Here is where things stand, as best as I can tell from the public record.

The timeline, updated

Neither Phase I dialogue nor a first draft happened on schedule. IFC’s own Sustainability Framework page, last updated June 2026, now shows Phase I concluding in October 2026, with Phase II public consultation anticipated to commence “October–December 2026.” The internal drafting window IFC confirmed at its April 2026 CSO information session runs April-September 2026.

Which means the draft is being written right now, this month, and could land for public comment as early as October, roughly six weeks from today.

That’s not a lot of time to think about what you actually want to say to it.

What happened since February

The February dispatch mapped the main pressure points: climate, FI governance, remedy, rights language, biodiversity, transparency. Most of that picture holds at present. But one development since then has materially changed the terrain, and it would be odd not to mention it here given that this newsletter covered it in some detail in July.

The Board’s June 2026 decision in the Cambodia microfinance case, overruling the CAO’s compliance findings, asserting no policy non-compliance and suspending all CAO cases relating to financial consumer protection in microfinance, explicitly deferred the underlying policy question to the Sustainability Framework revision. The IFC Board statement is clear on this: financial consumer protection will be addressed “at the time of the Sustainability Framework update.”

So the drafting team currently writing the revised framework has been handed a specific, politically loaded question to answer: do microfinance borrowers (people reached through IFC’s FI clients) count as affected people under IFC’s policies or not? The Board declined to say yes. It also stopped short of formally saying no. The draft will have to do one or the other or find some creative way to not quite do either.

What makes this interesting is that IFC’s April 2026 CSO session (a summary of Phase I feedback themes) makes no mention of FIs or consumer protection among the thematic areas covered. The slides cover climate, biodiversity, extractives, disability, Indigenous Peoples, stakeholder engagement and Access to Information Policy (AIP). The FI question which is arguably the most consequential thing the revision now has to resolve doesn’t appear. Whether that is a deliberate framing choice or just the session’s scope, I genuinely don’t know. But it’s a gap worth noticing.

What the draft is likely to contain

Based on confirmed Phase I feedback from IFC’s own CSO session and the February assessment, here’s a reasonable expectation of what the first draft will and won’t do.

Climate provisions will be strengthened. There’s no realistic path to a revised framework that doesn’t expand GHG requirements, tighten Scope 3 disclosure expectations and strengthen Paris alignment language. Civil society has been unified on this, IFC has committed to Paris alignment publicly and the CAO’s own climate advisory note found the current standards insufficient. The question is architecture, i.e. whether it will be a standalone climate standard versus a significant restructuring of PS1 and PS3. Either is possible.

Biodiversity and supply chains will tighten. Alignment with the Global Biodiversity Framework, stronger critical habitat language, enhanced cumulative impact requirements — all confirmed in Phase I feedback. Regulatory pressure from mandatory due diligence regimes in major markets gives IFC additional incentive to move here regardless of CSO pressure.

Indigenous Peoples provisions will be expanded. IFC committed to a dedicated consultation mechanism for Indigenous Peoples organisations during Phase I. The April session reflected calls for stronger FPIC, explicit coverage of IPs in voluntary isolation and transboundary protections. Expect movement.

The AIP will be updated. Proactive disclosure of climate assessments, stakeholder engagement activities and sub-project information have all been raised. IFC has already updated its Disclosure Portal, signalling willingness to move on transparency. How far the exceptions regime gets narrowed is the live question, particularly for FI sub-projects where client confidentiality arguments are strongest.

Labour and resettlement language will be sharpened. PS5 has had dedicated CSO coordination and Inclusive Development International submitted detailed recommendations on land acquisition in early 2026. Expect revised language, though how enforceable it becomes in practice is another matter.

The FI and consumer protection question is unknown. See above. This is the one area where the February confidence assessment can’t be updated, because the Board’s Cambodia decision has made it politically delicate in a way it wasn’t six months ago. The drafting team has to resolve it, and whatever they write will be read very carefully by a set of people who are now paying close attention.

How to actually engage

The public consultation when it opens will invite written submissions. Based on how IFC handled Phase I, a few things are worth keeping in mind before you spend time drafting your response.

Submissions that are specific do more work than submissions that are general. “Strengthen PS2 on labour” is easy to aggregate into a thematic bucket and forget about. A submission that identifies a specific gap (a definition that creates a loophole, a threshold that doesn’t reflect current practice, a provision that conflicts with how the Equator Principles have been applied) is harder to absorb without engaging with it.

The first round of consultation is unlikely to resolve everything. The process runs in two rounds, with a revised draft expected before the second. Questions that are genuinely politically contested (e.g. FI consumer protection, the depth of remedy obligations or full FI sub-project disclosure) may be left deliberately vague in the first draft to see what pressure comes back in the first round of comments. It means the second round may matter more than the first for the hardest questions.

And for practitioners specifically: the revised framework will flow through into the Equator Principles which means changes in scope, thresholds or definitions will eventually affect how commercial lenders structure their own E&S requirements. Engaging in this consultation is an opportunity to influence the standards your own clients will be asked to meet for the next decade or more.

The Hamburg Score View

The revision has been framed as a “generational opportunity.” That phrase has been used by IFC, by civil society and by practically everyone who has written about the process.

However, generational opportunities come with generational risks. The last revision, in 2012, produced a framework that has since been documented by the CAO’s own advisory work as consistently failing to deliver remedy to the communities it was designed to protect, because supervision, disclosure and accountability mechanisms were not resourced or structured to match what the text said.

The Cambodia case made that gap visible at the level of a Board decision. The question for the SF draft that lands in October is whether the revision addresses that gap or produces better language that suffers the same fate.

Better language is easier to deliver. The thing worth reading the draft for is whether the supervision architecture, the disclosure requirements and the accountability mechanisms have been designed to hold in practice, not just on paper.

In 2022, eighteen Cambodian borrowers filed a complaint. They described what had happened to them: land sold to repay debts, families going hungry, children pulled out of school to work, threats against those who pushed back. They asked the IFC’s independent watchdog, the Compliance Advisor Ombudsman (CAO), to investigate whether IFC had met its own E&S standards when it made those investments.

Four years later, they had their answer. One of the eighteen did not live to receive it.

What the investigation found

The CAO spent four years on this case. Its 167-page investigation report, published in October 2025, found that IFC had failed to comply with its own Sustainability Policy across investments in six Cambodian banks. Specifically, IFC did not conduct E&S due diligence commensurate with known risks; it did not identify borrowers as vulnerable groups under its own policies; it did not require its clients to apply the relevant standards; and it failed to assess impacts on Indigenous Peoples’ communal land where banks had accepted community land titles as collateral without consent.

Under IFC’s own rules, a finding of non-compliance triggers a clear requirement: management must prepare a Management Action Plan (a remediation roadmap) which the CAO then monitors.

IFC’s management declined. Instead, it submitted a report rejecting the CAO’s findings and arguing that microfinance borrowers do not qualify as “affected people” under IFC’s policies and that therefore the E&S framework does not apply to them at all.

On 23 June 2026, the IFC Board of Directors sided with management. It approved what it called a Special Management Action Plan, a commitment to hire a local facilitator to help the eighteen complainants navigate existing national complaint mechanisms, while simultaneously stating that it “recognised that there has been no policy noncompliance under IFC’s Policy on Environmental and Social Sustainability.”

The Board proposed to remedy harm it officially said had not occurred.

The next day, Janine Ferretti resigned as CAO Director General. She had led the office since 2021.

No documented precedent exists, across the public record of independent accountability mechanisms at any major multilateral development bank, for a board formally overturning a positive compliance finding by its own watchdog.

The jurisdictional question

The Board’s decision rests on a specific legal and policy claim whose implications extend well beyond Cambodia.

IFC does not lend directly to Cambodian farmers or small traders. It lends to banks which then lend to borrowers. IFC’s argument endorsed by the Board is that this distance places end borrowers outside the scope of its accountability framework. They are customers of IFC’s clients, not IFC’s own project-affected people.

As Human Rights Watch noted in its response to the decision, the CAO’s investigation had specifically rejected this narrow reading of IFC policy. The Board overruled the CAO and endorsed management’s interpretation instead.

The consequences specifically apply to a subset of IFC’s financial intermediary portfolio. FI clients investing in higher-risk sub-projects (energy infrastructure, agribusiness, extractives, etc) remain required to apply the Performance Standards, and communities affected by those investments retain access to the CAO.

What the Board decision does is carve out financial consumer protection in microfinance specifically: end borrowers reached through microfinance institutions are not, under the logic the Board endorsed, IFC’s project-affected people. Given the scale of IFC’s microfinance exposure globally, that is a significant exclusion. The decision is about who counts — and for one large category of people the IFC reaches, the answer is now formally: not you.

Pattern, not aberration

The Board’s decision also leaves a broader group of people in limbo. Beyond the 18 original complainants covered by the Special Management Action Plan, the CAO’s seven-point decision identifies a further 11 complainants in Cambodia whose cases are now suspended pending the policy discussion, with IFC to inquire whether they would accept the same limited support offered to the original eighteen.

The override did not happen in isolation either. On 9 June (two weeks before the Cambodia decision) the WBG Boards approved the consolidation of the CAO, the Inspection Panel and the Dispute Resolution Service into a single Integrated Accountability Mechanism, under a stated commitment to “no regression” in accountability standards. The recruitment process for the new mechanism’s leader was also launched that day.

The CAO’s own website now confirms what follows: the CAO has suspended processing of all ongoing cases relating to financial consumer protection in microfinance, and will not accept new cases in this area pending a broader policy discussion. The IFC Board statement is explicit that financial consumer protection will be considered during the forthcoming update to IFC’s Sustainability Framework.

The policy question the Board declined to resolve through an accountability process will now be resolved through a policy revision process — with the outcome of the Cambodia case having already established, through a Board decision rather than a formal policy change, where IFC’s accountability perimeter sits.

There is a longer pattern here too, and the CAO has documented it itself. Its own 2023 Remedy Gap advisory note found that people lodging complaints with the CAO “often do not see their specific concerns addressed despite attempting dispute resolution or undergoing a lengthy compliance process” — left, in the CAO’s own words, “to suffer harm and bear the unintended cost of development projects, even in cases where CAO has found non-compliance.” The Board override in Cambodia is unprecedented in form. The outcome for complainants is consistent with what the CAO’s own research had already described as a systemic pattern.

What changed is that the failure is now on the record at the level of a Board decision.

What Ferretti’s resignation means

The Board’s statement on Ferretti’s resignation thanked her for her service and confirmed that the recruitment process for her successor — launched on 9 June, before the Cambodia decision — remains underway.

The new mechanism will be led by a Vice President, a rank within the WBG management structure. Its head will be appointed through a Board-led process, replacing the arrangement under which civil society participation in selecting the CAO Director General was part of the mechanism’s design. That the recruitment was already running before the override became public is itself a detail worth holding: the institutional restructuring was proceeding on its own logic, independent of how the Cambodia case resolved.

An accountability mechanism whose head holds a management title, appointed by the board it is meant to hold to account, operating under a mandate shaped by the precedent the same board has just set: these are the conditions under which the Integrated Accountability Mechanism will begin its work.

The Hamburg Score View

The advocacy organisations have called this a crisis of accountability, and they are right to do so. But the framing of “unprecedented” risks obscuring a harder truth: the CAO was already a mechanism that rarely delivered remedy. What the Board did on 23 June was make a pre-existing dysfunction visible.

