SC
Sunhee Choi
Content thumbnail From Principles to Actions
AI Content Chat (Beta) logo

From Principles to Actions

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION 6. FROM PRINCIPLES TO ACTIONS This project has con昀椀rmed that there are technical, legal and behavioral factors that will always lead to decentralization of data, so the objective of a single centralized open data ecosystem is unrealis- tic, even if it was desirable. A more realistic future vision for an effective supply chain data ecosystem is federated data spaces where different actors collect and process data in a relatively harmonized manner and are able to connect with each other directly or via data hubs or intermediaries where data sharing occurs in more or less extensive ways. As this network of connections grows, and trust builds, more data will 昀氀ow around the system, at greater scale, and for greater impact. The following sections set out, based on the insights from this project: 1. practical recommendations addressed to the various actors involved in the supply chain data eco- system to achieve better data sharing and a more effective data ecosystem over the long term 2. three areas in which Tech Against Tra昀케cking intends to drive further action in collaboration with relevant stakeholders, including those already engaged through this project. 6.1 Recommendations to Businesses, Policymakers and Civil Society Organizations Based on this project’s 昀椀ndings on the current siloed state of data sharing, and barriers and enablers that exist for effective data sharing, we have identi昀椀ed 6 areas where principled actions by different actors can help build a more effective data ecosystem to address forced labor in global supply chains. 4848

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION FIGURE 5: Principles for Effective Data Sharing 2. THE RIGHT RESOURCES Invest in data management Share costs of the data ecosystem equitably RESOURCES DATA BEHAVIORS 1 1. THE RIGHT DATA Standardize data collection 3. THE RIGHT BEHAVIORS for greater interoperatbility Focus on progress and Reverse presumptions in favor of data sharing impact, not just risk Build trust in the system Know your role 4949

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION 1. THE RIGHT DATA Standardize Data Collection for Greater Interoperability All actors that collect data related to instances of forced labor should seek greater alignment and harmonization in the way such data is collected. This means building on the ILO’s forced labor indicators to de昀椀ne common criteria, including questions and metrics, to evaluate the risk of forced labor in practice. This will also require clearer de昀椀nition, and agreement, on how the concepts of modern slavery, forced labor and human tra昀케cking intersect. Such efforts should involve leading providers of audit or other assessment standards and services, as well as the growing list of “worker voice” tools, to ensure any standardization is adopted as widely as possible and easily integrated into existing enterprise risk management systems. They should also involve other frontline organizations (such as police border forces, or civil society organizations) tasked with identifying cases of forced labor, and who grapple with the same ch allenges to detect and report indications of forced labor. Governments should support such efforts by framing relevant policies and regulations in ways that incentivize the collection of harmonized data. Initiatives such as the EU’s plans for a public database to support compliance with the proposed EU Regulation on banning products made with forced labor (and similar initiatives in the United States, Australia, and New Zealand) should be opportunities to promote standardized criteria for identifying forced labor risks. See Section 6.2 below on how Tech Against Tra昀케cking and BSR intends to support collaborative action on this recommendation. Focus on Progress and Impact, Not Just Risk The data ecosystem currently focuses on identifying risks of forced labor, which is understandable given that’s the focus of current global standards and emerging regulations requiring risk-based due diligence of supply chains, as well as banning products suspected of being linked to forced labor. The focus on risk too often leads businesses and solution providers to con昀氀ate the evaluation of risks of forced labor (and its adverse impacts on affected workers’ human rights) with risks to the business that would result from a connection to forced labor. The objectives, and mitigation strategies, are very different. This approach may have the unintended consequence of worsening the harm to individuals, for example by leading companies to withdraw business where it perceives risks are too high.28 An effective supply chain data ecosystem should enable an 28 In January 2024, the Federation of German Industries (BDI) released the results of a survey of around 400 companies on the German Supply Chain Due 5050

