The World Bank’s Environmental and Social Safeguards and the evolution of global order

Abstract

This article analyses the World Bank’s environmental and social Safeguards against the backdrop of changing paradigms of global legal order. In January 2017, a new ‘Environmental and Social Framework’ (ESF) entered into force and replaced older ‘Safeguard Policies’ that had incrementally emerged since the 1980s in response to harmful impacts of investment projects financed by the Bank. The Safeguards reform epitomizes the changing structures and geopolitical shifts that shape international law in the twenty-first century and provides a fascinating looking glass on the evolution of global order since the end of the cold war. In this perspective, we see the first generation of Safeguards, introduced since the late 1980s, as an element of incremental legalization in the emerging global governance regime, a regime characterized by unipolar multilateralism and geopolitical dominance of ‘the West’. The 2016 reform not only reflects the increased politicization of global governance by civil society but also the emergence of a more competitive multilateralism, characterized by counter-institutionalization on the part of emerging powers like China. A comparison of the old and new Safeguards thus allows us to analyse different forms of contestation and resulting normative evolution in the key area of global governance of development and finance.

1. Introduction: The Safeguards as a lens on the evolution of global order

In January 2017, the World Bank’s new ESF entered into force. It replaced the older ‘Safeguard Policies’ that had incrementally emerged since the 1980s in response to harmful impacts of Bank-financed investment projects. The new ESF requires the Bank and its borrowers to assess and manage environmental and social risks, to inform and consult with stakeholders, and to compensate certain project-affected people. These new rules are significant in their own right, as they affect investment projects in over hundred countries with an annual lending volume of US $45 billion. ¹ But beyond its economic impact, the safeguard reform has a much wider political and legal significance. Since the 1990s the Safeguards have become a global normative standard. They were emulated by multilateral development banks (MDBs), by the private sector, and they diffused within domestic legal systems of borrowing countries. Moreover, the Safeguards reform currently is one of the very few major multilateral lawmaking projects that attract quasi-universal support under new geopolitical conditions: Western donor countries including the US, emerging powers like China and India, as well as developing nations, all agreed to the reform package in 2016 and remain committed to it until today. The universal embracement is all the more surprising as the ESF covers a wide range of controversial policy issues, ranging from climate change, labour standards and indigenous peoples, to human rights and accountability of international institutions. Hence, the Bank’s new Safeguards represent a multi-issue compromise among old and new powers, which was achieved not only in the shadow of new competitors like the Asian Infrastructure and Investment Bank (AIIB), but also with significant civil society participation at the national and global levels.

In sum, the Safeguards reform epitomizes the changing structures and geopolitical shifts that shape international law in the twenty-first century. We, thus, argue that the Safeguards provide a fascinating looking glass on the evolution of global order since the end of the Cold War. In this perspective, we see the first generation of Safeguards, introduced since the late 1980s, as an element of incremental legalization in the emerging global governance regime, a regime characterized by unipolar multilateralism and geopolitical dominance of ‘the West’. If we turn our looking glass to the 2016 reform, then the Safeguards compromise appears as a multifaceted reflection of processes of contestation and change in international law since the 1990s: The reform not only reflects the increased politicization of global governance by civil society but also the emergence of a more competitive multilateralism, characterized by counter-institutionalization on the part of emerging powers like China. A comparison of the old and new Safeguards thus allows us to analyse different forms of contestation and resulting normative evolution in the key area of global governance of development and finance.

In this comparative analysis, we adopt two different perspectives: a geopolitical one, focused on the wider political and economic context as well as on the processes inside and outside the Bank when the Safeguards were first introduced and then reformed; and a legal perspective, analysing the content of the Safeguards along three dimensions that each reflect foundational but shifting concepts of international law: first, the relationship between Bank and member state law, shaped by the concepts of state sovereignty on the one hand and authority of international institutions on the other; second, the role of individuals, which is circumscribed by individual rights, especially human rights, and accountability for violations of such rights; and third, the thematic coverage of the Safeguards, which defines the relationship of ‘development’ to other international regimes and thus represents an element of either fragmentation or integration of international law across policy areas and international institutions.

To elaborate our argument and these perspectives, the article proceeds in three steps: Section 2 analyses the evolution of the Safeguards in the 1990s, outlining the geopolitical context and their legal content along the three analytical dimensions. Section 3 turns to the 2016 reform, investigating the profoundly changed context and briefly outlining the process and outcome of the reform. Section 4 then takes a closer look at the substance of the ESF, returning again to the three dimensions. Section 5 concludes.

2. First generation Safeguards – in the first generation of global governance

2.1 Context and process: World Bank and the incremental introduction of safeguards in the 1990s

The Safeguard policies emerged in the late 1980s and early 1990s as part of a transformation of the Bank from lender to norm-setter, a transformation that profoundly increased its role in the global order and its impact on borrowing countries. At its inception, the World Bank was conceived as a lending institution and its founders did not assign a normative role to it. Initially, the Bank, thus, operated chiefly as a lender of funding for infrastructure projects. It paid out loans to borrowing countries, which then implemented the project according to their own laws and standards. The loan agreements essentially required the Bank to pay out the loan and borrowers to repay it. ² This changed in the 1980s. This decade saw the introduction of the widely-known and much criticized structural adjustment programmes, based on a new legal instrument for budget support from the Bank. ³ It was also marked by another, equally important, normative development: The Bank started to enact more robust environmental and social standards for its investment projects, which became known as ‘Safeguards’. These Safeguards grew incrementally in response to an increasing perception of serious deficiencies in concrete projects: When a dam is constructed or a road is built, land must be acquired and possibly expropriated, communities are displaced, businesses lose customers; waters are polluted, trees cut, natural habitats lost; and occasions for corruption and waste generated. ⁴ With each ‘problem project’, awareness of specific risks for people and the environment grew. As a result, incrementalism characterized the legalization process at the World Bank.

This legalization process can be attributed to several factors and occurred at a historical moment characterized simultaneously by hubris and crisis. ⁵ Hubris inspired claims about the end of history: Western powers, in particular the US, dominated international relations. In the decade after the end of the Cold War, the global order was shaped by unipolar multilateralism and the emergence of a system described today as global governance. ⁶ The ‘West’ led by the US dominated economically, politically, and normatively. Economically, Latin America and South-East Asia were only slowly recovering from heavy debt crises. China, Brazil, and India were opening their economies but growth rates had yet to take off. World Bank borrowers remained dependent on public capital flows and the private capital these flows leveraged. ⁷ Politically, the US and their allies dominated international relations in the absence of their former rivals in the Eastern bloc. Normatively, American and European legal systems and ideas claimed superiority and influenced international lawmaking and bilateral law reform projects around the world. ⁸ Liberal internationalism favoured the foundation of new international institutions and a strengthening of existing ones.