The more consequential question is what happens next in the Sustainability Framework revision. The Board has now established (by implication) that microfinance borrowers reached through IFC's financial intermediary clients fall outside its E&S accountability perimeter. That position will either be ratified in the revised Framework, challenged and corrected, or, most likely, left deliberately ambiguous. Deliberate ambiguity has been the IFC's preferred instrument for managing accountability questions it does not wish to resolve.

Ferretti’s resignation is a signal that the informal checks that kept the system functioning — the professional credibility and institutional standing of the CAO’s leadership — have reached their limit. The new mechanism will inherit the precedent, the narrowed jurisdictional perimeter and a mandate shaped by the board that just demonstrated its willingness to use it.

Whether microfinance borrowers reached through IFC's financial intermediary clients will have meaningful recourse under that new architecture is the question the Sustainability Framework revision will answer. The revision is underway.

The piece grew out of coursework during my master’s at Vrije Universiteit Brussel. My course instructor, Bhaskar Nath, who was also a co-editor at the journal, was the one who told me there was a proper article in the essay and pushed me to develop it and submit it. Without that, it would almost certainly have stayed a piece of coursework nobody ever saw again.

I want to be upfront that this is a departure from what I usually write here. Most of what goes out under this newsletter concerns IFC Performance Standards and other E&S lender standards. But the concept behind both is the same: watching what institutions actually do with the doctrine they write down rather than what they say they believe. Twenty years ago I did that with two churches. The story of what happened next, I think, is worth telling and it has very little to do with what I expected.

What I thought I was looking at

My argument back then was fairly simple. The Russian Orthodox Church had, in its year 2000 document The Bases of the Social Concept, an entire chapter (Chapter XIII), devoted to the environment. Unusually so. Most Orthodox churches did not have anything like it. The chapter talked about humanity’s responsibility for creation, about the dangers of consumerism, about restraint. It was cautious and suspicious of Western environmentalism which it tended to see as another secular import, but it was there, on paper, in a formal document approved by a church council.

The Ethiopian Orthodox Church had nothing comparable in writing. No equivalent chapter, no equivalent council document. But it had something the Russians did not: centuries of monasteries surrounded by forest, patches of genuine old-growth woodland preserved by the simple fact that nobody was allowed to cut down a sacred grove. Conservationists were starting to notice that some of the best-preserved forest fragments in the Ethiopian highlands were the ones with a church in the middle.

So my paper made a case that’s probably obvious in hindsight: one church had the theology without the practice, the other had the practice without the theology, and maybe each could learn something from the other. I thought I was watching the early stages of something. Two religious institutions edging, in their own very different ways, toward taking the environment seriously.

I was still a young researcher finding my feet. I believed in trajectories.

What actually happened to the Russian text

Chapter XIII never went anywhere. There was no diocesan environmental programme, no follow-up document, no real institutional life given to those paragraphs about restraint and stewardship. It just sat there, technically still part of church doctrine, doing nothing in particular.

What happened instead was chapter VIII.

This chapter is about war and peace. It lays out conditions under which the church considers war justified — broadly, the kind of “just war” thinking you’ll find in a lot of Christian traditions. I did not pay it much attention in 2003. It was not the chapter I was writing about, and nothing in the political moment suggested it would matter.

It matters enormously now. Patriarch Kirill has spent the last several years building a case that Russia’s war in Ukraine is not just permissible but holy, drawing on exactly this part of the Social Concept, dressed up further through the World Russian People’s Council and the broader “Russian World” idea, the notion that Russia is defending a distinct civilisation against a hostile West. The same document I once combed through for hints of environmental seriousness turned out to contain the doctrinal materials for justifying an invasion.

What I find genuinely striking, reading about this now, is a piece of recent scholarship comparing the war provisions in the Russian text against the equivalent section in the Catholic Church’s social doctrine. Olga Nedavnya, a Ukrainian religion scholar, published a close comparative analysis in 2024 of exactly this question. The two documents look similar on the surface — both set out conditions for a just war. But looked at closely, the Russian text’s conditions can be read to justify the side doing the attacking, not just the side defending itself. The Catholic version cannot really be read that way. It’s not that Kirill twisted a peaceful document to mean something it never could. The door was built into the text from the start. Nobody just happened to notice it was unlocked in 2022.

There is a small irony in this for me personally. In 2005 I did a close, careful reading of this document looking for evidence of environmental seriousness. It turns out the document’s real significance, twenty years on, lies in a completely different chapter, doing something nobody, including me, saw coming.

It’s worth saying that this was not the only path available to Orthodoxy. The Ecumenical Patriarchate in Constantinople went the other way entirely. Patriarch Bartholomew has spent decades building a reputation as the “Green Patriarch,” and in 2020 his church produced its own social document, For the Life of the World: Toward a Social Ethos of the Orthodox Church, written at Bartholomew’s request and endorsed by the Holy and Sacred Synod of the Ecumenical Patriarchate — far more developed on the environment than anything Moscow ever wrote. Two patriarchates, both Orthodox, both starting from a broadly similar theological inheritance — and twenty years on, one is a genuine voice in global climate conversations, the other is providing the doctrinal cover for a war.

What actually happened in Ethiopia

The Ethiopian story runs on a strange parallel track.

On the conservation side, my 2003 instinct held up better than I had any right to expect. The church forests are real and they matter more than I knew at the time. There’s now a solid body of published research — botanists and ecologists doing actual fieldwork — documenting these forest patches as genuine biodiversity refuges. A 2024 review in the journal Sustainable Environment pulled together decades of this work and confirmed that areas around the churches host endangered species found nowhere else in the surrounding agricultural land, sustained by a church doctrine that treats destroying any plant or creature on church grounds as a sinful act. The practice I noticed as an interesting curiosity in 2003 has, if anything, become more clearly important with the science behind it.

But the institutional story is much darker, and it rhymes with Russia’s in a way that truly unsettles me.

The war in Tigray broke out in November 2020 and lasted two brutal years. The Ethiopian Orthodox Church, like the Russian one, found itself standing very close to state power at exactly the wrong moment. Senior church figures were seen publicly blessing the federal army. Reports surfaced of monasteries and churches in Tigray being shelled and destroyed, of clergy among the dead, while church leadership largely stayed quiet. In May 2024, the Holy Synod itself sent a formal apology to the Tigrayan church and its followers for failing to push for an end to the war in time and for not being physically present in Tigray during it. The Tigrayan side did not accept it as adequate.

The consequences did not stop there. Tigrayan archbishops, accusing the wider church of complicity and abandonment and of doing nothing to protect churches and monasteries from destruction, declared their own ecclesiastical independence — first in May 2021, then more formally as a separate synod in October 2024. As one analysis put it, the split mirrors almost exactly what happened between the Russian and Ukrainian Orthodox churches over the same war in Ukraine — a mother church’s wartime conduct triggering a formal break by its own regional branch. A church that had stood as a single institution for the better part of two thousand years now has a breakaway branch, born directly out of how it behaved during a war.

So: a church whose practical, on-the-ground environmental stewardship is more impressive than I gave it credit for in 2003, sitting inside an institution that fractured under the pressure of war and its own closeness to political power. Conservation continued in the forests. The church itself did not hold together.

The pattern, twenty years on

Put the two stories side by side and something becomes hard to avoid. In both cases, the practical, often monastery-level work — Russian environmental theology on paper, Ethiopian forest conservation on the ground — turned out to be the least consequential part of the story. What actually shaped these churches over the past twenty years was war, and each church’s relationship to state power when war came.

Neither church became the kind of environmental actor I imagined writing about back in 2003. Both got pulled into something much larger and much more dangerous — nationalism and civilisational politics in Russia’s case, ethnic conflict and institutional collapse in Ethiopia’s. The Ecumenical Patriarchate’s example shows this was not inevitable for the Orthodoxy as a whole. It’s something more specific to how each of these two churches positioned itself relative to its own state.

A note to close on

I’m not sure what to do with a 20-year-old paper that turned out to be asking the right question about the wrong chapter. There’s something humbling about it. No, I don’t think the original argument was foolish. But I assumed institutions develop along the lines you can see forming at the time you’re watching them. Mine did not. They went somewhere else entirely, somewhere a young researcher writing in 2003 had no way of anticipating.

More than twenty years into a career that has taken me from project finance to independent consulting and back, I think that’s probably the most honest thing I can say about institutional prediction in general: you can read the documents as closely as you like, and the chapter that ends up mattering is very often not the one you were reading.

What it also does, almost as an aside, is reveal a significant organisational change in how IFC now governs E&S risk internally. That change has received almost no external commentary, and it deserves some.

1. What Changed

The key sentence appears in the manual’s first section, framed as context rather than announcement:

“This version reflects the integration of E&S risk management and opportunities development into IFC’s regional vice-presidencies, with a dedicated central E&S Policy and Risk department providing a ‘second line of defense’ risk management function and supporting regional E&S teams.”

Under the previous model, IFC’s E&S specialists were perceived (at least from the outside) as a relatively distinct technical function, sitting alongside but somewhat separate from investment operations. The new model embeds E&S teams explicitly within regional vice-presidencies, alongside investment officers, advisory teams and upstream operations. A central department — CES, the E&S Policy and Risk department — provides oversight, policy interpretation and quality assurance from the centre.

In short: E&S has moved closer to the deal. Whether that is good, bad or simply a different set of trade-offs is the question this Dispatch sits with.

2. The Governance Model IFC Has Imported

The manual’s language is saturated with governance vocabulary — oversight, escalation, processing levels, risk tiering, quality assurance, decision gates, clearance. It reads more like a banking risk manual than an environmental handbook.

And indeed, the model IFC has adopted is recognisable from financial risk management: a three-lines-of-defence structure.

The first line is the regional E&S teams, now embedded within operational vice-presidencies. They conduct due diligence, prepare key deliverables, engage with clients and manage projects through the investment cycle. Investment Officers lead project teams; E&S specialists are part of those teams.

The second line is the CES Risk Officers (ESROs) who provide independent review and clearance of key deliverables, particularly for high-risk projects. The manual is explicit about their function: ESROs “drive global consistency, quality and harmonization of approaches across and within regions.” For High-Risk Projects, ESRO review and clearance is mandatory before documents are shared with the transaction team or the client.

The third line, implicitly, is the accountability and evaluation architecture that sits outside IFC’s operational chain — CAO, the Independent Evaluation Group and the Board.

This is a coherent governance design. In banking, the three-lines model has become standard precisely because it separates the people doing the work from the people checking it, while keeping both inside the same institution. It has worked reasonably well for credit risk, market risk and operational risk. The question is whether E&S risk behaves enough like those risks for the model to transfer cleanly.

3. Where the Analogy Gets Complicated

Credit risk can be quantified. A loan either performs or it doesn’t. The metrics exist, the data accumulates, the escalation triggers are reasonably legible. A risk officer reviewing a credit decision is working with numbers that (however uncertain) occupy a shared analytical framework.

E&S risk is messier. A community’s relationship with a project — its sense of whether it has been genuinely consulted, whether its concerns have been heard, whether its land rights have been respected — does not produce a number. Social conflict, Indigenous rights disputes, gender-based violence risks, grievance patterns: these require judgement that is contextual, often contested and difficult to standardise across geographies.

The manual acknowledges this implicitly. It places considerable weight on the ESRO review function as the mechanism for ensuring consistency. ESROs are described as providing not just clearance but advice, interpretation and harmonisation across regions. For genuinely complex situations, the manual envisions early engagement between the Lead E&S Specialist and the ESRO. That is a reasonable design for managing complexity.

