many multinationals incentivised by its lower corporate tax rates, offering a diverse sample of firms across various sectors. Thirdly, Ireland has not implemented any domestic-specific mandatory human rights and environmental due diligence regulation. This means that, unlike other EU countries such as France, Germany or The Netherlands, which have some form of existing framework that will contribute to CSRD reporting, investigating Irish-based companies allows the effects of the CSRD as a regulatory shock to be better isolated and investigated. This study context better complements the theoretical dimension of this study’s aims.

Additionally, Ireland is a traditionally reactive implementer of human rights regulations, deferring to its obligations under UN and EU law (Darcy, 2021). As the development of the second national action plan on business and human rights is currently underway in Ireland, and in light of the CSRD and future human rights regulations, there is also an opportunity to demonstrate how SCM research can contribute towards public policy in addition to managers and academics (Richey and Davis-Sramek, 2022).

In total, 21 Irish sustainability reports were evaluated, and 32 interviews were used in the analysis. Ethical approval was granted prior to data collection, and all participants provided informed consent. The findings are presented in the following section.  Appendix 2 provides more details on the reports analysed, while  Appendix 1 gives an overview of the respondents. Due to participant anonymity concerns, no linkages were established between respondents and reports.

The findings are presented in two parts. Section 1 details how many companies reported S2 requirements, and evaluates the substance of the HRDD disclosures produced. Section 2 then triangulates these observations with internal practitioner experiences to provide deeper insights into some of the limitations experienced by companies in conducting, and consequently being able to report on, HRDD.

Prioritisation of material IROs was done by identifying which disclosure topics were high in both financial and impact materiality, through conducting the DMA. Using this approach, 38% (8 / 21) of organisations considered workers in the value chain a material issue. Additionally, several organisations reporting on S2 provided a minimum disclosure requirement, essentially stating they were not currently able to report on the sub-topic but were in the process of developing management systems. Some basic HRDD data, such as policies, for firms that did not report under S2, could be extracted from supplier management disclosures under their governance (G1) disclosures. It is also worth noting that there were significant variations in reporting format. Some reports listed each criterion under an S2 header akin to an itemised receipt. Others developed a longer narrative with stock images and case studies/examples. Table 2 shows the findings by criterion for the Irish companies evaluated.

Across 21 Irish reports, S2 coverage is uneven. Although policies were widespread, with 90% of the sample having some policy relating to value chain workers, very few engaged with value chain workers or had mechanisms in place to track the effectiveness of actions taken with regard to workers in the value chain. At a high level, this is immediately consistent with a gap between policy and practice.

Almost every company had some policy related to child labour, modern slavery or exploitative working conditions. These came in the form of codes of conduct, human rights policies or human rights and environmental due diligence policies. It is worth noting that many of these policies were blanket statements, such as generalised human rights policies, while others were explicitly developed with supply chains in mind. These results largely align with previous research on codes of conduct (Brockhaus et al., 2019) and the substance of modern slavery disclosures (Schaper and Pollach, 2021; Pinnington et al., 2023).

This was the least disclosed element of the S2 reporting. Although at a general level, all companies within the sample detailed how they engaged with stakeholders, as this is a requirement of the general ESRS disclosures, information on how the companies engaged with value chain workers specifically was scarce. In addition to the three companies in Ireland that reported on this criterion, two provided a minimum disclosure requirement (MDR), highlighting they did not currently do this and were in the process of developing engagement channels. Two companies disclosed only supplier communications, without indicating any direct worker engagement.

While some organisations provided MDRs for S2 engagement, others avoided explicitly stating this and provided ways they engage with the supplier, which may or may not involve directly engaging with workers. Furthermore, there were several methods through which the focal firms “engaged” with workers in the value chain. In evaluating the S2-2 disclosures, four areas emerged. These were:

1. proxies who acted on behalf of the workers for the focal firm’s engagement;

2. tools used by the focal firm to engage with workers in the value chain;

3. stakeholders that were involved in the consultation process; and

4. supplementary activities and initiatives the focal firm participated in.