This overall situation contributed to the consolidation of a global governance system characterized by a set of normative assumptions. These assumptions included the belief in the possibility of international authority beyond the state and a rudimentary notion of a global common good. These ideas offered a normative justification for the increased exercise of public authority by international institutions. These justifications obscured, for quite some time, the extent to which the existing global governance system continued to institutionalize inequality. ⁹ At the World Bank, this inequality had always taken the form of weighted voting, a significant departure from sovereign equality: The US and other OECD countries continued to enjoy voting rights intended to represent their share in the world economy but not automatically adjusted to the changing economic realities. This was even exacerbated by the cyclical process of International Development Association (IDA) replenishments, i.e., the drives for additional funding from wealthy donor states. Only these donors participated in the pledging conferences and thus influenced the policy orientation of the Bank even more directly without being counterbalanced by any borrowing or recipient country participation. ¹⁰

The hubris of the 1990s, however, also contrasted with a moment of crisis that began to affect international institutions in general and the World Bank in particular. ¹¹ The Bank faced scathing criticism from different constituencies: For one, civil society organizations (CSOs) became increasingly aware of the problematic effects of World Bank activities. Structural adjustment programmes elicited fierce critique, especially in Latin America. The more traditional Bank-financed investment projects also came under fire in the late 1980s and early 1990s. The Bank’s problems are illustrated by the often-cited example of the Narmada dam in India: Its construction displaced more than 140,000 people, affected the livelihoods of many more, and sparked worldwide controversy. National resettlement rules and their implementation proved inadequate, and the Bank’s own standards and implementing practices were found to be insufficient. ¹² CSOs began to organize ever more vocal protests. This was true for protests in the Global South, but there was also increasing awareness of lacking social and environmental protection in the Global North. In particular, CSOs in the US were able to influence Bank policy indirectly through the US Congress’ power of the purse. ¹³ In addition to these external pressures, the crisis also came from within. The Bank’s management was becoming increasingly aware of the fact that many of its projects had become inefficient, with substantial time and cost overruns due to deficient internal organizational and procedural structures. The critical Wapenhans report of 1991 testified to these deficits and documented severe governance and compliance problems within the Bank. Taken together, these problems posed a serious legitimacy crisis for the Bank.

As a response to this crisis, the Bank adopted a strategy of increasing legalization and juridification as part of a larger strategy of technocratic legitimation. It decided to overhaul its policy framework and give it a more formalized – and legal – structure. Internal legal acts were classified, new standards were enacted as ‘Operational Policies’ and ‘Bank Procedures’ (OP/BPs), processes for enactment and amendment were standardized, and publicity requirements introduced. ¹⁴ Over time, a total of 11 OPs and BPs, collectively known as Safeguards, provided protections against particular risks (e.g., resettlement) and for particular groups (e.g., indigenous people) or resources (e.g., forests, natural habitats). Besides, an independent review mechanism, the Inspection Panel, was established to hear non-compliance claims by project affected individuals, as outlined below.

The concrete processes in which these early Safeguards were adopted reflected the larger political context. In the already unequal institutional structure of the World Bank, its Articles of Agreement do not clearly set out the procedure for secondary lawmaking. ¹⁵ The process, thus, evolved over time, reflecting a balance between Bank management under the President, on the one hand, and the Board of Executive Directors representing member states, on the other. ¹⁶ In this system, the Board would lay out the general principles of the safeguard policy to be adopted and management would prepare the actual text of the policy. While this balance shifted back and forth over time, US influence over the process remained a constant: The US appointed the President of the Bank and held the largest share of votes in the Board. Largely excluded from this process were those member states that did not hold larger voting blocks, especially all borrowing and recipient countries, even if they were important borrowers and therefore directly affected by new Safeguards (India and China for example). However, their position, or rather opposition, towards the Safeguards did not stop the process.

Equally excluded were external actors and civil society ¹⁷ – with one significant exception: Domestic US interests were able to influence Bank policy indirectly through the US Congress’ power of the purse. The Bank depended on regular replenishments to refinance its interest-free credit by its concessional lending arm, the IDA. Appropriations by Congress, thus, were a leverage for interests represented in it, which was used extensively since the 1970s. ¹⁸ An illustration of US influence is the introduction of environmental impact assessments (EIA) into Bank Safeguards. During the early 1990s US CSOs pressured Congress, which resulted in the so-called ‘Pelosi’ amendment in the early 1990s. This amendment to an appropriations bill, proposed by Democratic Representative Nancy Pelosi, made US contributions to IDA dependent on the introduction of environmental impact assessments into Bank procedures. The Bank duly enacted such a Safeguard policy (OP 4.01), which largely followed the US model mandated by the federal National Environmental Policy Act (NEPA) of 1969. ¹⁹ In sum, the Safeguards process was, thus, driven (for different reasons each) by the Bank management, Western states, and bottom-up pressures in concrete projects and was enabled by the unipolar structure of the global order at the time.

2.2 Nature and content: Original Safeguards as a reflection of their unipolar geopolitical context

The political dominance and normative pull of the ‘West’ also shaped the nature and substantive content of the Safeguard policies. They are, thus, another example of the extent to which ‘Western’ states acted as norm-shapers and the rest as norm-takers. This can be demonstrated along the three dimensions identified in the introduction, each of which points to a structural element of the international legal system, viz sovereignty/international authority, the role of the individual, and thematic fragmentation or integration. These dimensions not only represent features of international law in general but also principles of the law of development co-operation that has emerged over recent decades. ²⁰

2.2.1 Relationship between Bank and member states: From sovereignty to shared responsibility

The relationship between the Bank and its member states is structured by the principle of sovereignty, on the one hand, and the public authority exercised by the Bank over its borrowers, on the other. In the context of development co-operation, sovereignty can be reframed as a principle of collective autonomy, which supports claims for ownership and discretion of member states in their interplay with the Bank. ²¹

Before the Safeguards were adopted, international and domestic law remained relatively separated in Bank projects, which were implemented by borrowers according to their own domestic law. Beyond the negotiation of the loan agreement, collective autonomy of borrowers was not limited, at least not legally. The Safeguards changed this and increasingly imposed requirements on how member states were to design and implement projects: EIAs were to be conducted, indigenous peoples to be consulted, resettlement plans to be drafted, project affected people to be compensated etc. The scope for sovereignty and collective autonomy of member states shrank. The multilevel interplay between international and national law was increasingly determined in practice by the Bank’s internal law, at the expense of domestic constitutional law.

To understand this shift one needs to trace the particular legal techniques through which the Safeguards become binding on borrower states. ²² As OP/BPs, Safeguards have internal legal effects only and are directly binding only on World Bank staff (enforced through bureaucratic hierarchy, performance assessments, promotions and disciplinary measures, and through the Inspection Panel). However, they also become binding on borrowing member states in two steps: First, if a government wants to receive a loan or grant from the Bank, it is required to prepare a funding application in line with the procedural, environmental and social requirements stipulated in the Safeguards; they, thus, function as an ex-ante conditionality. Second, once the loan agreement is concluded, it incorporates the Safeguards and the concrete environmental and social measures agreed upon during project preparation. Loan agreements concluded with member states are considered treaties under international law. ²³ By virtue of this bilateral treaty, the borrower is under an international legal obligation to comply with applicable Safeguards. In case of non-compliance, the Bank is entitled to withhold or call in the loan.