What it requires, however, is that ESROs (a relatively small central function) maintain sufficient bandwidth, regional knowledge and sector expertise to provide meaningful second-line review across IFC’s entire global portfolio. Whether that is achievable in practice is a question the manual does not address.

4. The Human Capital Question

The manual is detailed about governance architecture but is notably sparse about the people expected to operate it.

The closest it comes to specifying qualifications for regional E&S specialists is this, in the description of the Lead E&S Specialist role (para 3.13): “The LESS is appointed according to the expertise required to address the project’s technical, sectoral and regional issues.” For Regional Industry Leads, the description (3.12) is: “senior technical resources for E&S risk management for their defined sector and region.”

That is broadly it. There are no provisions in the document for minimum years of experience, mandatory sector expertise, required familiarity with specific Performance Standards, professional accreditation, competency matrices or independence criteria for specialists.

This observation requires a qualification. The ESRP Manual is a procedures document, not an HR framework. IFC almost certainly maintains internal position descriptions, grade requirements and technical competency standards that sit elsewhere in its systems. The absence of competency requirements from this document does not mean they do not exist.

What it does mean is that the ESRP (the document that describes how E&S risk management now works) rests on an implicit assumption about the people performing it. The governance model is described in considerable detail. The human capital question is left to other systems.

In banking regulation, this would be unusual. Prudential frameworks for financial risk management typically place significant emphasis on the competence, authority and independence of risk staff alongside the governance architecture they operate within. Here, IFC has described the architecture carefully and left the staffing assumptions largely offstage.

5. The Geography Problem

Embedding E&S teams within regional vice-presidencies means that the quality of E&S appraisal now depends on who is available in those regional markets.

IFC operates across more than 100 countries. The calibre of E&S expertise available locally varies enormously. Attracting experienced senior specialists to field locations — in fragile states, conflict-affected environments or markets where international development finance experience is thin — is genuinely difficult. It has always been difficult. The previous model, in which E&S was more centralised, was not immune to this problem. But centralised functions carry their own quality assurance mechanisms, including proximity to institutional knowledge, peer review and the accumulated expertise of a specialist department.

Regional embedding places greater weight on the judgement of individual specialists operating closer to the ground, in more varied and sometimes more challenging contexts, with the ESRO second line providing review from a distance. That may produce better outcomes in some cases — earlier engagement, deeper client relationships, more contextually sensitive appraisal. It may produce more variable outcomes in others.

The manual’s quality audit mechanism (CES audits of regional team supervision efforts) is the structural response to this risk. Its effectiveness will depend on frequency, coverage and the willingness of the system to act on what audits find. The manual describes the function; it does not describe its reach.

The Hamburg Score View

The ESRP Manual reflects a considered organisational choice. Integrating E&S into regional operations is not an obviously wrong decision — there are real arguments for putting specialists closer to where projects happen and decisions are made. The three-lines-of-defence model is a serious governance framework, not a fig leaf.

But the model raises questions it does not answer.

A governance architecture that depends on layered review and escalation is only as strong as the judgement of the people operating it at the first line. If the quality of that judgement varies significantly across IFC’s regional footprint, then the ESRO second line carries a heavier burden than a small central function can easily sustain across a global portfolio.

There is also a subtler question about incentive structures. An E&S specialist embedded within a regional deal team understands the project better and engages earlier. The same specialist may also feel the pull of the team’s primary objective: getting the investment done. The manual is aware of this tension because it describes the ESRO function partly in terms of maintaining independence from that pull. Whether the architecture is strong enough to hold that line across all the contexts in which IFC operates, is something that only time and CAO’s caseload will reveal.

The Performance Standards revision currently underway focuses on what IFC requires its clients to do. The ESRP Manual is about something different: how IFC organises itself to ensure those requirements are applied. That question deserves as much scrutiny as the standards themselves. It has received considerably less.

There is a paradox at the heart of the energy transition that does not get nearly enough attention in policy circles. Getting to net zero requires a massive expansion of renewable energy — solar panels, wind turbines, electric vehicles, battery storage systems. All of these technologies depend on minerals: lithium, cobalt, copper, nickel, graphite, manganese. Getting those minerals out of the ground requires mining. And mining, wherever it happens and however it is managed, is one of the most socially and environmentally disruptive industrial activities that exists.

A April 2026 report from the UN University put this plainly: systemic global failures are allowing the costs of critical minerals extraction to fall disproportionately on some of the world’s most vulnerable communities, while the benefits accumulate elsewhere in the form of electric vehicles, renewable energy systems and AI infrastructure. The report is careful to note it does not question the need for clean energy. It questions who pays for it.

This is a real concern playing out in real time in specific places, and the E&S standards that are supposed to manage these impacts, including the IFC Performance Standards currently under revision, are being tested against a project pipeline they were not originally designed for.

1. The Scale of the Problem

The 2025 Transition Minerals Tracker, published by the Business and Human Rights Resource Centre, has been monitoring human rights risks linked to the mining of transition minerals since 2010. Between 2010 and 2024, it recorded 835 allegations of human rights and environmental abuse linked to the extraction of eight key minerals: bauxite, cobalt, copper, iron ore, lithium, manganese, nickel and zinc. The pattern it identifies is consistent: those most affected are workers, Indigenous Peoples, local communities and human rights defenders. Water contamination, loss of food security, destruction of surrounding ecosystems and the systematic bypassing of Free, Prior and Informed Consent (FPIC) feature across geographies and across minerals.

What makes the current moment distinctive is the speed and the geopolitics. As the SIRGE Coalition noted, mining is increasingly driven by national security imperatives, supply chain competition with China and the raw materials requirements of AI infrastructure and advanced defence systems. The European Commission’s 2025 Defence Readiness White Paper explicitly identifies critical raw materials as essential not just for climate goals but for drones and AI-based weapons platforms. In this environment, FPIC and community consent risk becoming inconvenient friction in a geopolitical race that nobody wants to lose.

The Serbian lithium saga illustrates the social licence dimension plainly. Rio Tinto's proposed Jadar mine in western Serbia, which would be Europe's largest lithium deposit, has been through two major waves of protests (2021 and 2024), a government cancellation of permits in 2022, a Constitutional Court ruling that the cancellation was unlawful in 2024 and a formal EU designation as a "strategic project" under the Critical Raw Materials Act in June 2025. In November 2025, Rio Tinto quietly placed the project on indefinite care and maintenance, redirecting resources to lower-risk assets. More than 60,000 protest signatures, years of community opposition and genuine environmental concerns about groundwater and soil contamination in a fertile agricultural valley were not the decisive factor — project economics were. As BankTrack has documented, the failure to build genuine community support (a basic requirement under PS1) was visible to investors throughout and contributed directly to the project's risk profile.

2. Where the Standards Fall Short

The IFC Performance Standards are the closest thing to a global baseline for E&S risk management in private sector infrastructure and extractives finance. They apply directly to IFC’s own investments, and indirectly (through the Equator Principles) to a large share of international project finance. They are, in that sense, the rules that govern how much of the critical minerals pipeline will be assessed and managed.

The problem is that the current PS were designed in 2012, around a project pipeline that looked quite different from the one emerging today. As the Bretton Woods Project has argued, IFC must use the PS revision to correct a failed top-down approach to communities affected by mining — one that has consistently prioritised developer timelines over community rights. The core critique is familiar: PS5 on involuntary resettlement requires livelihood restoration, not improvement. PS7 on Indigenous Peoples requires FPIC in defined circumstances, but its application has been inconsistent. PS6 on biodiversity requires no net loss in critical habitats, but its supply chain provisions are bounded by what clients can “reasonably” be expected to control.

A coalition of leading environmental and human rights NGOs, coordinated through Transport & Environment, has called explicitly for the PS revision to address transition minerals projects specifically, adopting a human rights-centred approach, strengthening supply chain due diligence and improving access to remedy for affected communities. The argument is straightforward: if IFC and the MDB community are going to finance the critical minerals pipeline, the standards governing that financing need to reflect what is actually at stake.

There is also a practical, non-ideological argument here that Global Witness made clearly in March 2025: mine delays and closures caused by community opposition and FPIC failures represent material losses for investors. Projects that fail to secure genuine community consent don’t just generate reputational risk — they generate project risk. The Jadar case is one example. There are others across Latin America, Africa and Asia-Pacific where the same pattern (inadequate early engagement, contested FPIC, escalating conflict, project delay or suspension) has played out at significant financial cost.

3. The Specific Gaps

Three areas in the current PS architecture are particularly exposed by the critical minerals challenge, and all three are on the table in the ongoing revision.

FPIC and Indigenous Peoples (PS7). The current PS7 requires FPIC for projects affecting Indigenous Peoples’ lands, resources and cultural heritage — but only where the project meets certain threshold conditions (e.g. physical relocation). In practice, as the Transition Minerals Tracker documents, disregard for FPIC is pervasive in transition minerals mining globally, exacerbated by economic incentives favouring rapid deal-making. The civil society ask for the PS revision to apply FPIC principles more broadly and consistently is a direct response to this pattern. Whether the revised PS7 will close the gap or merely clarify it will be one of the clearer tests of the revision’s ambition.

Resettlement and benefit-sharing (PS5). The current PS5 requires that affected people’s livelihoods be restored to at least their pre-project levels. For communities in extreme poverty which describes a significant proportion of those living near critical mineral deposits this means their post-resettlement situation may remain one of extreme poverty. There is no requirement in the current standards for communities to share in the financial benefits of a project built on their land. This gap has been flagged repeatedly by civil society and is one of the clearest misalignments between the PS and the MDB development mission.

Supply chain due diligence (PS1/PS6). The current PS address supply chain risks only where the client has “management control or influence” — a threshold that has historically been interpreted narrowly. For critical minerals, where the supply chain runs from artisanal mining communities in the DRC through to battery manufacturers in East Asia and vehicle producers in Europe, this boundary is inadequate. Growing mandatory due diligence regulation in the EU and elsewhere is raising the bar faster than the PS are moving.

The Hamburg Score View

The energy transition is not optional. The minerals it requires will be extracted. The question is under what conditions, with what protections for the communities and ecosystems in the path of that extraction, and whether the financial institutions providing capital for the transition are held to standards commensurate with what is at stake.

The IFC Performance Standards revision, due in draft form in Q3 2026, is the most significant opportunity in over a decade to answer that question with something more than the current framework provides. The signals from the consultation process — on FPIC, on resettlement, on supply chain diligence — suggest that the revision will move in the right direction. The question, as always, is how far.

There is also a harder observation worth making. The geopolitical pressure driving the critical minerals boom is not going to make any of this easier. When a contested mine is designated as a “strategic project” and permitting is fast-tracked, the implicit message to project developers is that community opposition is a problem to be managed around, not a signal to engage with seriously.

The revised PS will only be as effective as the supervision architecture behind them. That observation has appeared in every Dispatch on this Substack, because it remains true for every topic this revision touches. It is perhaps most true here.

Two recently adopted frameworks represent IFC’s most explicit attempt yet to address this: the Approach to Responsible Exit, published in October 2024, and the Interim Remedial Action Framework (RAF), approved by the Boards of IFC and MIGA in April 2025. Together, they form a paired accountability architecture that IFC did not have before. Whether that architecture functions as intended is a different question, and one that is already being tested in the field.