In relation to proxies, very few of the companies directly engaged with the workers; instead, they engaged with employee, supplier and union representatives. When employees were “directly” engaged, it was through interviews conducted by auditors. Turning to tools, most companies made reference to the use of audits as a means of engaging with value chain workers. Less common were “conversations” and “open dialogue”, although what these actually entailed was not provided. In two instances, self-assessment questionnaires for the suppliers themselves were listed under the S2-2 header. Reliance on proxies and audit tools – rather than engaging with workers directly – suggests a weak degree of coupling between a clear stance on human rights in the supply chain and the actions required to uphold these standards. Instead, traditional risk management tools seemingly substitute direct worker voice considerations.

Every company detailed the actors whom they consulted regarding workers in the value chain. These mainly included suppliers, NGOs, auditors, industry working groups and local authorities. Finally, some companies referred to initiatives and activities they participated in. These served the purposes of learning, exchanging information, developing industry initiatives and facilitating consultations with stakeholders. Furthermore, it is worth noting that these applied to the first-tier of the supply chain as opposed to the extended supply chain.

Situated more broadly in the literature, these findings reinforce the shortcomings in incorporating worker voices (Stephens et al., 2025) and effective stakeholder engagement (Soundararajan et al., 2021) into sustainable supply chain management (Reinecke and Donaghey, 2021).

Every company that reported on S2 specifically referenced a “Speak Up” tool or platform as a method for value chain workers to raise concerns. As part of the supplier audits and codes of conduct, suppliers have to make employees aware of these communication channels. That being said, one company acknowledged that they do not currently assess how aware the actual workers are of these, or the degree of trust they place in them, while none of the other companies commented on this potential issue.

One company explicitly referenced the concept of remediation. In cases of severe human rights violations, they will terminate the relationship with the supplier and inform the local authorities. In a similar vein, other reports also made reference to ceasing supplier relationships or introducing corrective action plans depending on the severity of the violation, and these companies appear to adopt a similar understanding of remediation to Stevenson and Cole (2018). Remediation appears to have been interpreted narrowly here, focusing on termination versus corrective action (Skerritt et al., 2025), as opposed to having a more victim-centric approach (Kach et al., 2025).

Although some of the recent literature has discussed the development of innovative digital worker engagement technologies to highlight labour violations (Chanani et al., 2022; Sorg et al., 2023), there was no evidence of these in the samples investigated. Furthermore, while the presence of grievance platforms was universal, no companies reported on their effectiveness. This raises concerns about the actual impact of worker voice tool adoption in supply chains. Universal hotlines with no effectiveness metrics, evaluations or potential awareness for workers suggest weak implementation of a remediation policy.

Disclosures under S2–4 were frequently vague. Several firms submitted MDRs, indicating no violations had been detected and therefore no actions were taken. Many “actions” referred to action plans or policies surrounding supplier breaches of codes of conduct. These are similar in nature to many of the “remediation” mechanisms referred to earlier, i.e. the termination of suppliers in instances where a corrective action plan is not suitable. The bundling of S2-3 and S2-4 disclosures in some cases could indicate conceptual uncertainty around what constitutes remediation versus action, reflecting a logic of reacting to incidents, as opposed to adopting preventative measures.

Additional actions identified, which are well-documented as practices in sustainable supply chain management (Yawar and Seuring, 2017), included auditing, supplier training, internal buyer training, risk assessments and supplier capacity building for human rights engagement. Risk assessments largely followed the OECD-recommended approach of identifying higher-risk areas/suppliers and then investigating them more thoroughly. A substantial number of companies use EcoVadis as their third-party risk assessment and monitoring platform. Interestingly, despite firms essentially reporting on their HRDD process through S2-1 to S2-5, a substantial number of reports briefly noted conducting due diligence in their supply chains as an action taken, without elaborating further.

However, nearly all companies failed to report metrics tracking the effectiveness of these actions. That being said, many of these actions were only implemented recently, indicating that the CSRD may have acted as a catalyst for the adoption of these measures. In a couple of instances, the number of risk assessments and/or audits conducted year on year acted as an indicator under the S2-4 heading; however, it is unclear how this informs the extent to which actual adverse human rights impacts have been addressed. This is in line with findings from Christ and Burritt (2023) and McMillan (2023), who note that companies struggle to define appropriate KPIs for modern slavery risk reduction.