As a result, borrowers implement projects under two sets of norms: Their own domestic law, and international obligations towards the Bank. If national law does not meet Bank standards, the borrower faces the choice of forgoing the loan or raising its national standards to meet higher Bank standards, at least for the project at hand. Under the economic and geopolitical conditions prevailing until the 2000s, most borrowers accepted, if only grudgingly, this trade-off between cheap loans and collective autonomy. Consequently, the conditionalities and the network of hundreds of loan agreements the Bank concluded with a total of 146 countries over time effectively multilateralized the safeguard regime. ²⁴ The Bank had turned from lender to international lawmaker. ²⁵

This increased normative reach of the Bank was controversial for two reasons: For one, borrowers not only had to accept a new set of Bank norms, they also had little say in the development of these international norms in the first place. ²⁶ A second aspect was the fact that the Safeguards applied uniformly across all borrowers, irrespective of the quality of their domestic legal system and their administrative capacity: Mali and India, Suriname, and China were treated alike, at least as a matter of law. While critics suggested that the Bank did, in fact, differentiate between large and small borrowers in practice, such differentiation had no legal basis in the initial Safeguards. In 2005, the Bank did introduce a policy on the ‘use of country systems’ ²⁷ which, in theory, allowed replacing parts of the Safeguards with domestic law. However, the policy imposed such strict requirements that it was rarely applied in practice. A main obstacle was the requirement that the national system be legally equivalent to the Safeguards, which was hardly ever the case. At least on paper, the Bank, thus, subjected not only itself to relatively stringent requirements but also all of its borrowers and recipients.

This, however, did not necessarily mean that the Safeguards actually protected the environment and project-affected people effectively in practice. In fact, the original Safeguards suffered from serious implementation deficits: Although they were integrated in loan agreements and states formulated project-specific plans to mitigate social and environmental risks, the implementation of these obligations and plans was entirely entrusted to the member states. In contrast to other donors like the UN Development Program, which can execute its own projects, the Bank only finances projects. Execution by national authorities is only a modality of implementation. While the Bank could relatively effectively police borrower compliance with ex ante requirements for project preparation before the loan was approved, the tables turned once the loan agreement was signed and money started flowing: Supervising the implementation of hundreds of projects worldwide proved difficult, and institutional incentives militated against actually withholding or calling in loans once implementation problems occurred. This problem of the original Safeguards was known as ‘frontloading’, whereas there was a lack of ‘downstream focus’. ²⁸ It was, thus, in the implementation phase that borrowing countries retained, and regained, some of the collective autonomy and discretion they had lost to the Bank in earlier phases of the project. Despite this informal pressure valve, the Safeguards did change the overall equation between the sovereign powers of member states and the Bank’s international legal authority.

2.2.2 Role of the individual: Rights and quasi-judicial accountability

The original Safeguards advanced the normative project of Western donors in another significant way: They transformed individuals from passive beneficiaries of development projects into active rights holders and legal subjects under international law. This transformation was in line with the increasing role of human rights in the international legal order and is encapsulated by another principle of development co-operation law, that of individual autonomy. ²⁹ Unlike human rights law, however, this transformation was based on international institutional law, and it created relatively concrete individual entitlements: These were mainly procedural rights for affected people to participate and be heard in the development of a project and its implementation. Besides, they established substantive guarantees, especially for compensating resettled residents or indigenous peoples. ³⁰ These entitlements can be seen as an incipient regime of ‘simple’ individual rights beyond the framework of human rights law or more broadly as part of an emerging ‘global administrative law’. ³¹

The Safeguards were accompanied by a mechanism for their enforcement: In 1993, the Bank’s Board of Executive Directors established the Inspection Panel, mandated to hear non-compliance complaints from project affected people. After the Wapenhans report had detailed enforcement gaps in the Bank’s existing standards in the Narmada project, several types of accountability mechanisms were discussed. While European countries initially favoured a strengthened internal evaluation function, they eventually followed an American proposal for an independent panel hearing individual complaints. ³² The Inspection Panel subsequently grew into a quasi-judicial review institution effectuating the Bank’s environmental and social accountability towards external individuals. ³³ It also provided an avenue to make the Safeguards more effective in the implementation phase by mobilizing individuals with a direct interest in compliance. This type of decentralized ‘fire alarm’ control, as opposed to ‘police patrol’ enforcement by Bank management, can be seen as another element of US influence and catered to the interest of major donor countries. ³⁴

While human rights are not explicitly mentioned in the Safeguards, project-affected people and CSOs supporting them quickly turned to rights language in the complaints they brought to the Panel. And indeed, the Panel relied, explicitly or implicitly, on human rights when interpreting safeguard standards on resettlement, indigenous peoples, and public participation. ³⁵ Even though civil and political rights remained an anathema for the Bank for a long time, the Safeguards and the Inspection Panel signaled that economic and social rights became increasingly accepted by the Bank. Overall, the first generation of Safeguards mirrored the increased role of the individual in international law, leveraging the Bank’s economic and legal authority vis-à-vis member states to empower individuals while at the same time using rights and review to place some checks on that very authority of the Bank.

2.2.3 Thematic coverage: Fragmentation and integration of international legal regimes

One of the traditional features (and weaknesses) of the international system that also shaped early forms of global governance is the fragmentation of policy fields. Each policy area (security, environment, trade, development) is regulated in separate regimes and by separate institutions, and there is no place akin to the plenary of a national parliament to integrate these functionally differentiated regimes. With the proliferation of international law, institutions, courts, and tribunals in the 1990s, this fragmentation became an increasing concern for international lawyers. ³⁶ In the area of development, the general issue of fragmentation vs. integration translated into policy debates about the notion of development – should the focus be on economic, human or sustainable development? – and into legal questions as to how broadly or narrowly the mandate and objectives of development institutions should be interpreted: should they be understood as aiming at economic growth, or should the World Bank and its peers pursue a more holistic understanding of development that includes environmental and social concerns, good governance and human rights?

At a general level, these questions correspond to two other principles of development co-operation law: first, the principle of development, which enshrines a more holistic notion of development focused on poverty reduction, and second, the principle of coherence, which requires donors and international institutions to co-ordinate their activities. ³⁷ This is reflected at the institutional level of the World Bank: the Articles of Agreement mandate that only economic considerations be relevant for projects but this provision has been interpreted broadly to allow for activities in agriculture, education, good governance and the like. While this ‘mission creep’ was well under way in the 1990s, the Safeguards were significant in that they legalized and codified the social and environmental considerations relevant in project appraisal and approval. From this perspective the Safeguards can, thus, be seen a legal attempt at overcoming fragmentation and moving towards greater integration.

At the same time, this attempt at integration faced the Bank with significant questions: Which non-economic issues should be included in Safeguards? And should the Bank develop its own standards, or should it rely on norms produced by other international institutions, such as the International Labour Organization or UN human rights organs? Essentially, the Bank opted for a cautious incremental approach that preserved its institutional autonomy. The original Safeguards made a first tentative step to integrate mostly environmental concerns and some social issues, often proceduralizing them by requiring impact assessments or by granting participation rights. In terms of social and economic rights, the Safeguard policy on resettlement included substantive protections for (some) property rights, whereas labour rights or health issues remained largely uncovered. Likewise, civil and political rights were clearly outside the range of the Safeguards, and human rights language was avoided entirely. This limited scope is partly explained by the incremental process in which the Safeguards were introduced; in this process, they emerged mostly in isolation of wider international law regimes. A second element was the absence of inter-institutional co-ordination: In its lawmaking process, the Bank rarely consulted systematically with other international institutions, nor did it follow the normative views of other international institutions, courts or tribunals in the interpretation of its Safeguards. In fact, the reverse became true: With its increasingly dense and sophisticated norm production, the Bank itself took on a new leadership role as a global norm-setter, influencing legal developments well beyond the confines of its own projects.