1. Why This Needed Addressing

The pattern that prompted these frameworks was documented with some care by CAO’s Responsible Exit: Insights from CAO Cases (December 2023), which analysed IFC’s approach to exiting investments through the lens of complaints handled over a decade. The findings were not flattering. IFC had a tendency to exit projects that became problematic, leaving communities in the middle of active accountability processes, without informing affected stakeholders and without ensuring that outstanding ESAP commitments had been addressed. In some cases, communities that had raised concerns with CAO found themselves abandoned mid-process when IFC sold its stake or allowed a loan to run off.

The 2020 External Review of IFC/MIGA E&S Accountability had already identified this as a systemic weakness and prompted IFC to develop both frameworks. That it took five years from that review to the RAF’s approval in April 2025 gives some indication of the institutional complexity involved.

The canonical example cited repeatedly in civil society submissions and CAO documentation is the Alto Maipo Hydroelectric Project in Chile. IFC divested from the project while construction failures resulting from inadequate due diligence continued to cause serious harm: tunnels bored through the Andes diverted three rivers, jeopardised the water supply of communities outside Santiago and destroyed glaciers. As CIEL notes, the fifth tunnel separation was reported just weeks before the Responsible Exit Approach was published. IFC’s exit did not end its responsibility in any meaningful sense; it simply ended its formal involvement.

2. What the New Frameworks Actually Say

The Responsible Exit Principles set out five considerations IFC must weigh before exiting an active investment. It should evaluate whether the project’s development impact has been achieved and is sustainable; assess the status of outstanding ESAP commitments and any E&S risks that could materialise after exit; use its contractual leverage to push clients to address those risks before departure; consider its institutional mandate and any precedential implications; and assess how its exit might affect co-investors it has mobilised. The Principles apply to active exits only; they do not cover normal loan repayments or pre-payments made without IFC’s consent.

The RAF adds the complementary obligation around remedy: as CIEL notes, it is the first explicit IFC/MIGA policy on remedy and its adoption formalises a commitment that IFC had previously resisted making. The RAF acknowledges that IFC and MIGA must play a role in the “remedial action ecosystem”, recognising, in the language of international law, that institutions that contribute to harm should contribute to remedy.

The two frameworks are complementary: responsible exit is the mechanism by which IFC manages its departure, and the RAF is the mechanism by which harm is addressed before, during and after that departure. In practice, however, BIC notes that the RAF does not reference the Responsible Exit Principles, leaving the relationship between the two frameworks undefined — a gap that matters when the decision to exit and the decision about remedy need to be made in the same breath.

3. What the Frameworks Do Not Say

Several significant limitations are already visible on the face of the documents and have been catalogued by civil society with precision.

The RAF avoids committing IFC to direct financial contributions towards remedy, acknowledging that remedy often requires financial resources, while declining to provide them. The justification offered is legal, operational and reputational risk. This leaves the RAF heavily reliant on client-side action, without strong enforcement mechanisms to compel clients who are unwilling to engage. The framework also excludes projects that have already closed, even where harm is documented — meaning that many communities with legacy grievances will find no entry point in the new architecture.

The Responsible Exit Principles leave their application to staff discretion on a case-by-case basis. BIC observes that IFC staff may face conflicts of interest when applying the Principles in cases where they are simultaneously responsible for determining whether to exit due to unresolved E&S issues. The discretionary approach, however understandable in a complex portfolio context, limits communities’ ability to hold IFC accountable for how the Principles are applied.

Perhaps most practically significant: IFC does not plan to disclose the Responsible Principles Analysis that will be incorporated into Project Completion Reports. As Recourse noted in a civil society statement ahead of the RAF’s approval, without transparency about which projects have been considered under the frameworks, what decisions were made, and whether communities regard the outcome as adequate remedy, it is very difficult to assess whether the frameworks are functioning as intended.

4. The First Real Test: Santa Rita

The Santa Rita Hydroelectric Project in Guatemala has emerged as the first live test case. As BIC documents, IFC holds a 35% equity stake in the project through an investment in a fund that financed the project developer. In 2014, CAO received a complaint from Indigenous Maya Q’eqchi’ communities citing the absence of proper Free, Prior and Informed Consent. Construction began in 2013, was halted due to conflict with affected communities, and has not resumed since. Over a decade later, the situation remains unresolved.

Santa Rita is an FI investment, channelled through a fund, in which IFC is two steps removed from the project developer. This is precisely the accountability gap that CAO has documented as the most persistent weakness in IFC’s portfolio. The RAF provides limited guidance on how to operationalise remedial measures through these complex financial instruments — and Santa Rita is testing that gap in real time. As of the time of writing, IFC has not yet applied either the RAF or the RE Principles to the project.

The Hamburg Score View

The Responsible Exit Principles and the Remedial Action Framework represent a genuine institutional step. IFC and MIGA are the first MDB to develop frameworks specifically on remedy and responsible exit — a fact that is both genuinely notable and a mild indictment of how long this took. The frameworks acknowledge obligations that IFC had previously avoided formalising, and they create at least some procedural basis for affected communities to point to when IFC’s behaviour falls short.

But the limitations are tangible. A remedy framework that declines to commit to financial contributions; responsible exit principles applied at staff discretion without public disclosure of the analysis; an exclusion of legacy harm; and a near-complete absence of guidance on FI sub-projects — these are not drafting imperfections that will be resolved in guidance notes. They reflect genuine institutional reluctance to accept the full implications of what responsible exit and remedy actually require.

The connection to the ongoing Performance Standards revision matters here. CIEL has argued explicitly that the PS revision is the right moment to embed both the RAF and the RE Principles into the standards themselves, rather than leaving them as supplementary policy instruments. Civil society has made the same point. Whether IFC agrees will be visible in the draft text expected later in 2026.

In the meantime, Santa Rita is watching. So is every other community that filed a CAO complaint and then watched IFC sell its stake.

1. The Structural Reorganisation: What Has Actually Changed

The most visible element of the One World Bank push is operational. From January 2026, the knowledge teams of IBRD/IDA and IFC merged into a single unit organised across five thematic verticals: People, Prosperity, Planet, Infrastructure and Digital. Treasury operations, HR, IT, budget, real estate, communications and eventually legal functions are being unified across the Group. Leadership in country offices has been consolidated, with a single joint representative replacing separate public and private sector presences.

Banga has been consistent about what this is not: the five institutions — IBRD, IDA, IFC, MIGA and ICSID — remain legally separate, with their own balance sheets, governance structures and debt issuance arrangements. This is rather integration of function than merger of legal entity.

For E&S practitioners, the operational changes have a specific implication: the public and private sector E&S regimes — the World Bank’s Environmental and Social Framework (ESF) on one side and the IFC/MIGA Performance Standards on the other — are products of distinct institutional cultures, distinct legal obligations and distinct client relationships. Merging the people who work on them into shared thematic verticals does not, by itself, align the standards. But it creates the organisational conditions under which alignment becomes easier to pursue.

IFC’s own Approach Paper for the Performance Standards revision lists convergence with the World Bank ESF as an explicit objective. Whether this convergence produces stronger or weaker protections is one of the more consequential open questions in MDB standards at the moment.

2. The E&S Standards: Convergence in Progress

The ongoing revision of IFC Performance Standards is the primary vehicle through which E&S alignment is being pursued. The February 2026 Dispatch on this Substack covered that process in detail; a brief update is warranted here.

The April 2026 deadline for a first public draft has passed without a draft. Civil society tracker Recourse now places the first formal draft at Q3-Q4 2026 at the earliest. Phase II public consultations are formally underway, but the substantive text remains with IFC.

What the convergence agenda actually means in practice is still being defined. IFC’s Approach Paper signals thematic alignment on resettlement, biodiversity, Indigenous Peoples, vulnerable groups, gender and GBV, labour, associated facilities and cumulative impacts — areas where the ESF has somewhat different (and in some cases more explicit) requirements than the current PS. The direction of travel suggested by civil society demands and CAO advisory outputs is towards more prescriptive requirements in several of these areas, not less.

The civil society “non-regression” principle articulated clearly in the Bank Information Center coalition letter of December 2024 is partly a response to the convergence agenda itself. If alignment with the ESF means adopting ESF language that is more flexible on borrower systems and risk-based thresholds, then convergence could mean dilution. Whether that concern proves justified will only be apparent when draft text is available.

On 14 April 2026, OECD Watch led a joint sign-on letter from over 50 organisations calling for the revised PS to align with the updated OECD Guidelines for Multinational Enterprises on responsible business conduct. The letter explicitly frames three parallel processes (the IFC PS revision, new OECD guidance on development finance and the One World Bank integration) as a single moment of opportunity for policy coherence.

Whether IFC treats it as such remains to be seen.

3. The Accountability Mechanisms: The Most Consequential Thread

The least-covered but arguably most consequential element of the One World Bank integration (for those who care about outcomes rather than text) is the proposal to merge the Group’s independent accountability mechanisms.

At present, the World Bank Group runs two separate systems. The World Bank Accountability Mechanism handles complaints from communities affected by IBRD/IDA-financed projects; it comprises the Inspection Panel (compliance reviews) and the Dispute Resolution Service. The Compliance Advisor Ombudsman (CAO) handles complaints relating to IFC and MIGA projects and has three functions: dispute resolution, compliance review and advisory work. The CAO has over the years built a considerable body of advisory output — on climate, on financial intermediaries, on remedy — that has directly shaped the IFC PS revision process. It is one of the more effective accountability mechanisms in the MDB landscape.

The Task Force on Integration of World Bank Group Accountability Mechanisms, established by the Boards of the World Bank, IFC and MIGA, published a draft report for public consultation on 26 March 2026. CAO published its own comments on that draft on 24 April 2026, only last week. The public consultation on the draft report closes in the coming weeks.

The civil society concern is specific and well-founded. As one participant at the October 2025 Annual Meetings discussion put it plainly: CAO has a notably low eligibility bar — it is genuinely accessible to affected communities in a way that the Inspection Panel historically has not been. Integration could mean that CAO’s accessibility is extended across the Group which would be a meaningful improvement. It could equally mean that the eligibility threshold is set somewhere between the two existing mechanisms which would represent a regression for IFC/MIGA complainants. The Task Force has been asked to adhere to principles including no regression and no dilution, but those principles require watching.

The timing is pointed. The accountability mechanism integration process is running in parallel with the IFC PS revision, with the IFC Remedial Action Framework (adopted April 2025) still being assessed for adequacy, and with CAO’s advisory and monitoring work continuing to document implementation gaps. Compressing or diluting the accountability architecture at precisely the moment when the standards themselves are being renegotiated is a combination that civil society has flagged as concerning, and not without reason.

The Hamburg Score View

The One World Bank Group initiative is a sensible response to organisational fragmentation. A Group that presents a single face to clients in country, pools its knowledge across public and private sector work and stops duplicating corporate functions is probably more useful than one that does not.

The E&S implications are less tidy. Three processes — the operational integration, the IFC PS revision and the accountability mechanism merger — are running simultaneously, each with its own timeline and its own set of stakeholder pressures. They are connected in ways that matter: the standards being revised will be supervised by a mechanism that is itself being redesigned, and both are being shaped by an institutional culture that is (at least at the level of presidential messaging) more focused on private capital mobilisation and speed of delivery than on the safeguard architecture that took thirty years to build.