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION understanding of the latter risks, i.e. of impacts of forced labor on affected workers. With that in mind, an effective data ecosystem should not only surface risks but also progress and where practices are improving in global supply chains. Those who contribute data to the ecosystem (such as civil society organizations, auditors, or other actors engaging directly with workers) should balance their focus on violations and non-compliance with an effort to understand and report on improvements. In conducting due diligence on suppliers and other partners, companies should seek dynamic risk data that captures the “direction” of risk, taking into account positive signals of progress, as well as risks. Similar dynamics affect governments’ willingness to share data that could deter foreign investment and trade. Strategies to encourage governments to share data more openly (as recommended below), should also focus on indicators of progress, not just risk. 2. THE RIGHT RESOURCES Invest in Data Management It is a precondition to an effective data ecosystem that data is “昀椀t” for sharing. This starts with determining what valuable data your organization holds, for its use, and potential use by others in the supply chain data ecosystem. In this context, organizations need to invest appropriate technical and human resources to ensure they can collate, clean, and organize (potentially also analyze and/or aggregate) their data in such a way that it can be easily exported to others, or is easily “discoverable” if it can be made public. Many companies are doing this as part of general strategies to digitalize their business and to capture 昀椀nancial value in their data, but few are focusing these efforts on unlocking the potential impact of their data for human rights risk management objectives. Governments, many of whom are publicly committing to strengthen their data management practices, have a critical opportunity to identify and disclose what relevant data they hold across agencies (e.g. labor inspection 昀椀ndings, migration data) and make that data more easily accessible to external stakeholders. Such efforts should include consulting with businesses and civil society organizations to understand what data would be valuable to access (e.g. more customs data to increase the accuracy of automated supply chain mapping solutions). Governments should also support SMEs in accessing the 昀椀nancial and technical resources to enhance their data management practices, so they can in turn connect and contribute to the global supply chain data ecosystem. This should include supporting capacity-building efforts where appropriate, with the support of third parties, including other governments and the private sector. Diligence Act (LkSG). According to the survey, 14% of respondents said they are leaving high-risk countries, 24% are trying to reduce their number of suppliers, 24% are avoiding suppliers that are di昀케cult to check, 39% are avoiding risky suppliers. 5151

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION Recognizing that some governments will be more technically and 昀椀nancially constrained than others to take these actions, governments with more resources and more developed data practices should also encourage and support other governments to improve their data 29 practices. This could be through trade agreements, ILO discussions, or participation in initiatives like Alliance 8.7 or the Bali Process (which include speci昀椀c commitments to leverage 30 technology to combat forced labor and human tra昀케cking ). The resourcing and technical expertise needed to enable more data sharing by governments and other actors present in sourcing countries should be given due attention in public policy dialogues. Share Costs of the Data Ecosystem Equitably Collecting and sharing data has human, technical, and 昀椀nancial costs, which well-resourced companies and government agencies have more power to do, compared to smaller NGOs or companies, exploited workers and members of the public. This means the type of data being collected and shared is driven by the interests of those more powerful actors in the data ecosystem, and valuable data may be left untapped. Governments should invest more in their data management but also support smaller actors in connecting to the supply chain data ecosystem, e.g. by supporting technical capacity building so smaller enterprises can participate in data-sharing initiatives such as Open Supply Hub or the ITC’s sustainability map. In particular, governments and international organizations committed to supporting efforts to combat forced labor and tra昀케cking should support more investment in technology and data management for less technically equipped (or 昀椀nancially resourced) actors in source countries, including companies, civil society organizations, and governments. Technologically advanced companies should look for ways to support governments, smaller companies, and civil society organizations with the digitalization needed to improve their data management, for the ultimate bene昀椀t of other actors in the ecosystem, including themselves. In some cases, this may mean supporting training on basic data management skills such as collecting and analyzing data in spreadsheets. Private companies should support business models for more “open” collaborative solutions, e.g. paying for a solution shouldn’t exclude the possibility that some of the data “generated” is shared for the public good. 29 Noting that several governments participating in Alliance 8.7 are committed to improving their data collection and management systems related to forced labor risk. See for example F椀樀i’s efforts to develop a paperless labor inspection system or Argentina’s plans to produce statistical data on human tra昀케cking and forced labor. 30 The Bali Process’ 2023 Ministerial Statement included a commitment by participating Ministers to “enhance the capabilities of law enforcement authorities to leverage technology to combat people smuggling, human tra昀케cking, and related transnational crimes”, “strengthen cooperation through facilitation of policy dialogue, information-sharing, and capacity building” and “continue collaboration with the private sector, with particu- lar emphasis on digital technology”. (Source: https://www.foreignminister.gov.au/minister/penny-wong/media-release/bali-process-eighth-min- isterial-conference-co-chairs-statement, 11 February 2023) 5252