2.3 Diffusion: The Bank as global norm-setter

To fully understand the significance of the Safeguards, one needs to look beyond projects and take into account the epistemic authority the Bank came to wield in national and international development discourses and policy making. ³⁸ With the Safeguards, the Bank also built up its social and environmental expertise and enhanced its reputation for legal expertise in these fields. An internal Bank evaluation concluded in 2010 that the main benefit of the Safeguards for the World Bank was the ‘recognition of its leadership role in setting and promoting benchmarks for environmentally and socially sustainable projects’. ³⁹ Indeed, at the time of their initial enactment, the Bank’s Safeguards on environmental assessments, resettlement, indigenous peoples etc. were considered the ‘gold standard’ in development and project finance.

As a result, the Safeguards diffused horizontally into the laws and practices of other MDBs and private banks and vertically into domestic legal orders beyond individual projects. ⁴⁰ When other MDBs encountered similar problems in their projects, they often copied the Bank’s Safeguards. Today, the Inter-American Development Bank, the Asian Development Bank, the African Development Bank, and the European Bank for Reconstruction and Development have adopted not only environmental and social standards modeled upon the World Bank’s safeguard polices but also Inspection Panel-type accountability mechanisms. ⁴¹ The OECD export credit financing conditions include Bank-type social assessments. ⁴² Through the Bank’s private sector arm, the International Finance Corporation (IFC), they also influenced the private banking industry, whose ‘Equator Principles on Sustainable Finance’ are largely modelled upon the IFC’s environmental and social performance standards. ⁴³

Member states also modeled their own laws on Bank standards and incorporated their experience with implementing Bank projects in policy processes. While systematic empirical studies of this aspect of the Bank’s epistemic authority are still missing, the causal mechanisms developed in the diffusion literature, namely coercion, competition, learning, and emulation provide a plausible explanation for observed convergences in national legislation in borrowing countries. ⁴⁴ One example is the spread of environmental impact assessments: ⁴⁵ They were adopted in jurisdictions like Mexico, Brazil, and South Asia. ⁴⁶ India introduced a legal basis for environmental impact assessments with the Environment Protection Act 1986. ⁴⁷ A second example is assessments to manage social risks, mainly in resettlement. Two Safeguard Policies mandate specific social assessments. ⁴⁸ In China, the 2002 Guidelines of the National Development and Reform Commission require social assessment in feasibility studies and development investments, and the practice of social impact assessments has expanded considerably since then. ⁴⁹ In 2013, India enacted the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act (LAA), whose overall structure and individual provisions on social impact assessments bear strong resemblance to the Bank’s OP 4.12 on Involuntary Resettlement. ⁵⁰ While the Bank did not play a prominent role in the drafting and there is no mono-causal explanation in either of these cases, it is quite plausible that experience gained in Bank projects as well as mechanisms of emulation and learning played a role.

In any event, it is safe to say that the Safeguards had become a globally diffused normative model for socially and environmentally sound development. Hence, they not only provided a looking glass on the evolution of global order but had also come to influence to some extent that very legal order.

3. The 2016 Safeguards reform: Context, process and outcome

3.1 The changing context: From unipolar to competitive multilateralism and multiple contestations

When the Bank embarked on its Safeguard reform process in 2012, it was motivated by several considerations. In part, the reform responded to various criticisms that the Safeguards had attracted over the years. At the same time, it was a reaction to the considerable changes in the geopolitical and economic context, which put the Bank’s entire business model as well as its international authority into question. To some extent, the Safeguards had also become a victim of their own success: Other MDGs not only copied the Bank’s standards, but also developed and improved them, so that the Bank’s old system had actually fallen behind, especially in comparison to the IFC. ⁵¹

While the Safeguards had considerably increased the normative reach and impact of the Bank, they had in equal measure triggered contestation. Instead of resolving questions about its legitimacy and satisfying critics, the Safeguards had led to new kinds of criticisms and concerns. ⁵² Rising powers as well as weaker states criticized the Safeguards as bureaucratic burden and inappropriate conditionality from the outset. They were considered as too detailed a corset for borrowers and as imposing specific normative values on them, for instance the fields of indigenous peoples or resettlement. The Indian government, for example, by no means welcomed World Bank standards such as those on resettlement. Like many other governments, it opposed in particular compensation for informal occupants, often considered illegal squatters under national law. ⁵³ Like structural adjustment, the Safeguards came to be seen as a conditionality. Within the Bank and in concrete projects on the ground, states pushed back or adopted strategies of selective (non-)compliance. ⁵⁴ The Inspection Panel, designed as institutional guardian of the policies, faced criticism for its role in enforcing the Safeguards.

At the same time, the rise of China, India and other emerging economies initiated a gradual transition from a unipolar to a multipolar global system. ⁵⁵ Economic growth especially in the BRICS countries, but also in some other transition states, has increased their share of the global economy and is set to surpass that of the OECD countries in the foreseeable future. ⁵⁶ Private capital flows to emerging markets have soared and let flows of Official Development Assistance (ODA) pale in comparison. ⁵⁷ Emerging powers increasingly mobilize domestic capital for investments at home: In 2010, the Brazilian Development Bank (BNDES) disbursed the equivalent of roughly US$100 billion, surpassing the World Bank’s worldwide disbursements of US$40 billion by far. ⁵⁸ Even as they integrate into the global economy, China, Brazil and India promote a variation of the ‘developmental state’ as an alternative to the Washington Consensus-type development model. ⁵⁹ China has been at the forefront of this development. ⁶⁰ Emerging economies have become capital exporters and development donors themselves. In the relations with recipients, they at least rhetorically emphasize sovereignty and non-intervention, offering an alternative to established donors’ conditionalities and infusing competitive pressures into the system. ⁶¹ Indeed, they assert an alternative vision for wider international law that is more sovereignty oriented and less individual-centred, as epitomized by the Declaration of the Russian Federation and the People’s Republic of China on the Promotion of International Law, signed in Beijing on 25 June 2016. ⁶² The new powers want to move from the position of norm takers to that of norm shapers. ⁶³

While the rise of emerging powers has thus ended the unipolar moment of the 1990s, the new multipolar system does not end multilateralism but rather transforms its nature. The BRICS countries pursue a double strategy: They combine engagement in established institutions with the creation of new competing outfits, trying to gain voice within the existing system but also contesting it from without. ⁶⁴ This indicates a transition towards a more contested or competitive multilateralism. ⁶⁵ China exemplifies this development. On the hand, it increasingly engages with existing multilateral institutions. Its own ecological problems have pushed it to embrace environmentalism in regional trade agreements, and it emerged as a champion of the 2015 Paris Agreement after the US’s exit. ⁶⁶ It has also stepped up its engagement with the World Bank, became an IDA donor in 2009, and increased its IDA contribution in the 2012 replenishment round. ⁶⁷

Despite this increased engagement, the Chinese voting share in the Bank has remained largely unchanged. Resistance from developed countries, country classifications and conservative indicators for economic growth impeded substantial shifts in voting power, and US Congress refused to ratify the Bank’s 2010 ‘voice’ reform altogether until late 2016. ⁶⁸ At the IMF, China did not fare much better: Even though the IMF’s voice reform went ahead, China’s voting share reached only 3.81 per cent, a fraction of its 12.4 per cent share in world GDP. ⁶⁹ In fact, from 2009 to 2014 developed countries actually gained voting shares in the Bretton Woods institutions relative to their share of world GDP. ⁷⁰ When a new President was set to be appointed at the Bank in 2012, the US government again imposed its own candidate, Jim Yong Kim, against increasing resistance from recipient countries.