None of this means the outcome will be regressive. The IFC PS revision has strong institutional momentum towards strengthening climate, FI and rights provisions, and there is a credible civil society and CAO-led evidence base pressing in that direction. The accountability mechanism integration could, if designed well, genuinely improve access to remedy across the Group.

But “designed well” is doing a lot of work in that sentence. The decisive variable — as with the PS revision — will be whether the institutions responsible for implementing and overseeing these standards are structured, resourced and genuinely incentivised to do so.

That question has not yet been answered. Q3 2026, when the first PS draft is expected, will begin to answer part of it. The accountability mechanism decision will answer another part. Watching how the two land together seems worth the effort.

Because IFC has committed to publishing only aggregated, de-identified summaries of stakeholder comments, a precise mapping of who said what to IFC during Phase I is not yet possible. What is possible is a structured reading of the public record: IFC’s own framing documents, Compliance Advisor Ombudsman (CAO) advisory outputs prepared specifically for the review and stakeholder submissions that have been independently published. That is what this brief attempts.

1. Process Architecture

IFC has structured the update as a two-phase process (see the Approach Paper, April 2025, and the Stakeholder Consultation Framework):

Phase I (Dialogue): Thematic and stakeholder engagement, concluding Q1 2026.

Phase II (Public Consultation): Release of first draft for global comment, anticipated April 2026 onward, running into 2028.

IFC’s Approach Paper signals that the update will address: modernising PS language; clarifying the interface with national law; convergence with the World Bank’s Environmental and Social Framework (ESF); integration of lessons from CAO investigations; and potentially more agile mechanisms for future updates via Guidance Note revisions. That scope already signals this is not a cosmetic exercise.

The FAQ document and the October 2025 CSO information session deck in which IFC/MIGA summarised high-level civil society feedback by theme are the most substantive publicly available windows into Phase I at this point.

2. The Major Pressure Points

2.1. Climate and GHG Integrity

This is the area where institutional signals and stakeholder demands converge most clearly and where the evidence base is most robust.

The CAO’s climate advisory note (2024) prepared explicitly to inform the Sustainability Framework review concludes that the current Performance Standards do not provide a sufficient basis for reducing GHG emissions, pointing to inadequate carbon sink protection and outdated reporting commitments. A letter from 28 civil society organisations (January 2025) calls for full Scope 1-3 emissions accounting, robust alternatives analysis and independent oversight. The IFC/MIGA summary deck reflects civil society requests for a standalone climate standard, stricter GHG accounting and third-party auditing.

IFC has also committed publicly to Paris alignment across its operations, which makes it difficult to leave the climate provisions of its core standards materially unchanged.

Assessment: High confidence that climate provisions will be materially strengthened, either through a standalone climate standard or significant restructuring of PS1/PS3, with expanded Scope 3 triggers, stronger disclosure expectations and tighter requirements for an alternatives analysis.

2.2. Financial Intermediaries

The FI accountability gap is the most persistent fault line in the PS architecture, and this revision cycle has produced the most focused institutional attention to it yet.

The CAO’s monitoring report on FI E&S risk management (July 2025) documents ongoing weaknesses in sub-project oversight and questions whether IFC’s current approach is delivering “do no harm” outcomes at the sub-project level. The IFC/MIGA summary deck reflects civil society calls for a standalone FI standard, enhanced sub-project transparency and independent monitoring for high-risk sub-projects. Inclusive Development International’s recommendations paper (2020) explicitly supports stronger disclosure of grievance mechanisms and CAO access for FI sub-projects.

Assessment: High confidence that FI provisions will tighten materially, whether through a standalone standard or reinforced requirements, particularly emphasising transparency and independent oversight.

2.3. Remedy, Responsible Exit and Accountability

Remedy has shifted from a peripheral topic in previous cycles to a central one. CAO’s Remedy Gap advisory note (2023) and its Responsible Exit insight series frame exit during active harm cases as a systemic problem. IFC’s own Approach Paper references building on recent measures to strengthen accountability and mainstream a responsible exit approach which signals that some of this work may be integrated rather than left entirely to supervision practice.

The Bank Information Center coalition letter (December 2024) signed by multiple civil society organisations anchors the process-level demand: non-regression as a baseline, improved compliance transparency and publication of monitoring outcomes.

Assessment: High confidence that supervision and accountability architecture will be strengthened. Lower confidence that remedy obligations will be deeply embedded in PS text itself; more likely that changes sit in the Sustainability Policy, supervision practice and disclosure expectations.

2.4. Rights Framing: Labour, Land and Indigenous Peoples

The language of “rights-holders” is increasingly visible in public commentary, though its translation into enforceable PS obligations remains uncertain.

The IFC/MIGA summary deck reflects civil society calls for stronger FPIC application, customary land rights protection, resettlement avoidance and improved livelihood restoration, with resettlement explicitly identified in the Approach Paper as an ESF convergence topic. The ITUC statement (December 2024) characterises the PS2 situation in stark terms, arguing that employers on IFC projects have violated the standards with impunity and that workers have been left without recourse. The ITUC calls for stronger freedom of association protections, improved worker grievance access, whistleblower safeguards and clearer alignment with ILO conventions. The Bretton Woods Project’s commentary (October 2025) links the rights-holder framing to just transition mineral supply chains, where Indigenous rights and FPIC disputes are common.

Assessment: Medium–high confidence of strengthened language across labour rights, resettlement and Indigenous Peoples provisions. The degree to which new obligations become genuinely enforceable (rather than clarified in guidance) is the main open question.

2.5. Biodiversity, Supply Chains and Scope

IFC’s Approach Paper names biodiversity loss and ecosystem degradation as core global challenges the updated framework should address, and lists biodiversity as an ESF convergence topic alongside associated facilities and cumulative impacts. The IFC/MIGA summary deck reflects civil society calls for stronger critical habitat protection, enhanced supply-chain due diligence and recognition of culturally significant ecosystems. A Friends of the Earth report (November 2025) positions the Sustainability Framework review as an opportunity to correct existing omissions in biodiversity protection.

The supply-chain dimension is reinforced by external regulatory pressure. The growing body of mandatory due diligence regimes and deforestation rules in major markets creates alignment incentives that go beyond stakeholder advocacy.

Assessment: Medium–high likelihood of tightening around biodiversity and cumulative impact boundaries, though implementation may fall partly through Guidance Notes rather than PS text revisions.

2.6. Transparency and Access to Information

The Access to Information Policy (AIP) is being revised in parallel with the PS, and the two are explicitly linked in IFC’s communications. The Law and Democracy submission (October 2025) provides detailed comparative analysis of right-to-information practice and recommendations for improving IFC’s AIP. Civil society feedback (as summarised in the IFC/MIGA deck) calls for broader proactive disclosure, narrower confidentiality exceptions, better FI sub-project transparency and improved handling of information requests and appeals.

IFC acknowledges a tension in the Approach Paper: strengthening transparency while respecting the confidentiality expectations of private sector clients. That tension will likely shape what is achievable on FI sub-project disclosure in particular.

Assessment: High probability of AIP strengthening on proactive disclosure and definitional clarity. The extent of FI sub-project disclosure remains politically sensitive and is likely to be negotiated hard.

3. The Equator Principles Dimension

One constituency that has not published detailed position papers but whose interests are structurally significant: the Equator Principles signatories. The 2024 Equator Principles Activity Report notes that the PS remain the foundational benchmark for EP implementation. This means the revision has implications well beyond IFC’s own portfolio. Changes in scope, thresholds or requirements will ripple through project finance globally. Signatories have strong incentives to support predictability and interoperability which both reinforces the case for ESF alignment and creates a constituency for manageable implementation burdens.

4. Directional Assessment

The evidence currently available supports the following trajectory:

High-confidence changes:

• Expanded climate/GHG requirements: Scope 1-3 clarification, stronger disclosure and a likely structural upgrade (standalone standard or dedicated PS1/PS3 module).

• Tightened FI governance: a standalone FI standard or materially reinforced requirements, with independent monitoring for high-risk sub-projects.

• Strengthened AIP: broader proactive disclosure and clearer definitions.

• Greater clarity around accountability, supervision and responsible exit.

Medium-confidence changes:

• Hardening of resettlement and labour language, with clearer ILO alignment.

• Stronger biodiversity and supply-chain expectations.

• Clearer treatment of cumulative impacts and associated facilities.

Lower-confidence or politically sensitive areas:

• Deep embedding of remedy obligations directly into PS text.

• Radical redefinition of communities as rights-holders with formal decision-making authority beyond current PS architecture.

• Full public disclosure of FI sub-project information.

The direction does not suggest deregulation bur rather targeted hardening in the areas where implementation gaps have become most publicly visible.

The Hamburg Score View

The revision is not being driven primarily by textual ambiguity in the current standards. It is rather being driven by institutional credibility.

CAO findings, Paris alignment commitments, persistent scrutiny of FI portfolios and the interoperability agenda with the World Bank ESF have together created a set of pressure points that are now difficult to defer. The forthcoming draft will likely reflect more explicit climate governance; firmer intermediary accountability; clearer disclosure architecture; and sharpened rights language in labour, land and Indigenous Peoples provisions.

The decisive question, however, will be whether supervision, disclosure and remedial mechanisms are resourced and structured to match what the text says. That gap (between policy and practice) is precisely what CAO’s advisory work has been documenting for years, and it is the real test of whether this revision produces outcomes or just better language.

We will know more in April.

And yet, not all development banks operate this way.

This new Hamburg Score Briefing looks at four institutions — the Saudi Fund for Development (SFD), the Kuwait Fund for Arab Economic Development (KFAED), the Abu Dhabi Fund for Development (ADFD) and the China Development Bank (CDB) — that finance large-scale infrastructure across regions without publishing standalone safeguard frameworks comparable to those of the World Bank, IFC, EBRD or ADB.

The question is not whether environmental and social issues are considered at all. The question is how risk governance is structured when safeguards are not codified as institutional policy.

Key Findings

• The three Gulf-based funds rely primarily on host-country regulatory systems, project-level instruments (ESIAs, ESMPs) and the safeguards of co-financiers where joint financing occurs. They do not publish portfolio-wide E&S standards with binding thematic requirements.

• Sustainability language is present in strategic communications, but issues such as resettlement, labour standards, biodiversity or stakeholder engagement are not governed through publicly articulated institutional benchmarks.

• China Development Bank operates under China’s domestic Green Credit Guidelines and sustainability disclosure rules. These mandate internal environmental risk management and reporting. However, for overseas lending, CDB does not publish a project-level safeguard framework comparable to MDB standards.

• In co-financed transactions, safeguard-based MDBs effectively supply the E&S architecture. In standalone operations, risk management is shaped primarily by national law and internal discretion.

• The result is a dual system within development finance: one model built on codified, publicly disclosed safeguards; another anchored in national regulation, internal procedures and transaction-specific arrangements.

The Hamburg Score view

This is not a question of rhetoric or intent. It is a question of institutional design.

Where safeguards are codified, they create predictability, comparability and a shared reference point for borrowers, communities and co-financiers. Where they are absent as formal architecture, E&S governance becomes contingent and dependent on jurisdiction, partnership structure and internal practice.

Both models operate within the global development finance landscape. They reflect different governance choices.

As multilateral institutions continue to update and expand their safeguard frameworks, the distinction between codified and contingent E&S governance will remain visible.

The policy narrative across these institutions is orderly and sustainability-oriented.

The Hamburg Score points to a different structural logic beneath.