Taken together, this demonstrates a relatively weak implementation of management systems for HRDD, with some well-defined “responses” to the detection of supply chain human rights violations, without necessarily being proactive about mitigation or prevention. This is further problematic, as if this “reactive” approach to modern slavery is adopted, then the burden is placed on identification efforts, with several companies declaring they detected no code of conduct violations. As Meehan and Pinnington (2021) note, finding no human rights violations in the supply chain typically reflects a failure to detect, rather than the absence of violations.

In addition to the five companies in Ireland that had targets associated with workers in the value chain, four provided MDRs. Three main targets emerged from these companies, albeit worded differently. The % of suppliers agreeing to adhere to the supplier code of conduct was presented as a target in three of the companies. The % or number of suppliers who have been third-party audited was another metric; some based this on the total % of the supplier base, while others based it on the % of suppliers based on spending – i.e. X% of supplier spending has been audited, rather than the number of suppliers proportionate to the supplier base. Finally, the year-on-year change in audit scores was presented as a target.

The audit score was the closest target to measuring human rights performance that emerged from these disclosures, but many companies did not disclose whether any severe violations had been detected, i.e. forced or child labour. However, audits are controversial, and the (in)effectiveness of audits in addressing labour abuses in supply chains is well documented (LeBaron et al., 2017; Islam et al., 2018). At this point, for the majority of S2-5 disclosures, corporate targets appear to be centred around auditing the (first-tier) supply base, as opposed to addressing, mitigating or preventing human rights violations across the extended supply chain.

Taken together, these results demonstrate that supply chain HRDD is a novel concept for many firms, with a significant amount of the sample not considering it a material matter to currently engage in. Those that have begun reporting on HRDD currently limit their efforts to their tier-1 suppliers, with no references to multi-tier or extended supply chain activities to identify and address adverse human rights impacts. Practices and corporate approaches to remediation largely reflect those captured in the existing literature (Yawar and Seuring, 2017; Stevenson and Cole, 2018) with the representation of worker voices (Stephens et al., 2025) and human rights targets (Bartels and Schramade, 2024) in the supply chain being highly underdeveloped.

Two key strands of findings emerged from the interviews. Firstly, some results from the Gioia analysis, particularly around the DMA (Question B2), upon comparison, overlapped heavily with neo-institutional coercive, mimetic and normative pressures. Secondly, respondents’ accounts of trying to engage with, collect data from and report on the supply chain illuminate how the loose policy−practice coupling observed in the S2 disclosures may be a consequence of internal and supply chain capabilities.

Beginning with the role of neo-institutional pressures, the CSRD, through its design, facilitates coercive and mimetic pressures to operate on the firm. A potentially unintended consequence is the introduction of normative pressures. Neo-institutional pressures were not coded directly in the Gioia analysis, but the axial codes that emerged from the data were retrospectively evaluated through a neo-institutional lens. Figure 3 demonstrates how the coders linked the second-order categories in the data to neo-institutional pressures. This is conceptually anchored on the DMA, as this is where these neo-institutional effects were most commonly observed. A more detailed Gioia analysis structure is available in  Appendix 4.

Beginning with coercive pressures, an abstract perception of regulation as something that needed to be complied with, and the third-party limited assurance associated with the DMA and disclosed data seemed to be the driving factors for engaging in CSRD implementation. The term “abstract” is used here, as managers were not concerned with non-compliance penalties at this point, but rather viewed regulations as having rule-like statuses that had to be adhered to, as one manager stated:

Instead, a far more common occurrence was the pressure associated with receiving a limited third-party assurance. This appeared to be a significant motivator across the sample for meaningfully implementing CSRD, and was coded as a coercive pressure as it was a mandatory requirement of the regulation. One manager explained the role of third-party assurance providers, and why they may lead to a better connection between disclosures and actions, as follows:

Turning to the mimetic pressures, the first step of the DMA was a business environment scan. As managers and consultants explained, this covered an analysis of peers, industry trends and media articles to identify an initial list of potential sustainability IROs for the company. On the premise that mimetic pressures emerge from copying peer behaviours in the face of uncertainty (DiMaggio and Powell, 1983), the peer analysis code that emerged in the data analysis was linked to mimetic pressures. Although this data was collected cross-sectionally, managers felt that in the coming years, contrasting DMA results within industries could influence practices, making it harder to produce below industry-average disclosures:

Additionally, throughout the DMA and further CSRD implementation, a heavy reliance on third-party consultants to fill the knowledge and technical gaps within businesses on these matters seems to have given way to the introduction of normative pressures. Many companies highlighted the use of third-party consultancy firms for technical interpretation and methodological guidance on carrying out the DMA, and to a lesser extent, on writing up the finished CSRD disclosures. As one manager stated:

A consultant at a major firm providing CSRD services elaborated on this, describing their engagement as:

Having interviewed CSRD consultants for a significant portion of the European market share in this study, the authors observed a high degree of homogeneity in how they provided guidance to businesses on DMA execution and CSRD compliance. These codes were ultimately categorised as normative pressures, as opposed to mimetic pressures, as consultants are effectively setting and enforcing the normative standard around DMA methodologies and CSRD reporting. This expectation of conduct is only reinforced as many of these major consultancies also act as third-party assurance providers for companies they have not assisted in this process, indicating that their expectations of auditors would be in line with their company’s expectations as CSRD consultants. This was supported by the interviews with assurance providers:

Overall, this creates a strong, self-reinforcing system of professionalisation and normalisation around the DMA procedure and CSRD implementation more broadly.

While the coercive and mimetic pressures discussed here speak more to the RQ2 of this study, on the coupling of policy and actions under HRDD, these normative pressures demonstrate how consultants are shaping the interpretation of CSRD.

Turning to experienced challenges in implementing S2 reporting and developing practices reflective of policies in the supply chain, several facts emerged regarding the DMA process that may explain why companies have a policy framework around human rights in supply chains, but may not be implementing meaningful practices. Presented in Figure 4, several factors were captured during interviews and the analysis process that demonstrate how practices in the lead up to determining if S2 was a material topic (Steps 2–4), played a role in deprioritising HRDD. Commonalities here were the role of bias and engagement challenges. After these, the actual challenges managers experienced in trying to work with their supply chains (Step 5) are explained.

One key risk in the CSRD’s design is its reliance on double materiality, which allows firms to exclude S2 reporting if it does not score highly on the DMA. The second step in determining this materiality rating is through the internal stakeholder consultation process. At a broader organisational level, every internal stakeholder felt their subject area was the most important to the business, biasing the identification and quantification of IROs. As one manager stated:

In contrast, existing literature has demonstrated that supply chain managers, who would have been the internal stakeholders for S2, often perceive human rights risks as less likely in their supply chains (Simpson et al., 2021; Cakir et al., 2025). Triangulating this with consultants, they reaffirmed this effect:

Similar issues can manifest during the external stakeholder consultation process (Step 3). As noted in the analysis of disclosures, proxies were somewhat commonplace as a substitute for suppliers and workers in the supply chain. Several managers stated that they did not choose to include external stakeholders in their DMA process for this year. Some provided the logic that they were in an exploratory phase and that external stakeholders would overcomplicate the process; others were satisfied with the data collected internally. Regardless, this means that an internal stakeholder was used as a proxy for identifying the actual and potential IROs on human rights in the supply chain. This could result in significant biases. As one consultant stated:

When companies proceeded to engage in external stakeholder engagement, a key challenge was supply chain engagement. Poor response rates to questionnaires and general engagement from suppliers were endemic across the sample. This, in turn, led to the use of proxies, inviting downward bias in the scoring of the S2 IROs.

More intentional was the artificial deflation of material topics in the scoring reduction process (Step 4) by the governance or steering committee. Several consultants and managers noted this, with one illustrative quote provided:

Turning to companies that have begun work on S2 disclosures, the overwhelming evidence indicates that the capabilities of the firm bottleneck the SSCM practices and outcomes of the supply chain. In particular, our sample consists of very large companies, almost all of which expressed the size of the supply base as a limiting factor in their due diligence efforts. One company, which has been subject to media pressures for approximately two decades, had a multi-tier SSCM in place; otherwise, every company stated they limited their supply chain due diligence to a fraction of their first-tier suppliers:

Additionally, some managers were aware of the need to go further into the supply chain, but felt that the technology and infrastructure did not exist yet:

Taken together, there is evidence that, in some instances, greater coupling is occurring between policy and practice in the HRDD context. However, these practices are still in an early stage of development, and are being bottlenecked by a limited internal appreciation of HRDD-related IROs in addition to supply chain engagement barriers. In other instances, certain practices in the DMA process work to keep policies and practices weakly coupled, enabling the deprioritisation of HRDD. Consequently, neo-institutional pressures have worked to some extent in driving greater policy−practice HRDD coupling, but limitations still exist.