Against this background, China started to pursue an alternative strategy of contestation through counter-institutionalization. This strategy aims at creating competing multilateral institutions, which not only help bypass the established institutions but also put competitive pressure on them. The Chinese government has led the establishment of the Asian Infrastructure Development Bank (AIIB) and of the New Development Bank (NDB) by the BRICS states. ⁷¹ China holds the largest share in the AIIB and is on a par with the other BRICS members of the NDB, hosts both headquarters (in Beijing and Shanghai, respectively), and the Chinese Jin Liqun became the first AIIB president (whereas the NDB is headed by an Indian national). Attempts by the US government to prevent its allies from joining the AIIB failed embarrassingly. ⁷² The AIIB and the NDB offer not only a competing source of capital investment but also an alternative development model based, at least rhetorically, on a stronger role for sovereignty and the state. ⁷³ This puts the Bank’s business model under economic pressure, given the importance of rising powers as clients, and also challenges the Bank’s normative role. This double challenge posed a dilemma for the Bank and for its Safeguards reform: Can the Bank compete as an efficient lender while at the same time retaining its leadership role as a global norm-setter?

The task of the Safeguards reformers was further complicated by the fact that counter-institutionalization by rising powers was not the only form of contestation the Bank experienced. The increased normative and epistemic authority of the Bank also faced criticism and protest from a range of societal actors. CSOs continued the critical observation of the Bank, focusing not only on its projects but also on its normative role and knowledge activities. ⁷⁴ Trade unions criticized the Bank’s approach to labour rights, embodied in its influential Doing Business indicators. ⁷⁵ Environmentalists levelled increasing criticism at the Bank’s approach to climate change and brought cases against coal-fired power plants to the Inspection Panel. ⁷⁶ Human rights activists mounted campaigns against displacement in Bank-financed resettlement projects. ⁷⁷ Overall, this form of societal contestation led to an increasing politicization of Bank policy in general and of the Safeguards reform in particular. ⁷⁸

3.2 The 2016 reform process and objectives: From incrementalism to systematic overhaul and organized multi-stakeholderism

In light of this context, Bank management chose to frame the objectives of the Safeguards reform as ‘enhancing development effectiveness’ in project financing. The stated objective of the new generation of Safeguards was to define aims and responsibilities more clearly, to develop an integrated and uniform normative structure, to accommodate the different needs of the diverse borrowers, and to cover additional environmental and social risks, for instance, in the domains of climate and labour. ⁷⁹

The reform process lasted four years and involved not only intergovernmental negotiations but also three public consultation rounds. In these hearings, representatives of governments, civil society, indigenous groups, academia, and other stakeholders were consulted on different aspects of the new Safeguards (objective and structure, the text of the first draft, and feasibility). They involved around 8,000 stakeholders in 64 member states. ⁸⁰ This makes the Safeguards reform the Bank’s most inclusive and transparent legislative process thus far and it is read by some observers as an instance of global deliberative democracy. ⁸¹ The organized multi-stakeholder process is also a far cry from the incremental approach during the 1990s: The new ESF is the outcome of a single legislative process and systematically covers all thematic areas and structural questions considered relevant.

This participatory approach, however, also made consensus building difficult. The documentation of the consultation feedback from representatives of various countries offers insights into the contrasting positions taken during the negotiations. ⁸² Generally speaking, civil society organizations and donor states urged that protections standards be preserved or even raised. ‘No dilution’ became their rallying cry. In contrast, borrower states sought greater flexibility in the implementation of the standards and greater autonomy for their own national legal systems. Some thematic standards were particularly controversial in the negotiations, such as land acquisition and resettlement, indigenous peoples and labour rights. Among the disputed structural aspects were the introduction of a non-discrimination principle, explicit references to human rights, and the use of borrower country systems. ⁸³ In the consultations, participants from China emphasized its role as third-largest shareholder well positioned to play a ‘constructive role’; they claimed to speak for developing countries when they argued that many of them were frustrated at the onerous work required to fulfil the standards. They insisted that the Bank stick to its development mandate and not adopt a one-size-fits-all approach to borrowers. ⁸⁴ Participants from India and Brazil also complained about excessive implementing cost. ⁸⁵ South African representatives explicitly worried that with added requirements, the Bank risked becoming uncompetitive compared to the AIIB and the NDB. ⁸⁶ As a result, the final version of the ESF must be viewed as a compromise package and the outcome of a complex attempt to reconcile the interests of 189 states, particularly those of the main shareholders, while also satisfying CSOs in the North and the South.

3.3 The outcome of the reform: Overview of the new structure

The outcome of the reform is a unified legal text, the ESF, whose structure differs from the previous Safeguards. Previously, they were spread over a total of nine OP/BPs. ⁸⁷ An OP typically listed the duties of both the borrowers and the World Bank, as well as material standards and procedural principles; the corresponding BP spelled out the procedural steps and the internal responsibilities within the Bank. The new ESF differs in that it is a so-called ‘integrated framework’ that includes three components: A vision statement, an Environmental and Social Policy (ES Policy), and ten Environmental and Social Standards (ESS).

The ‘Vision Statement’ is novel and has no equivalent in the old safeguard system. ⁸⁸ In this statement, the World Bank expresses its commitment to the principles of ecological and social sustainability and to the Universal Declaration of Human Rights. ⁸⁹ These principles apply to all Bank activities, and not just to investment projects; ⁹⁰ in this respect, the scope of application thus extends beyond the previous Safeguards, which only covered project financing. However, the vision statement specifies no legally binding obligations. In fact, it has an aspirational thrust and specifies a non-binding guidance – and potentially interpretive – function. ⁹¹ The second component, the ES-Policy, summarizes all the duties of the World Bank in one instrument, whereas before they were scattered over a number of OP/BPs in no particular order. ⁹² The third component, the ten ESS, stipulate the obligations incumbent on the borrowers.

ESS 1–10 define the material standards of protection, procedural requirements, and individual rights of the project-affected communities, which borrowers must comply with and whose fulfilment the World Bank must supervise and guarantee. The new standards carry over numerous environmental and social requirements from the previous Safeguards, modify some of the older requirements, and introduce entirely new ones. Procedural rules concerning environmental impact assessment are now extended to include social impacts (ESS1 – previously OP/BP 4.01). Standards on land acquisition and resettlements (ESS5 – OP/BP 4.12), biodiversity (ESS6 – OP/BP 4.04, 4.36), indigenous peoples (ESS7 – OP/BP 4.10) and cultural heritage (ESS8 – OP/BP 4.11) are carried over and in part modified. Newly formulated Standards include labour and working conditions (ESS2), resource efficiency, environmental pollution, and climate change (ESS3), as well as health and security of project-affected communities (ESS4 – in part in OP/BP 4.37). Finally, the procedural rules for financial intermediaries (ESS9) and for stakeholder participation and information (ESS10) are carried over and expanded in independent standards.