The Hamburg Score Briefing 3 No Policies Feb 2026

2.41MB ∙ PDF file

Download

Download

.

Most E&S practitioners are familiar with the general concept. What is less often acknowledged is that very few people agree (sometimes even between different international finance institutions) on where the boundary actually lies.

This Dispatch looks at how the major multilateral development banks currently define associated facilities, how those definitions are applied in practice and why the real difficulty has less to do with wording and more to do with judgement, leverage and the absence of technical guidance.

The current definitions

It is worth starting with the definitions themselves, if only to clear away a common misconception: the main MDBs are not saying radically different things.

The World Bank Environmental and Social Standard 1 defines Associated Facilities as facilities or activities that are not financed as part of the project and that are:

• directly and significantly related to the project;

• carried out, or planned to be carried out, contemporaneously with the project; and

• necessary for the project to be viable and would not have been constructed, expanded or conducted if the project did not exist.

All three criteria must be met.

The IFC Performance Standard 1 uses a slightly shorter formulation, but the logic is the same. Associated facilities are those that are not funded as part of the project, that would not have been constructed or expanded if the project did not exist, and without which the project would not be viable. IFC provides (in a footnote) a non-exhaustive list of examples: roads, railways, pipelines, transmission lines, utilities, logistics terminals.

The Asian Development Bank, in its new Environmental and Social Framework, mirrors the World Bank almost word for word: facilities not funded as part of the project, directly and significantly related, developed contemporaneously, necessary for viability and absent in the counterfactual where the project does not exist.

The EBRD’s E&S Requirement 1 phrases things slightly differently, referring to facilities or activities not financed by the Bank but which are significant in determining the success of the project or in producing agreed project outcomes, and which would not be planned or carried out without the project and without which the project would not be viable.

Up to this point, the story is fairly straightforward. Different drafting styles, the same underlying test: counterfactual dependence and project viability.

The EIB, however, has moved in a different direction in its latest Environmental and Social Standards adopted in 2022. The current standards no longer offer a standalone definition of “associated facilities” at all. Instead, they require the assessment of impacts and risks to take into account, where relevant, ancillary or associated works that form an integral part of the project, supporting or enabling activities under the control of contractors, and facilities owned by separate legal entities without which the project would not be technically viable.

This seems to reflect a conscious shift away from categorical labelling and towards impact scoping.

Too much leeway?

If one reads these definitions carefully, it becomes hard to argue that divergent practice is driven by divergent policy language. The core elements are shared across institutions.

The difficulty starts after a facility has been identified as “associated” in principle.

That difficulty usually crystallises around a deceptively simple question: Would this have existed anyway?

That question does an extraordinary amount of work. In straightforward cases, it behaves well.

A dedicated transmission line to a mine.

A captive power plant serving a single industrial facility.

A project-specific access road cut through otherwise undeveloped land.

Are there less straightforward cases? Yes, plenty. And in those cases, the question begins to fray.

Borderline cases

Most practitioners will recognise some version of the following (all taken from my own practice — and in reality there are many more):

• A port terminal financed by the state and described as “strategic infrastructure”, delivered with admirable punctuality just in time for a single private project that will use most of its capacity for the first decade.

• A regional access road justified as part of a broader development programme, whose alignment, timing and specifications happen to coincide perfectly with the needs of a single project sponsor.

• A “temporary” construction camp that slowly acquires permanence: housing, clinics, schools, access roads, utilities — and a resident population whose impacts outlast the project that brought them there.

• A shared waste facility serving multiple users, one of whom produces most of the waste, most of the risk and most of the grievances.

In most of these cases, the technical analysis is not especially difficult. Everyone in the room understands what is going on. The argument is mostly about responsibility.

An application issue

This is where institutional practice starts to separate.

The World Bank explicitly links requirements to the borrower’s level of control or influence. Where control is limited, the Bank allows for a graduated response (assessment and mitigation to the extent feasible, proportionate to leverage).

IFC places stronger emphasis on the client’s responsibility to seek to ensure that associated facilities meet the Performance Standards, even where the client does not own or control them. In practice, this often translates into negotiated commitments, information-sharing arrangements or contractual covenants.

The EBRD tends to frame associated facilities through the lens of “reasonable efforts” and outcomes. If an associated facility is outside the client’s control, the focus often shifts to whether risks can be understood and managed sufficiently to protect project outcomes, rather than on strict compliance.

ADB, particularly in its updated framework, has moved towards greater clarity and consistency, but still relies heavily on professional judgement in determining how far requirements should extend in cases of weak leverage.

The EIB, by contrast, largely sidesteps the classification debate. Rather than asking whether something “is” an associated facility, it asks whether its impacts should reasonably be included in the project’s assessment scope. Where that answer is yes, impacts are assessed. Where leverage is minimal and impacts are remote, facilities may simply fall outside the assessed footprint.

The leverage problem (and why it keeps resurfacing)

At some point, associated facilities force an uncomfortable conversation.

Private developers are often asked by lenders to manage E&S risks arising from facilities they do not own, do not operate and cannot meaningfully influence, the facility owners frequently being state-owned utilities or public infrastructure operators with limited transparency and weak accountability.

Is this a developer responsibility? A lender responsibility? A joint exercise in pragmatism?

Most MDB E&S frameworks gesture towards “best efforts” or “reasonable influence”. But few really explain what that looks like when leverage is close to zero and risks are not.

The missing technical layer

What is striking, once one steps back, is that there is no shared technical guidance on how to delineate associated facilities in practice. Indeed, we have all sorts of guidelines on adjacent matters — detailed instructions on ESIA structure and content; sector-specific Environmental, Health and Safety Guidelines; step-by-step manuals on E&S management for financial intermediaries; increasingly elaborate guidance on climate risk and resilience; human rights impact assessment toolkits; stakeholder engagement handbooks; labour and gender notes; biodiversity offsets frameworks; cumulative impact good practice handbooks.

Yet when it comes to one of the most routinely contested questions in project appraisal — where the project boundary actually lies — practitioners are largely left to work it out for themselves. Current practice relies on precedent, negotiation, independent consultant judgement and institutional risk appetite. For Equator Principles projects, the boundary is often settled through discussion between sponsors, lenders and IESCs. For MDB projects, it is frequently resolved case by case, project by project.

The Hamburg Score view

Associated facilities are not difficult because they are poorly defined. They are difficult because they expose the limits of policy when confronted with real-world governance, ownership and power.

They remain one of the most reliable stress tests of how E&S safeguards actually operate — and not on paper, but in negotiation rooms, appraisal missions and supervision visits.

If there is a need for targeted technical guidance in the next phase of safeguards evolution, associated facilities would be a good place to start.

The decision to extend the consultation period by a further two weeks suggests an institution keen to be seen as inclusive and receptive, and perhaps still calibrating the level of external engagement it wants around its safeguards.

That alone is notable. But to understand what this update might realistically deliver, it helps to step back and look at where the CEB is starting from and where its peers already are.

A policy shaped by 2016

The current Environmental and Social Safeguards Policy was adopted in November 2016. It reflects a moment when most MDB safeguard frameworks were still primarily principle-led rather than system-led.

The policy is compact, principled and firmly anchored in European norms. It draws on Council of Europe human rights instruments, EU environmental law and the European Principles for the Environment. In tone and ambition, it is clear and careful.

What it does less clearly is explain how those principles are converted into consistent, observable behaviour at project level over time. Due diligence is proportional, disclosure is largely appraisal-stage, and monitoring is framed as a shared responsibility rather than a governance function.

That was not unusual in 2016. It is more unusual now.

A late mover in a crowded safeguards landscape

By the time the CEB opened this consultation, several peers had already completed (or embarked upon) major safeguards revisions. The World Bank, EBRD, ADB, EIB and IFC have all moved towards more explicit treatment of climate alignment, disclosure and life-of-project accountability.

The CEB is therefore deciding how far to follow.

That makes this update more interesting, not less. Late movers have the benefit of hindsight but also the temptation to modernise language without fully absorbing what those earlier updates changed in practice.

Why the consultation matters

The most significant signal here is not the promise of “alignment”, but the decision to consult publicly at all.

This is the first time the CEB has invited external stakeholders to comment on a core safeguards policy. The consultation questions themselves are very precise and focused on clarity, relevance, effectiveness and gaps — not on redefining ambition.

That framing suggests an institution testing its footing. It signals awareness that safeguards today are judged not only by what they say, but by whether they can be seen to work.

In that sense, the consultation is less a normative exercise than a question of institutional legitimacy.

The Hamburg Score view

From a Hamburg Score perspective, the key issue is not whether the revised policy (and any associated standards) will look familiar. They almost certainly will.

The real test will be whether the update strengthens the operational spine behind the principles.

When the draft revised ESSP appears in 2026, four things will be worth watching particularly closely:

• Does the policy clearly explain how gaps between national compliance and Bank standards are identified and addressed?

• Are disclosure and monitoring treated as life-of-project obligations, rather than appraisal-stage formalities?

• Is Paris alignment operationalised (with consequences), rather than acknowledged in principle?

• Is reliance on discretion reduced where predictability would improve accountability (especially for borrowers and intermediaries)?

If the update mainly refreshes the vocabulary of the 2016 policy, it will signal continuity. If it tightens these operational links, it will mark a genuine shift in how the CEB understands safeguards.

Either way, the consultation, albeit modest and late, tells us that the CEB knows the question is no longer whether to update its safeguards, but what such an update is meant to achieve. And that, for a bank like this, is already something worth paying attention to.

And yet, try to trace those safeguards over time.

This is the puzzle at the heart of BSTDB’s E&S framework. Information appears at the moment it is procedurally required, then quietly disappears. Projects move from appraisal to implementation — but their E&S commitments leave little public trace behind.

This new Hamburg Score Briefing examines BSTDB’s safeguards not by asking whether they exist, but how they function over the life of a project, and what happens once decisions are taken.

Key Findings

• BSTDB operates with a single, principle-based Environmental and Social Policy and an exclusion list, but no binding performance standards or sector-specific requirements.

• Category A projects are assessed and consulted on prior to approval, yet E&S documentation is disclosed only as part of time-bound consultation processes, not as part of a permanent project record.

• Once financing decisions are taken, public access to E&S information rapidly diminishes, limiting accountability during construction and operation when impacts typically materialise.

• E&S risk management is largely delegated to national systems and co-financiers, without systematic gap-filling where regulatory capacity or enforcement is weak.

• Where more prescriptive co-financiers are present, robust safeguards appear; where BSTDB acts alone, E&S commitments are often difficult to verify post-approval.

The Hamburg Score view

BSTDB’s safeguards system is largely procedural. It is designed to legitimise decisions at the point of approval, rather than to sustain transparency and accountability throughout the project lifecycle. This is a governance choice that prioritises flexibility, delegation and institutional restraint over standardisation and retention.

As regional development finance grows more prominent in politically complex environments, the question is no longer whether safeguards are disclosed — but whether they endure.

The Hamburg Score Briefing 2 Bstdb Jan 2026

2MB ∙ PDF file

Download

Download

Less visibly, ADB has just disclosed the full set of Guidance Notes supporting the ESF. So the architecture is now complete, with the ESF setting out what is required, and the Guidance Notes explaining how those requirements are expected to be applied in practice.

Taken as a whole, this is a significant institutional shift. The ESF moves ADB decisively towards the now-familiar MDB model: a single, integrated risk classification; clearer separation of ADB and borrower responsibilities; explicit attention to contextual risks (including digital and civic space issues); a formal ESCP/ESAP architecture; and much stronger treatment of disclosure, Sexual Exploitation, Abuse and Harassment (SEAH) issues and stakeholder engagement.