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION 3. THE RIGHT BEHAVIORS Reverse Presumptions in Favor of Data Sharing Too often, organizations default to a position of not sharing unless they are compelled or convinced to do so. All actors should seek to reverse this presumption, by articulating and embedding in their organizations the business case for why sharing data may be valuable to their objectives. In practice, this will likely require organizations to reverse their legal counsel’s burden of proof to demonstrate that data cannot be shared (i.e. presume that it can be shared unless proven otherwise). This recommendation has strong synergies with the wider corporate culture change towards more transparency that is needed to meet the expectations of emerging human rights due diligence and reporting laws such as the EU’s Corporate Sustainability Reporting and Due Diligence Directives. For civil society organizations and solution providers (be they for pro昀椀t or not), this means asking whether their business model or mission statement and objectives, can accommodate or should be revised to promote a more open model whereby one actor can share (more of) the data that is held, for greater impact. This may start with funders and investors, who should make some form of data sharing for the public good a more common condition of funding. Private companies should also support business models for more “open” data solutions: paying for a solution should not exclude the possibility that some of the data collected or generated be shared as a common public good. Build Trust in the System The lack of trust between actors as to how shared data might be used, and the fear that it may be used against them, was a recurring theme among interviewed stakeholders, especially from the private sector and civil society. Trust and data sharing will be di昀케cult to foster between organizations whose interests and roles are opposed (e.g. law enforcement and companies at risk of non-compliance, or companies and judicially active NGOs). But more trust could be built, and in turn, more data shared, if: • Companies openly recognize when forced labor is likely to take place somewhere in their supply chain and focus their due diligence and reporting efforts on 昀椀nding it and taking appropriate actions when they do, rather than trying to disprove that the risk exists. • Governments adopt policies and laws that reward transparency by companies about forced labor risks in their operations or supply chains, provided this is combined with demonstratable commitments and actions to prevent and mitigate these risks. Business stakeholders interviewed a part of this project shared that the effect of the US’s import ban on products suspected of being made with forced labor was to drive companies to divert 5353

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION their supply chains away from high-risk locations, rather than engage with suppliers to mitigate risks and improve conditions on the ground. Laws that punish admission by a company that forced labor takes place, or is likely to take place, in their supply chains are counterproductive and are unlikely to address the root causes of exploitation and modern slavery. • Governments lead by example, by proactively sharing valuable data they hold with trusted partners. While it is true that government data may be highly sensitive, as explained in the preceding section of this report, there are data privacy-compliant ways to share the substantive value of data with other stakeholders. • All actors adopt more nuanced approaches to “zero-tolerance” statements. Strong commitment to preventing forced labor is important, but interviewed stakeholders reported that “zero-tolerance” positions can dissuade certain actors from being forthcoming with information about forced labor risks in their operations or supply chains, out of fear of commercial, reputational, or legal consequences. For example, a supplier might fear losing their customer’s business, or a company might fear legal action or adverse media from an NGO. Strong commitments to prevent forced labor should be paired with a commitment to engage with and support any supplier, partner, or other stakeholder who proactively shares evidence of forced labor risk with the intent to enable collaboration in addressing the issue. • All actors seek to demonstrate reciprocity. Any actor receiving valuable data from the ecosystem should reciprocate by sharing valuable data in return. This could include disclosing what actions they have taken that were enabled by the data that was shared with them, to demonstrate the value of sharing such data with them. This is particularly relevant for companies and governments relying on data produced by NGOs, for whom evidence of impact is an important incentive to collect and share data and engage in trusting and collaborative relationships. Emerging due diligence and reporting regulations such as the EU CSRD and CSDDD require companies to meaningfully engage with “affected stakeholders” (or organizations representing their interests). This will require a step change in how companies engage with NGOs and other civil society organizations to inform their due diligence. Companies and civil society organizations should put trust, reciprocity, and data-sharing at the core of these new relationships. Know Your Role Distrust and frustration among actors in the supply chain data ecosystem are often the result of misaligned expectations as to what other actors should or should not be doing. As outlined at the beginning of this report, each actor has a role within the ecosystem and should take responsibility for what is expected of that role, without overstepping into roles best played by other actors, or shifting their responsibilities to other actors. For instance, businesses and solution providers should not seek to replace the important role of grassroots organizations in collecting and interpreting data through direct engagement with workers and local communities. 5454