Like the old Safeguards, the ESS are not directly legally binding on the borrower but specify conditions to be fulfilled before a loan agreement can be concluded and become legally binding when they are incorporated into the loan agreements. This incorporation now takes the form of a so-called Environmental and Social Commitment Plan (ESC-Plan), prepared in negotiations between the Bank management and the government of the borrower country. ⁹³ The ESC-Plans constitute the legal link that binds Bank regulations to borrowers’ duties. They identify specific measures that borrower states must undertake to comply with the ESS in the execution of the project, and lay the legal basis for corresponding information and control rights, as well as rights to legal remedies, accorded to the Bank. The ESC-Plans consolidate in a single standard document several types of plans under the old Safeguards that are now deemed mandatory for all projects. ⁹⁴ The ESC-Plans are not just a technical, managerial feature of the ESF, but a decisive legal locus where the relative bargaining power of Bank and borrower will play out in the future.

4. The Safeguards reform as a reflection of competitive multilateralism

For a more in-depth analysis, we now return to the three dimensions used above to analyze the original Safeguards. The new ESF reconfigures: (i) the relationship between the Bank and (some) borrowing states; (ii) has clarifying as well as ambiguous effects on the legal role of individual; and (iii) materially expands thematic integration while eschewing closer inter-institutional co-ordination. Against the historical and conceptual background developed above, the question is now how these changes reflect the new geopolitical context and the various forms of contestation.

4.1 Relationship between Bank and member states: Institutionalizing inequality in the Safeguards

The relationship between the Bank and its borrowing member states was one of the main battlefields in the negotiations between donor states and Bank on one side and borrowing countries (rising powers as well as weaker states) on the other side. The outcome of this battle now reconfigures this relationship through two countervailing legal changes in the ESF: The first change is the extension and intensification of Bank requirements for the implementation phase of the project, which further curtails borrower state collective autonomy and expands Bank authority over this hitherto problematic aspect of the Safeguards. The second change is a much more permissive approach towards the use of country systems: The ESF makes it much easier to replace some or all ESS with the national law of the borrower. This ‘escape clause’ is mainly aimed at emerging powers with relatively strong legal systems and administrative capacities and significantly enhances their flexibility and ownership. At the same time, it institutionalizes inequality between borrowing states at the level of secondary law: In the future, Safeguards will not apply equally to all borrowing members but discriminate between the strong and the weak among them.

The extension of Safeguards requirements for project implementation addresses the weakness of compliance in the application of Safeguards. As analysed above, the first Safeguards mostly focused on upstream regulation, i.e., had stringent requirements to be met before a loan or grant agreement was approved but left compliance to the borrower. ⁹⁵ The ESF does not change the basic model of Bank-financing combined with national execution, as this model is fundamentally linked to the Bank’s identity and role as a financing institution, and not as an administrative agency specialized in road construction, dam building, urban planning etc. ⁹⁶ But ESS10 now requires the participation of stakeholders throughout the entire project cycle, including the implementation phase. In addition, borrowers must institute project-based grievance mechanisms which allow affected people to bring complaints on the ground in the implementation phase. ⁹⁷ The ES-Policy stresses the duty of the Bank to continuously supervise project implementation, in addition to the due diligence obligation in the preparation phase. ⁹⁸ If taken seriously, these requirements go some way to addressing the implementation deficits of the Safeguards – if they apply at all, that is.

This expansion of Bank requirements is contrasted by a second, potentially dramatic, change in the ESF: The application of the ESF can now differ from borrower to borrower. The central instrument to manage this is the use of country systems (UCS). UCS was previously applicable only to pilot projects, under the restrictive conditions of OP 4.01 and thus, rarely implemented in practice. ⁹⁹ In the Safeguards reform negotiations, China, India, Brazil, and South Africa emerged as strong champions of the UCS. China used the language of ‘mainstreaming the use of national systems’, indicating that it saw the use of its own law as the new rule rather than the exception. ¹⁰⁰ In a similar vein, South Africa demanded the Bank’s point of departure should be the host country’s environmental and social parameters. ¹⁰¹ India declared that its own environmental and social system was more robust than the proposed ESF. ¹⁰² Brazil demanded even more flexibility in the UCS than initially proposed. ¹⁰³

The ESF eventually adopted now offers more opening clauses facilitating the use of the country system instead of the World Bank ES-Standards. This possibility also includes high-risk projects and does not restrict the volume. The pre-condition for using country systems is that the environmental and social standards of the country achieve ‘objectives materially consistent with’ the ESS. ¹⁰⁴ This requirement is results-based. The Bank is required to check in advance if the requisite outcome equivalence is present and, if necessary, it must agree on supplementary measures with the borrower to develop capacity. ¹⁰⁵

The use of borrower systems is potentially the most consequential change in the reform. The increased use of the country system corresponds, on the one hand, to the principle of national ownership and allows the Bank to adjust environmental and social project management to the different capacities and needs of the borrower, as efficient elements of the country’s legal and administrative systems can take the place of the corresponding parts of the ESS. It also provides the Bank with leverage to raise standards with national law generally rather than promoting isolated projects with high standards. ¹⁰⁶ Furthermore, given the diffusion of the Bank’s environmental and social standards into domestic legal systems, as described above, it seems logical from the perspective of the Bank to assume that national law today is better able to handle complex projects than it was 20 years ago.

On the other hand, allowing a too-generous handling of the opening clause carries the risk that World Bank standards will be undermined. This can also represent a risk for project-affected people and the environment, if they are not adequately protected under the country system. How real this risk is depends on how generously the management interprets the criterion of material consistency regarding country systems. If country law is to be applied, it is also necessary to ensure that this is brought on par with the protection levels offered by the ESF before the project begins to be implemented. Even if the Bank found national systems to guarantee material consistency, it remains important to continuously monitor this consistency during the implementation phase and, where necessary, to strengthen country systems before such risks can materialize. Previous experiences with the UCS show that national regulations and processes, in particular in the social realm, are generally not as protective as the World Bank standards. ¹⁰⁷ Likewise, effective law enforcement and compliance with rule of law principles are not per se guaranteed.

Ultimately, the extension of UCS can be seen as a legal institutionalization of inequality in the Safeguards regime. The application of the Safeguards, and thus borrower autonomy, now depends on the capacity of legal systems, which must ensure ‘outcomes materially consistent with the ESS’; the capacity of the national legal system thus becomes the criterion for differentiation among borrowers. This criterion differs from the standard used in the voice reform at the level of primary law, which depends on economic capacity as measured by share in the world economy. In sum, rising powers have left this main battlefield in the negotiations with considerable success. Though it is not openly laid down, it is to be expected that the domestic systems of stronger powers are more likely to get approved as ‘materially consistent’ and hence exempted from the ESF. China gets a better deal than Mali – and the collective autonomy of some is better protected than that of others. This is also in the interest of the Bank, as it helps it to remain competitive and attractive to its major clients, while still claiming a normative leadership role. Ultimately, UCS is also likely to have facilitated compromises and solutions in other areas, especially with regard to the role of the individual.