And yet, once the Guidance Notes are read alongside the ESF, the most interesting questions are about implementation behaviour: how the new system will operate in complex, constrained or politically charged project environments.

What follows are six areas where the framework is clear in intent, but where implementation behaviour is still not fully spelled out.

1. Adaptive management

ADB formalises adaptive management through the ESCP/ESAP and explicitly allows some E&S work to be completed during implementation rather than before approval. That flexibility comes with conditions: deferral is only allowed where it is unlikely to alter the project’s E&S conclusions and does not undermine overall E&S readiness. This is a sensible and clearly stated safeguard.

The unresolved issue is that this still leaves a large middle ground where judgement calls will matter enormously:

• What, in practice, counts as an “E&S conclusion” that must not change?

• How will “readiness” be demonstrated (e.g. in sovereign programmes, FCAS contexts or politically sensitive projects)?

• What happens when material impacts emerge late but addressing them would disrupt delivery?

The ESF also adds a second safeguard: where actions are deferred into the ESCP/ESAP, borrowers must not proceed with activities that could cause material adverse impacts until those measures are completed.

This really matters. The question is whether it becomes a real brake on premature implementation or whether it ends up as legal reassurance on paper. Anyone who has worked on implementation knows how easily construction schedules become the dominant reality, with the ESCP struggling to keep pace. The ESF is clearly trying to resist that dynamic. Whether it succeeds will be one of the early tests of the new system.

2. Audits of projects with existing permits

Where projects are already permitted, under construction or supported by national EIA processes, ADB requires an audit to assess alignment with the ESSs, with additional studies, consultation or disclosure where gaps are identified.

This is where MDB standards most often collide with reality.

Once permits are in place, designs fixed and contractors mobilised, the practical scope to revisit alternatives, reroute infrastructure or fundamentally redesign mitigation can be very limited. In many countries, national EIA processes are also highly procedural: compliant on paper but weak on social and cumulative impacts, associated facilities, labour supply chains, biodiversity baselines or meaningful consultation.

The ESF is right to require an audit. What it does not fully answer is how far ADB will push when audits show that national compliance is not the same as ESS compliance.

That will be decided case by case. Which is precisely why this remains an unresolved but consequential gap.

3. Contextual risk classification

One of the strongest advances in the ESF is its explicit inclusion of contextual risks in classification: fragility and conflict, governance constraints, capacity limitations, climate and ecosystem vulnerability, sector-specific risks, and, importantly, digital risks and civic space constraints affecting stakeholder engagement.

The missing piece is operationalisation:

• What evidence will be considered sufficient to assess civic space risk?

• How will digital risks be evaluated in a way that goes beyond box-ticking, especially where national frameworks are weak?

• How will these risks translate into staffing, supervision intensity, or real “stop/go” decisions?

The ESF states that resources will be allocated proportionate to risk and that classifications will be reviewed over the project cycle. The open question is whether contextual risks will genuinely shape project handling, or remain acknowledged but largely inert.

4. “Materially consistent” systems and “most stringent” common approaches

ADB places greater emphasis on the use of borrower systems and common approaches with co-financiers, anchored in the idea of achieving objectives “materially consistent” with the ESSs.

At the same time, the ESF states that common approaches should apply the requirements that are “most stringent or protective” of people and the environment.

Individually, both principles make sense. Together, they create a practical tension:

• “Material consistency” allows flexibility in how outcomes are achieved.

• “Most stringent” implies a hierarchy that may pull projects towards the highest standard in the room, even when standards are structured differently or excel in different areas.

The ESF envisages early agreement and performance tracking through the ESCP/ESAP. That works neatly on paper. What remains unresolved is what happens when common approaches become negotiated compromises and who ultimately concludes when protective requirements pull in different directions.

5. Disclosure and consultation rules

The ESF sets out disclosure timelines with welcome clarity:

• For High, Substantial and Moderate risk projects: disclosure as early as possible, and no later than appraisal or final credit approval.

• For High Risk projects: ESIA disclosure at least 120 days (sovereign) or 60 days (non-sovereign) before Board consideration.

These timelines are reinforced in the ESS10 Guidance Note.

At the same time, the ESF allows certain E&S documents to be prepared and disclosed after approval if this is agreed through the ESCP/ESAP.

So the real question is not about the rules but about how they are used. Applied narrowly, deferral is a pragmatic tool. Applied routinely, it risks hollowing out meaningful consultation by shifting key information into a phase where decisions are already effectively locked.

The ESF attempts to guard against this by requiring that disclosed drafts still provide enough detail to inform engagement and decision-making. But “adequate” is a flexible standard, and one that will be tested early.

6. SEAH’s scope and capacity constraints

ADB deserves credit for its treatment of SEAH. It is clearly defined in the ESF and linked across multiple standards. The Guidance Notes go further than many frameworks, setting out expectations on risk identification, contractor accountability, codes of conduct, training, survivor-centred referral pathways and dedicated grievance handling.

Two limitations remain visible even with this added detail.

First, the ESS4 Guidance Note limits SEAH assessment to risks involving project workers and interactions between workers and affected communities. That boundary may be too narrow in contexts where SEAH risks are mediated through informal labour, third-party actors, security forces or wider community dynamics.

Second, effective SEAH systems depend on local services and trusted reporting channels. Guidance can require referral systems, but it cannot create capacity where none exists.

The framework is materially stronger than before. Whether it delivers meaningful protection will depend on whether SEAH is treated as real risk management, rather than compliance formality.

The Hamburg Score view

Taken together, the ESF and Guidance Notes reflect serious institutional effort. The system is recognisable, modern and broadly well aligned with MDB practice: integrated risk logic, contextual analysis, adaptive management, clearer disclosure rules, greater use of borrower systems and stronger social protections.

What remains unresolved are rather the pressure points where discretion, incentives and capacity will shape outcomes.

In the first year of implementation, I would watch closely:

• how often deferrals appear in ESCPs/ESAPs and how lightly they are used;

• how tough audits are on already-permitted projects;

• whether contextual risks meaningfully influence classification and resourcing;

• how “material consistency” is interpreted in borrower systems and common approaches;

• whether post-approval disclosure remains the exception; and

• whether SEAH requirements translate into functioning systems on the ground.

That is where the ESF will either settle into familiar MDB comfort — or prove itself as a genuinely stronger instrument for E&S outcomes.

But frameworks matter as much for what they exclude as for what they enable.

A careful reading of the TFF shows a document that is deliberately narrow by design, cautious in its accounting, and selective in what it is prepared to count, claim or accelerate. That selectivity does shape where capital can — and cannot — flow under Deutsche Bank’s transition banner.

Below are five things the Transition Finance Framework quietly leaves out, and why each matters.

1. Entity-level transition finance is explicitly excluded from the headline target

The most consequential exclusion is also the most clearly stated, albeit easily missed.

Deutsche Bank defines three parameters of transition finance:

• Parameter 1: Activity-level transition finance

• Parameter 2: Entity-level transition finance

• Parameter 3: Sustainability-linked solutions

Yet only Parameters 1 and 3 will count towards the €900bn cumulative target from January 2026 onwards. Entity-level transition finance (Parameter 2) is not counted towards the target.

This seems to be a deliberate structural choice in how the target is defined.

Entity-level transactions (general corporate purpose financing for companies pursuing a credible transition strategy) are acknowledged as a legitimate category of transition finance but they are kept out of the target calculation. Deutsche Bank states that it intends to report these volumes separately once a “system-based solution” is in place.

This excludes a large class of real-world transition activity: companies whose capital expenditure, emissions trajectory and governance reforms point towards transition but where funds are not tied cleanly to a specific “transition activity”.

In effect, the message here is, “We recognise this finance, but we will not count it (yet)”.

A credibility safeguard, yes. But also a constraint.

2. Transition pathways without measurable activity anchors do not qualify

Closely related to the first point is what the framework does not reward. It does not reward broad, narrative-led transition strategies without transaction-level anchors.

For activity-level transition finance (Parameter 1), Deutsche Bank requires financing to be linked to specific activities that:

• are not yet “pure-play sustainable”,

• enable material emissions reductions, and

• are required in a net-zero economy.

This excludes transition claims that rely primarily on intent, aspiration or future alignment, rather than identifiable assets, technologies or upgrades.

For sustainability-linked solutions (Parameter 3), the bar is similarly concrete: performance-linked instruments must be tied to ambitious, credible KPIs, not generic ESG language.

What is quietly ruled out here is transition-by-proxy — financing justified mainly by sector membership, public commitments or long-dated net-zero pledges without operational teeth.

The framework seems to favour engineering over storytelling.

3. Transition finance that cannot be validated in-quarter is not counted

Another exclusion sits in the mechanics of accounting rather than the philosophy.

Deutsche Bank makes clear that, where validation against its frameworks cannot be completed before the end of a reporting quarter, volumes are only recognised once validation is finalised, even if the transaction itself has already occurred.

This matters because transition finance, by its nature, often involves:

• complex eligibility assessments,

• evolving technical criteria, and

• third-party opinions or internal committee sign-offs.

The practical result is that speed alone is insufficient. Governance completion determines recognition. Capital that moves faster than validation does not count (at least not immediately).

Again, this is about integrity. But it also means that certain fast-moving or bespoke transition deals may fall outside headline reporting windows, even if they are strategically important.

4. Nature-related finance is framed as ambition, not yet as a core accounting pillar

Alongside climate transition, Deutsche Bank introduces a nature transaction ambition: 300 transactions by the end of 2027 aligned with biodiversity, ecosystem conservation and restoration.

This is significant and is also carefully positioned.

Nature-related transactions are not yet integrated into the main sustainable and transition finance accounting architecture in the same way as climate-linked activities. They are described as an ambition, shaped with input from a Nature Advisory Panel, and connected to emerging instruments such as biodiversity credits.

What is quietly excluded is the treatment of nature finance as a fully mature, target-driving category equivalent to climate transition finance.

5. Some high-risk transition technologies remain structurally difficult to include

Finally, there is an exclusion that appears not in explicit language but in what the framework requires for eligibility.

By insisting on:

• clear activity definitions,

• measurable emissions outcomes,

• robust governance, and

• credible transition pathways,

the TFF seems to make it harder for certain high-uncertainty transition technologies to qualify (e.g. where lifecycle impacts, scalability or emissions performance remain contested).

This does not mean such technologies are ruled out forever. But it does mean that until they can be assessed against the framework’s parameters with confidence, they will struggle to qualify as counted transition finance.

The Hamburg Score view

Taken together, Deutsche Bank’s TFF is less expansive than some market narratives around “transition” might suggest, and that is exactly why it deserves attention.

What it quietly excludes is:

• entity-level transition finance from headline targets (for now),

• narrative-led transition without activity anchors,

• finance that outpaces validation,

• nature finance as a fully integrated accounting category (yet), and

• transition claims that cannot survive technical scrutiny.

The framework seems to be designed to defend the credibility of the numbers it does produce.

For clients, the message is clear: transition finance at Deutsche Bank will increasingly require specificity, evidence and patience.

For the market, the more interesting question may be whether other IFIs follow this deliberately constrained path, or whether they choose to count first and clarify later.

In Hamburg terms, this is a scorecard that refuses to applaud until the bout is properly fought.