This study sets out to provide early evidence on supply chain HRDD disclosures under the newly implemented CSRD, and to provide an analysis of the early effects of regulatory design on the degree of coupling between business policies and practices. Section 4.1 evaluated what firms are disclosing on their HRDD, while Section 4.2 shed some light on why these gaps potentially exist, and the challenges associated with aligning practices with policy in HRDD.

Firstly, considering the findings with regard to what was disclosed under these reports. Most companies had policies prohibiting the violation of human rights standards in supply chains, which are widespread and well-documented in the literature (Brockhaus et al., 2019). Worker voice, or the lack thereof, was also observed, and is interesting to note against the backdrop of this becoming an increasingly important topic on the agenda of supply chain management research (Soundararajan et al., 2021; Stephens et al., 2025). This analysis revealed an almost complete absence of direct engagement between focal firms and workers within their supply chains, accompanied by superficial and generic grievance mechanisms that were not tailored to supply chain workers. “Remediation” mechanisms, actions taken and targets were largely in line with the established social SSCM literature (Yawar and Seuring, 2017; Stevenson and Cole, 2018; Schaper and Pollach, 2021).

Turning to the findings related to the application of neo-institutional isomorphic pressures. One implication of these findings is that it positions auditors and consultants as institutional gatekeepers, shaping procedural norms and reinforcing peer norms, questioning non-aligned materiality decisions and scrutinising to-be-disclosed data, with, an albeit currently abstract, risk of legal penalties for procedural negligence. Penalties were not the core motivator in any interviews held, rather it was accountability to the auditors, perhaps stemming from legitimacy concerns, that was the key coercive mechanism behind CSRD implementation at this point.

From a regulatory design perspective, at an abstract level, the CSRD can be distinguished from the MSAs by reducing the agency of how firms can disclose. This is driven by more direct regulatory requirements, heightened transparency requirements enforced by third-party auditors and increased pressure exerted by the industry. One potential concern arising from this, as also noted by Rogerson et al. (2020), is that mimetic pressures may lower the quality of reporting rather than raise it. However, participants felt that the assurance process would prevent this downward pressure.

An additional factor that requires some consideration, from a regulatory design perspective, is the DMA. This mandatory process distinguishes the CSRD from a purely disclosure-based regulation, such as the MSAs. However, it also does not place a burden on firms to address any identified impacts, like full due diligence regulations. As such, although firms could state they are not currently taking actions on a material topic under the CSRD, they have to demonstrate that an internal mapping and stakeholder consultation process was carried out, subject to limited assurance. This has the potential to address the awareness barrier identified by Rogerson et al. (2020). However, this study also identified several practices in the DMA which, in tandem with supply chain knowledge gaps, may perpetuate a continued underappreciation of human rights issues in the supply chain, legitimising keeping HRDD policies and practices loosely coupled.

As documented, firms have already discovered ways to legitimately manipulate the DMA process to exclude supply chain IROs. This enables the perpetuation of selective disclosures (Meehan and Pinnington, 2021), albeit through new means, and could enable companies to stay wilfully ignorant of human rights issues in their supply chains (LeBaron et al., 2017), allowing it to remain an abstract “not in our supply chain” phenomenon (Simpson et al., 2021). Given that consultants and auditors have legitimised this process, correcting this practice requires regulatory intervention or further clarification from EFRAG for assurance providers on the permitted use of proxies and steering group scoring deflations.

More pragmatically, this study identified several practical limitations firms face in trying to recognise and address human rights violations in their supply chains. For the overwhelming majority of companies interviewed, the CSRD was what drove them to consider the sustainability of their supply chains. Consequently, SSCM appears to be at a nascent stage for many of these large Irish firms. When human rights in the supply chain were an area of interest for their CSRD implementation and reporting, it was practical limitations around engaging the supply chain that kept policies and practices loosely coupled.