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION Rather than developing separate mechanisms, it may be more effective to seek ways to support those activities and to lower the costs and burden of making that data more easily accessible to more stakeholders. Relatedly, the importance of context in interpreting raw data to convert it into actionable intelligence means such interpretation is likely to be more effective when done “close” to where the data has been collected. For example, data about potential child labor needs to be understood against local legal, social, and cultural norms. If the goal is to enable better decision-making and actions, it may be more effective to share insights and intelligence (i.e. interpreted data) rather than raw data, provided that any interpretation or aggregation has been done by quali昀椀ed actors. 6.2 Looking Ahead: Collaboration Opportunities Considering the principles and recommendations set out in the previous section, Tech Against Tra昀케cking is of the view that an effective federated supply chain data ecosystem will require: 1 Standardized and interoperable data on potential and actual forced labor Scalable, cost-effective, and accessible technologies that enable a federated 2 ecosystem of purposeful data 昀氀ow between organizations, including smaller and less technologically advanced actors Further dialogue between the public sector and 3 corporate sector (as well as civil society) to identify how governments can better support an effective supply chain data ecosystem 5555

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION 1 Standardized and Interoperable Data Policies, practices, and systems to collect and gather forced labor risk management data are generally not aligned across the ecosystem. Many seek to align with the ILO’s forced labor indicators, but there remains a signi昀椀cant diversity of interpretations and data collection protocols that undermine the ease with which data collected by different parties can be shared and aggregated. Tech Against Tra昀케cking wishes to support the standardization of forced labor risk data, building on the ILO’s forced labor indicators and working in partnership with key standard- setting bodies like the ILO and the IOM, and leading providers of supply chain risk data (e.g. supplier audit services, worker voice tools). Such standardization efforts should aim for alignment and consensus among actors across the supply chain data ecosystem on qualitative and quantitative data points to seek when looking to identify forced labor risk. For example, drawing on the experience of actors involved in collecting data to identify actual cases of forced labor (e.g. auditors, police o昀케cers, labor inspectors) to de昀椀ne practical questions, and quantitative or qualitative responses, that are most useful to identify evidence of forced labor. 2 Scalable Technologies for a Federated Ecosystem Many initiatives are already exploring ways in which some of the legal and/or behavioral barriers to data-sharing (e.g. linked to con昀椀dentiality, competition, or privacy concerns) can be overcome through more “federated” or decentralized data spaces, where different actors can share some of their data, and contribute its value to the broader ecosystem while retaining control of potentially sensitive elements. A future supply chain data ecosystem based on such federated data spaces will require a scalable model for such data architecture that is affordable and accessible beyond large multinational companies and their supply chain risk management solution providers. Tech Against Tra昀케cking hopes to leverage its network of technology experts committed to demonstrating how technology can help anti-tra昀케cking and forced labor efforts to advance efforts to understand how more federated data spaces could be enabled across the supply chain data ecosystem. This should also consider how more purposeful data 昀氀ows can be enabled end-to-end, from raw data collection from frontline workers to various forms of data sharing between corporates and other organizations. 3 Public Sector Dialogue This project rea昀케rmed the signi昀椀cant siloes (or one-way data 昀氀ows) between the public sector and the rest of the supply chain data ecosystem, and the need for better mutual understanding between governments and businesses as to what data is available to identify risks of forced labor and inform effective prevention and mitigation strategies. 5656

EXECUTIVE SUMMARY 1. INTRODUCTION 2. SUPPLY CHAIN DATA 3. KEY ELEMENTS 4. CURRENT LANDSCAPE 5. EXAMPLES 6. PRINCIPLES TO ACTION Tech Against Tra昀케cking wishes to build on its engagement with policymakers in the preparation of this report, and the multi-stakeholder discussions at its November 2023 Summit, to foster more dialogue between governments, civil society, and the corporate sector on how governments can support a more effective supply chain data ecosystem. Areas of focus for discussion could include: • facilitating the greater visibility of, and access to, government-held data • designing policies and laws that promote transparency and data sharing • developing public databases that meet the needs and expectations of the broader data ecosystem 5757