4.2 Role of the individual: Expansion and differentiation

The role of the individual and the protection of human rights were the central question for most of the CSOs and Western donor states when the Safeguards reform began. The original Safeguards had made tentative but important steps by introducing individual rights in the context of an international institution in the first place and by establishing the Inspection Panel as an institutional enforcement mechanism. The new ESF clarifies the Bank’s general commitment to human rights, strengthens and expands prior procedural guarantees, and establishes new entitlements in the areas of labour and non-discrimination. At the same time, enhanced UCS signifies a new differentiation between individuals in borrower states with strong and weak national systems. This differentiation implies new ambiguities and potentially reduces greatly the role of individuals and of the Inspection Panel as an accountability mechanism for stronger borrowers.

In a first for the Bank, the ESF Vision statement expresses an explicit commitment to human rights. ¹⁰⁸ The inclusion of this commitment in the non-binding Vision Statement and the concrete formulation represent a compromise that must be read against the backdrop of the non-political mandate of the World Bank and diverging views on human rights among donors and borrowing countries. ¹⁰⁹ China, in particular, had insisted that the Bank stick to its non-political development mandate and flagged human rights, along with labour rights, as ‘politically very sensitive’. As the reference to human rights is made only as a non-binding commitment, it does not live up to the standards of the binding Human Rights Impact Assessments (HRIA), practiced, for instance, in the UN Development Programme and other UN agencies. ¹¹⁰

Yet the ESF represents a major step forward for human rights in two respects: First, human rights are concretized through explicit reference to the Universal Declaration of Human Rights (UDHR). This is significant because the UDHR guarantees not just economic, social, and cultural rights but also civil and political rights. Previously, the Bank had always denied that civil and political rights were covered by its legal mandate. ¹¹¹ Second, the wording of the commitment – ‘prevent any negative impact on human rights’ – for the first time recognizes the dimension of respect for human rights as relevant for the Bank itself. Previously, the Bank had only unequivocally recognized its role in supporting member states in the fulfilment of social and economic rights and considered the respect dimension only in the event of gross violations by member states. ¹¹² Now the Vision statement offers an additional argument to corroborate widespread legal interpretation in academic writing, according to which World Bank is already under the obligation to respect human rights by virtue of general international law. This interpretation holds that the Bank is prohibited from aiding and abetting human rights infringements committed by its borrowers, particularly by financing projects that are problematic from a human rights perspective. ¹¹³ It is possible to build on this progress in the practical application of the ESF: the Vision Statement opens up the possibility to interpret the Bank’s binding obligations in the ES-Policy in the light of human rights law, an interpretation that the Inspection Panel already pursues in its own practice. This will become relevant for interpreting, for instance, rules on non-discrimination and the protection of housing in resettlement processes. ¹¹⁴

The individual ESS strengthen and expand the procedural and substantive entitlements for project affected individuals and other stakeholders. ESS10 defines the notion of ‘meaningful consultation’ with greater precision by integrating aspects of the Inspection Panel case law, extends consultation requirements into the implementation phase, and introduces mandatory project-based grievance mechanisms. ¹¹⁵ ESS5 on land acquisition requires borrowers to give women due consideration and clarifies that forced evictions are prohibited unless all provisions of national law and ESS5 have been fully complied with. ¹¹⁶ ESS2 on labour for the first time establishes minimum standards for the protection of workers and bans child labour in Bank projects. ¹¹⁷ ESS1 introduces mandatory social impact assessments for all projects and recognizes a non-discrimination principle. ¹¹⁸ This non-discrimination assessment is a novelty, and its inclusion was controversial in the negotiations until the very end. In particular, member states were unable to agree on a conclusive definition of discriminated and vulnerable groups. In light of that, ESS1 includes a footnote outlining some abstract criteria and giving non-exhaustive examples, namely older people and minors. ¹¹⁹ These criteria are somewhat specified in an internal Bank Procedure and a Directive that concretize the Bank’s due diligence obligations. ¹²⁰ This non-discrimination principle is an entry point for a human rights-conforming interpretation advocated above: Protected groups can be identified by referring to the anti-discrimination provisions of international human rights agreements ratified by the affected borrower state. ¹²¹

The role of the Inspection Panel, a central mechanism to effectuate the Safeguards as an individual rights regime, has become more ambiguous with the reform. On the one hand, the inclusion of new rights expands the Panel’s role and ability to interpret them. On the other hand, the increased possibility to escape the ESS and apply domestic law through the UCS mechanism may diminish its reach significantly. This will depend on the role the Panel will play if country systems replace the ESS. Two different interpretations are conceivable: a narrow interpretation would be that the Panel restricts itself to examining whether the management was justified in assuming that country systems achieved outcomes materially consistent with the ESS. This would restrict the scope of the Inspection Panel considerably and empower Bank management. ¹²² A second, wider, interpretation would be that the Panel not only controls the consistency test but also investigates violations of national laws and procedures insofar as they replace specific ESS. This would considerably enhance the role of the Panel. The first, narrower, interpretation seems more in line with wording of the Inspection Panel Resolution of the Board, which does not give the Panel a mandate for reviewing compliance with national law. On the other hand, oral statements by the Bank’s Operations Policy Country Services department (OPCS) seem to favour the second, wider, interpretation. Given these ambiguities introduced by the Safeguards reform, the Bank announced a separate reform process of the Inspection Panel in fall 2017, which might bring clarification on this point.

In sum, the overall trend is ambivalent. On the one hand, donor countries and CSOs seem to have stood their ground in this second negotiation battlefield. They have not only defended most pre-existing individual entitlements but also succeeded in expanding them in the ESS and in including an explicit reference to human rights in the Vision statement. The latter is a major achievement for CSOs, even though their demands had gone much further. On the other hand, the Vision statement is non-binding and the entitlements in the ESS potentially have limited application if country systems are used widely, in which case the Inspection Panel’s role is also likely to be diminished. In fact, the provisions on the role of the individual extend the inequality between weaker and stronger member states to the citizens of those states, with reverse effects: Citizens in Mali will be better protected under the ESF than under the old Safeguards. Citizens in China and India will not be able to rely on international protection to the same extent as before. Where country systems are applied, all will depend on the effective protection of individual rights in national legal systems.

4.3 Thematic coverage: Material integration, inter-institutional fragmentation

The last dimension, that of thematic coverage of the Safeguards and of integration vs. fragmentation, also embodies some significant change. The ESF covers a range of additional risks and issues and brings the Bank materially closer to other international legal regimes. In contrast to this material integration, fragmentation largely remains at the institutional level.