The CSRD marks a significant shift in sustainable supply chain reporting for many European businesses. Through an evaluation of 21 Irish sustainability reports, alongside 32 interviews, this study examined the effectiveness of the CSRD in aligning HRDD policies with practices. Findings indicate modest yet significant improvements in transparency, particularly concerning due diligence for adverse human rights impacts in the supply chain. How this increased transparency will translate into tangible improvements in rights outcomes remains an open question, particularly in light of persistent gaps in worker engagement and remediation.

Our findings demonstrated the effects of all three isomorphic pressures, particularly during the DMA process. Mandatory third-party assurances created a layer of accountability and acted as a coercive pressure on firms. How the behaviours of peers and the industry shaped the IRO identification and prioritisation of the focal firm was documented. The structured DMA process and the oversight of assurance suggest that peer comparisons and industry norms will play a more prominent role under the CSRD than under earlier frameworks such as the MSAs. The role of consultants as normative standard-setters, in the absence of internal capabilities, was also highlighted. This suggests that future research examining novel regulations that require non-traditional expertise should not ignore the role of these actors in shaping business behaviours.

As our analysis of S2 disclosures revealed, the CSRD has introduced a substantial amount of uncertainty that businesses are trying to navigate, further compounded by the Omnibus. This article provides a summary of how the various sub-codes of ESRS S2 are being approached by EU companies. It provides practical insights into common practices, targets and grievance mechanisms that companies can benchmark their own approaches against and avoid misalignment with emerging norms. This is particularly beneficial to companies coming into scope for Wave 2 or Wave 3 reporting, which will be in the early stages of CSRD implementation.

The CSRD has improved on previous reporting regulations in various ways. Limited assurance requirements appear to be a key regulatory lever in fostering internal organisational change. The highly prescriptive nature of the reporting requirements also seems to have had a positive impact on the quality of disclosures. Furthermore, implementing a duty of conduct, i.e. the DMA, seems to have had a significant impact on obliging companies to become more aware of their relationship with sustainability matters. That being said, the CSRD fell short in several ways. Firstly, the DMA allowed organisations to avoid reporting on S2, which raises significant concerns from a human rights perspective. Secondly, although the directive improved upon previous regulations, it stops short of requiring substantive remediation for affected workers, beyond supplier termination or corrective action. From a business and human rights perspective, how this contributes to remediation is unclear (Nolan and Pryde, 2024). Finally, as noted in the findings, the use of proxies and the ability of steering groups to artificially deflate material matters during the scoring reduction process introduce bias into the materiality process. Further clarification is required from EFRAG and the EU on how third-party auditors and companies should engage in this.

The CSRD represents a novel regulatory intervention that is having a substantial impact on corporate and supply chain sustainability across Europe. Through this study, the authors have identified several aspects of the CSRD process that remain underexplored and warrant future research.

Firstly, a more detailed investigation of the DMA process is needed. Interviewees described a process involving business environment scans and stakeholder consultations; however, the operationalisation of this process could not be explored in depth. Specifically, future research could investigate and theoretically engage with what makes a topic a material priority based on internal and external factors to the company. The finding that few Irish companies considered S2 material, described by one participant as an “Ireland-specific issue” (Manager 6), suggests institutional and geographic influences on CSRD interpretation and disclosure decisions.

Secondly, under S2-2, organisations have to disclose how they engage with supply chain workers. Our findings showed widespread use of proxies (audits, NGOs) with limited direct engagement with workers themselves. There is emerging research on innovative worker voice tools (Chanani et al., 2022; Sorg et al., 2023) and future research could explore their use in improving engagement with supply chain workers. Alternative avenues include a “human rights approach” (Cao et al., 2024), empowering workers to form associations and engage with buyers, as procedural rights are often underexplored in labour exploitation research (Amengual et al., 2020).

Thirdly, the results presented are early evidence of the CSRD’s effectiveness in improving human rights supply chain disclosures. As discussed, interviewees highlighted the role of normative and mimetic neo-institutional pressures in shaping the DMA outcomes. Future research could first investigate how these pressures influence the relationship between coercive regulation and sustainable supply chain orientation, through the DMA topics identified. Additionally, longitudinal studies could assess whether such institutional dynamics elevate disclosure standards over time (race to the top) or result in strategic minimalism (race to the bottom) or in the tighter coupling of policy and practice.

Finally, the use of third-party consultants to support CSRD implementation and reporting was widespread across the sample. The findings of this study demonstrated how these third parties create the norms around these procedures and approaches to sustainability. Given that in Europe a relatively small number of major consultancies dominate the market for supplying these services, further research could examine the role of these actors in shaping the way that CSRD is understood and implemented in multinational enterprises through nexus supplier or relational capabilities theory (Yan et al., 2015; Emberson et al., 2022).

This study was subject to several limitations. The main limitation is the sample size. Although the number of Irish sustainability reports appears modest compared to large-sample studies (Schaper and Pollach, 2021; Pinnington et al., 2023), it does constitute a significant portion of the known in-scope population, and all publicly available reports were reviewed. The deductive findings were further validated through interviews. Although the interview sample was limited, participants provided consistent responses that reinforced the results of the document analysis. To enhance reliability, the interview findings were also subjected to peer checking, whereby participants were provided with a summary of their interview, including key themes and discussion points, to verify accuracy. Altogether, this multi-method approach enhanced the reliability of our findings. In the Irish context, it supports the representativeness of our sample and the methodological rigour of our conclusions.

This research is supported by Atlantic Technological University through the Postgraduate Research Training Programme in Operations and Supply ChAin Research (OSCAR).

*There were contextual prompts and background text in the original version that have been excluded due to word count restrictions; the core questions remain unchanged.

QA1. How would you describe your role, and how does it relate to the implementation of CSRD and CS3D?

QB1. Has your team, department and/or company had to change its existing policies and practices to meet the reporting requirements? If so, what changes have occurred? (Areas of focus include infrastructure changes, resourcing, training, data flows and organisational alignment)

QB2. Can you describe your company’s approach to the double materiality assessment and the identification of material topics?

QB3. What were the initial challenges faced in meeting the CSRD requirements, and what are the current challenges being encountered?

QB4. Do you perceive any benefits associated with CSRD compliance and reporting? If so, what benefits, and more broadly, do you see the CSRD as more of a compliance risk or a value creation opportunity?

QB5. How have you leveraged third parties, such as consultants and technology providers, to work towards the CSRD requirements?

QB6. Has the CSRD influenced the way your company engages with its supply chain? If so, how and what challenges does the supply chain pose for your company’s CSRD reporting?

QC1. Does your organisation have a human rights and environmental supply chain due diligence policy in place? If so, how does it compare to the CS3D/OECD requirements and what changes are required?

QC2. Is/was your organisation preparing to comply with the CS3D? Has the Omnibus impacted this?

QC3. The CS3D due diligence process can largely be broken down into six steps. Considering the following, what is the current state of your policies and practices in these areas, and what are the key challenges associated with each step?

QC3a. Embedding due diligence policies in your and your suppliers’ operations?

QC3b. Identifying potential and actual environmental and social risks in the supply chain?

QC3c. Mitigating, preventing or ceasing both potential and actual adverse environmental and social impacts in the supply chain?

QC3d. Tracking the effectiveness of the above measures and developing associated key performance indicators (KPIs)?

QC3e. Remediating any social and environmental harms caused by suppliers to value chain workers and affected communities?

QC3f. Reporting on the five steps above?

QD1. Do you think the regulatory requirements introduce any novel challenges to your organisation, or are these existing challenges that have just been given increased urgency?

QD2. Has the CSRD improved disclosures, compared to prior legislation and voluntary frameworks?

QD3. How has the Omnibus impacted your organisation?

As noted in the methodology. This data was collected as part of a larger project on CSRD implementation. What is detailed below are the results of an analysis following the Gioia methodology up to the second-order categories, which were then linked to neo-institutional theory in Figure A1. In Figure A2, the Gioia analysis includes the third-order themes. These were bundled under a “challenges” construct in the full analysis. This coding process, with the exception of the neo-institutional pressures labelling, was informed by the interview schedule in  Appendix 3.