The expansion of thematic coverage occurred in particular with ESS2 on labour and with ESS3 in respect of climate change. ESS3 for the first time adopts explicit guidelines for the reduction of greenhouse gases in World Bank financed projects. ¹²³ China had agreed with this standard, simply demanding that ‘thresholds be acceptable to developing countries’, whereas Brazil suggested an even stronger alignment with the international climate regime, especially the Paris Agreement. ¹²⁴ Enhanced convergence was also achieved in ESS7 on indigenous people, which is now largely equivalent to procedural and material standards of the International Labour Organization (ILO)’s Indigenous and Tribal Peoples Convention, 1989 (No. 169) and the UN Declaration on the Rights of Indigenous Peoples, without expressly referring to them. In particular, ESS7 introduced the requirement of Free, Prior and Informed Consent (FPIC), in cases where indigenous land rights or cultural goods are particularly affected or in the event of impending resettlements. ¹²⁵

In contrast, ESS2 on Labour and Working Conditions proved to be among the most divisive issues in the negotiation process. ¹²⁶ Under the old Safeguards, the World Bank did not have a corresponding standard, and many borrowers strongly opposed the introduction of strong labour standards significantly exceeding their domestic labour law protections. China, in particular, argued that the Bank’s labour standards needed to be matched with the ‘level of development of the country’ and should not be used as a tool to intervene in the political sphere. ¹²⁷ A particularly controversial aspect in the negotiations was the relationship of ESS2 to the ILO’s minimum international law standards. Even if the Bank claims that the final ESS2 embodies the ‘core principles of ILO Fundamental Principles and Rights at Work’, ¹²⁸ no direct references are made in the wording of ESS2 to concrete ILO standards; instead, the Standard repeatedly refers to national law, without requiring national law to be compatible with ILO minimum standards. This is particularly evident in collective labour law: If freedom of association of workers is not guaranteed in the borrower state, this will not prevent project financing. ¹²⁹

Compared to the original Safeguards, the ESF has expanded substantive convergence with other norms of international law and re-integrated the Safeguards with other international legal regimes to some extent. One factor that contributed to this result was the inclusion of other international institutions in the consultation process. At the drafting stage, some inter-institutional co-ordination thus occurred. However, the application of the Safeguards in daily practice remains characterized by inter-institutional fragmentation: The Safeguards themselves do not stipulate any requirements for consultation with other international institutions in the application of the standards, nor do they prescribe interpretation according to the principle of systematic integration or require taking into account interpretations of parallel norms by other institutions, human rights mechanisms or international courts and tribunals. The Bank, thus, retains full authority to interpret its standards subject to its own logic and constraints. At the same time, a different form of inter-institutional co-operation might increasingly shape the global order of the twenty-first century: The Bank has recently concluded co-financing agreements with the AIIB making the new multilateralism look much less competitive than one might have thought.

5. Conclusion

The World Bank’s new environmental and social Safeguards are a reflection of the new geopolitical context in which the Bank operates. This context is best described as competitive multilateralism. For the Bank and the ESF process, this competition had several elements: Political contestation by rising powers, especially by China through the foundation of the AIIB and NDB; economic competition from other lenders (private banks, philanthropic lenders, national agencies of rising powers); and politicization through the increased observation and criticism by CSOs. The new ESF is one element of several reforms with which the Bank tries to reposition itself in this environment. These reforms include the new Safeguards and other elements of secondary law such as a new lending instrument, the Program for Results, a new Access to Information Policy, and the formalization of knowledge products. ¹³⁰

The common denominator of these repositioning efforts and general answer to this new context seems to be a strategy of cautious adaptation, which ultimately confirms institutionalized inequality. This much less affects primary law and the institutional structure of the Bank, such as voting rights and the selection of its president. One could say that the Bank avoids changes in the ‘engine room’, i.e., in institutional or power structure. Instead, institutional inequality is now introduced into secondary law Safeguards through the UCS. The overall intention of these reforms seems to be to satisfy new international forces, namely rising powers and civil society, without cutting too deeply into the flesh of established structures.

A comparison of old and new Safeguards along our three analytical categories allows three tentative conclusions on the evolution of global order and the role of international law in it. First, the Safeguards reform has strengthened member state sovereignty in those cases where country systems are used: international standards retreat and national law takes centre stage. At the same time, this strengthened sovereignty is an unequal one because only stronger borrowers will enjoy this privilege whereas weaker borrowers are subject to broadened and deepened standards of the Bank. Interestingly, the legal criteria for differentiating between strong and weak depends on the strength of the respective legal and administrative system – and not economic performance, which is the criterion for weighted voting in primary law. This shift from economics to law might be the ultimate victory of the Bank’s good governance agenda: Sovereignty becomes contingent upon good governance. The Bank has, thus, repositioned itself as the arbiter of sovereignty. This indicates that member state sovereignty and the Bank’s international authority are not in a zero-sum relationship: The Bank may have lost financial leverage but it has successfully built on its epistemic and legal authority in the Safeguards reform.

A similar dynamic can be observed with respect to the second analytical dimension, the role of individuals. While individual entitlements in the new ESF have been expanded, fewer individuals are likely to enjoy these expanded rights, as UCS will become the norm at least in stronger borrowers and hence individual entitlements based on the ESF retreat. This does not necessarily mean than individuals are worse off: national systems may offer equivalent (or even better) protection, at least in some cases. Conversely, expanded international rights do not necessarily improve the situation of individuals on the ground. To the contrary, critical legal scholars have long pointed out that ‘mainstreaming’ individual rights, and particularly human rights, in international institutions may muffle their emancipatory potential and actually result in institutional self-empowerment. Critics might be wary of the Bank deploying its epistemic and legal authority in ever new areas of rights such as labour, now empowered by the ESF. They may, for instance, discern a tendency to depoliticize labour rights through a focus on individual protections at the expense of collective mobilization rights.

Finally, the third dimension points to substantive integration but institutional fragmentation of international legal regimes. This may reinforce international norms, but given their indeterminacy, institutional interpretations will matter greatly. Hence, the enhanced interpretive reach of the Bank may ultimately reinforce fragmentation unless the Bank engages in a meaningful dialogue with competent institutions like the ILO, UN human rights bodies, and climate governance actors.

Taken together, the developments in these three analytical dimensions also point to conclusions for the role of international law in the new global order more generally. For one, legalization continues: borrowers and donors continue to choose the form of law to negotiate interests and express compromise. The Bank continues to use law in a functionalist logic: Reforming the Safeguards was about effectuating its development mandate. What changes is the nature of law and of the lawmaking process: In terms of form, secondary law becomes the decisive locus for negotiating and making global norms. Process-wise, organized multi-stakeholderism has replaced state-led lawmaking dominated by powerful donors. Substance-wise, the universalization of ‘Western’ norms through international lawmaking is yielding to a pluralization of international law, which becomes normatively and ideologically more differentiated and more complex. Whether this increases legitimacy remains an open question.

Finally, through the lens of the Safeguards reform, the debate about unipolar vs. competitive multilateralism requires some differentiation. On the one hand, the Safeguards reform confirms emerging powers’ double strategy of continued engagement in old institutions combined with the creation of new ones. On the other hand, it seems less clear that the new multilateral banks like the AIIB and the MDB will actually compete with the World Bank across the board. Early practice also indicates co-operation in the form of co-financing agreements, and the AIIB has emulated the Bank to some extent by establishing its own set of social and environmental standards that are not, at least on paper, indicative of a fully-fledged race to the bottom. In any event, the Bank retains a competitive advantage in respect of its own epistemic authority and legal expertise, which the Safeguards reform is likely to enhance. Reports on the death of World Bank are, thus, greatly exaggerated. Its role as a lender may diminish, but its international authority as a norm-setter and epistemic actor is likely to endure under new geopolitical conditions.