On paper, the EDB aligns itself with international standards: it references the Equator Principles, participates in green bond markets and claims compliance with sustainable development goals. But what happens when you actually read its Environmental and Social Framework?

This new Hamburg Score Briefing does just that. Based entirely on publicly available documentation, it examines how the EDB’s safeguard system is structured — and what that means in practice.

Key Findings

• The EDB allows environmental and social appraisal up to 12 months after loan signing, unlike IFC or EBRD which require it pre-approval.

• It relies entirely on national legislation without equivalence checks, even for high-risk (Category A) projects.

• No E&S documentation (e.g. Environmental or Social Impact Assessments) are published for major projects, including large-scale hydropower and gas pipelines.

• Its Exclusion List includes “case-by-case” exceptions for coal, nuclear energy, hazardous waste incineration and asbestos.

• Projects listed in sustainability reports often lack category classification, documentation or disclosure on stakeholder consultation.

The Hamburg Score View

Viewed as a whole, the EDB’s safeguard architecture presents a system that is procedurally light, transparency-averse and reliant on national frameworks, even for high-risk investments. The Bank’s public commitments to ESG and sustainability are not matched by disclosure practices, risk sequencing or accountability tools.

What the EDB’s ESF does provide is:

• A clear statement of reliance on borrower country systems, with minimal procedural overlay;

• A permissive framework allowing post-approval appraisal and discretionary project classification;

• A flexible exclusion regime that leaves room for high-risk sectors under internal interpretation.

For practitioners, the key question is not whether the EDB’s framework looks aligned in principle, but whether its actual governance model delivers the environmental and social assurance expected from a development bank of this scale.

The Hamburg Score_Briefing 1_EDB_Dec 2025

1.23MB ∙ PDF file

Download

Download

When the EBRD updated its Environmental and Social Policy (ESP) in 2024, Environmental and Social Requirement (ESR) 9 for Financial Intermediaries emerged with a clearer architecture: sharper governance expectations, formalised disclosure and a more structured framework for Category A subprojects.

But ESR 9 is still a policy requirement document. It tells partner financial institutions (PFIs) what must happen, not how.

That “how” is what the new ESR 9 Guidance Note is designed to explain. It does not create new obligations. Instead, it fills in the operational detail that ESR 9 itself leaves open.

Below is a factual account of what the new GN clarifies relative to the ESR.

1. ESR 9: what actually changed since 2019

The starting point is ESR 9 itself, not the Guidance Note.

The 2024 ESP and ESRs took effect on 1 January 2025. ESR 9 largely tracks the 2019 Performance Requirement (PR) 9, but with several headline changes that are worth keeping in view as the frame for the GN:

1.1. Performance Requirements are now ESRs

The architecture is the same (ten requirements), but the terminology has shifted to match the new ESP.

1.2. The referral list has gone

The 2019 combination of an Exclusion List plus a Referral List has been simplified: only the E&S Exclusion List remains in Annex A of the ESP. Subprojects that trigger Category A still have to be referred to the EBRD where required, but there is no separate formal “referral list” annex.

1.3. PFIs must now disclose their ESMS and Category A subprojects

ESR 9 paragraph 16 requires:

• a summary of the ESMS to be published on the PFI’s website; and

• disclosure of Category A subprojects financed with EBRD proceeds (with links to publicly available ESIA reports, where such projects exist).

1.4. External communications and a grievance mechanism are explicit ESMS features

ESR 9 paragraph 15 requires ESMS procedures for dealing with external communications on E&S matters, including an external grievance mechanism. This was previously implied; it is now clearly spelt out.

1.5. Category A via capital market instruments is now permissible – with conditions

Under the 2024 ESP, the EBRD may invest in capital market instruments (for example, green bonds) where the use of proceeds may include Category A subprojects, subject to “stringent preconditions” and additional Paris alignment tests.

1.6. ESRs 1-8 and 10 have been updated

For PFIs, those changes only bite where they finance Category A subprojects; most FI business remains primarily about national law + ESR 9, with ESRs 1-8 and 10 coming into play primarily where Category A subprojects are financed.

However, PFIs must comply with ESR 2 (labour and working conditions) and ESR 4 (health, safety and security) as they apply to the PFI’s own workforce, premises and internal operations. The GN points back to those ESRs and their own guidance notes without duplicating them.

1.7. Paris alignment and related frameworks are now part of the landscape

ESR 9 sits alongside:

• the EBRD’s Paris Agreement alignment methodology for intermediated finance;

• the growing ecosystem of national and EU requirements (CSRD, ESRS, supervisory climate guidance), which the GN acknowledges explicitly.

1.8. Renewable energy eligibility criteria (GET)

For FI portfolios that finance renewable energy, the GN points to separate E&S eligibility criteria for solar, wind, hydropower, bioenergy and geothermal. Where relevant, these sit alongside ESR 9 and the ESP and are explicitly referenced as additional filters for subprojects.

Two additional clarifications in the GN are worth flagging as part of this baseline:

1.9. Double materiality framing

Section 2 defines “E&S risk” using the language of double materiality found in the CSRD/ESRS:

• “impact materiality” – impacts on people and the environment;

• “financial materiality” – how those impacts can translate into credit risk for the PFI.

ESR 9 itself does not require a full double materiality assessment, but the note clearly situates FI E&S risk in that vocabulary.

1.10. Scope and product coverage

The GN offers:

• an indicative table of FI products within ESR 9’s scope (project finance, term corporate loans, SME/MSME lending above very short-term tenors, asset finance, private debt, etc.); and

• a list of products treated as negligible E&S risk for ESR 9 purposes (current accounts, credit cards, residential mortgages, interbank lending, advisory, plain-vanilla remittance services, and so on).

That is the frame. The rest of the GN is essentially about how to implement those expectations in a way that is commensurate with risk and consistent across instruments.

2. ESMS architecture: familiar process, tightened edges

The core ESMS architecture for FI operations is very recognisable from the PR 9 times. The new GN retains the same five-step process for integrating E&S into the credit and investment cycle:

1. Screening

2. Categorisation

3. Risk assessment

4. Risk mitigation and control

5. Risk monitoring and review

and expects PFIs to apply those steps to all subprojects financed with EBRD proceeds, with proportionality on depth and effort.

What changes in the new note is the clarity of the signposting. Specifically:

2.1. Screening is now anchored in the E&S Exclusion List

The GN confirms that screening must:

• check every proposed subproject against the E&S Exclusion List in the ESP;

• make an initial judgement on whether Category A is likely, with final confirmation coming after more information is gathered.

• The old notion of a formal “referral list” no longer appears; referral now follows from Category A status and the financing agreement, not from a separate annex.

2.2. Risk categorisation is bifurcated by business model

For:

• SME/MSME-focused operations, the note still recommends a three-tier low/medium/high system, with definitions very close to those in the previous guidance.

• Project finance / large infrastructure / higher-risk corporate lending, the note recommends adopting A/B/C categorisation, aligned with EBRD’s own definitions and the Equator Principles. The explicit EP reference is new in the GN itself.

2.3. Assessment, mitigation and monitoring are steady in substance

The note re-states:

• regulatory compliance checks as the baseline;

• use of ESDD, escalation of effort with risk, and the appropriate use of external consultants for complex or high-risk deals;

• ESAPs as the key instrument when a transaction can be brought into compliance over time;

• monitoring as a combination of compliance checks, performance indicators (accidents, grievances, emissions data, labour unrest, etc.) and early warning for deterioration or new risks.

2.4. Record-keeping is spelt out more explicitly

PFIs are expected to:

• maintain transaction-level E&S files (including ESDD reports and supporting evidence);

• keep those records available for inspection by EBRD and other lenders;

• present aggregate E&S performance data periodically to senior management, the board and the executive committee.

In other words: the process itself has not changed, but the GN gives a more structured description of how it should work in practice, and ties it more explicitly to the ESP’s Exclusion List, Category A definitions and the EBRD’s suite of briefing notes and tools.

3. Category A: from a single paragraph to a full playbook

Where the GN really adds value is in its treatment of Category A subprojects. Under PR 9, this was largely dealt with in a few lines. Under ESR 9, there is an entire chapter.

There are three main elements.

3.1. When Category A is “in scope” at all

The note starts with an important clarification: most micro, small and medium-sized enterprise (MSME)-focused credit lines will simply not permit Category A at all. That is explicit in many loan agreements.

However, it recognises several situations where Category A may legitimately appear in a PFI’s business, for example:

• equity investments in banks with significant corporate and project finance portfolios;

• infrastructure and private debt funds with greenfield renewables, transport or heavy industry in their strategies;

• FI green bonds where use of proceeds may include such assets.

So Category A is not expected to be ubiquitous, but it is not treated as an aberration either.

3.2. What ESR 9 requires when Category A is permitted

Where Category A subprojects are allowed, ESR 9 and the GN flesh out what must happen:

• Applicable standards.
  Category A subprojects funded from EBRD proceeds must:

  • undergo an ESIA consistent with ESR 1 and national law;

  • meet the relevant provisions of ESRs 1-8 and 10;

  • be referred to EBRD where the investment agreement requires it.

• Roles and responsibilities are clearly divided.
  The note spells out who does what:

  • The sub-borrower / investee company (project developer) is responsible for ESIA, management plans, stakeholder engagement, disclosure and project-level grievance mechanisms, and for reporting progress and incidents to the PFI.

  • The PFI is responsible for correct categorisation, appropriate E&S due diligence, integrating E&S covenants and conditions into the financing, monitoring performance (potentially with external experts) and maintaining its own FI-level grievance channel.

  • The EBRD provides a timely E&S opinion on referred Category A subprojects, may advise on additional work or mitigation, but does not replace the PFI’s due diligence or “clear” projects on its behalf.

• Consultation, disclosure and timing are given practical parameters.
  The following requirements are included in the GN:

  • Formal, participatory stakeholder engagement embedded in the ESIA;

  • PFIs are encouraged to ensure that ESIA documentation is publicly disclosed for at least 30 days before the PFI’s investment decision;

  • PFIs are recommended to refer potential Category A subprojects to EBRD as early as possible, ideally around 90 days before their own final decision; and

  • last-minute referrals are discouraged, with an indication that EBRD will typically need at least 10 working days to consider such cases.

• PFI-level disclosure of Category A.
  To operationalise ESR 9 paragraph 16, the note encourages PFIs to publish, within six months of signing, basic information on Category A subprojects on their websites:

  • project name and location;

  • Category A designation;

  • month and year of signing;

  • a link to publicly available ESIA report(s) issued by the project developer.

None of this changes the underlying requirement that Category A must meet ESRs 1-8 and 10; what it does is convert that into a fairly practical “how to” for PFIs and their clients.

3.3. What an “advanced” ESMS looks like for Category A

A particularly useful addition is the section on what the EBRD considers an “advanced” ESMS when deciding whether a PFI is ready to handle Category A subprojects with EBRD funding.

The GN sets out indicative criteria under three broad headings:

• Governance and resources:

  • evidence of board-level oversight of E&S risk (for example, a sub-committee with clear terms of reference);

  • integration of E&S (including climate and nature risks) into risk appetite frameworks;

  • demonstrable E&S capacity in corporate lending / project finance origination teams;

  • at least two or three qualified E&S staff with authority to escalate or pause transactions on E&S grounds;

  • access to suitable external consulting firms for ESIA review, lenders’ E&S adviser roles and site-based monitoring.

• Policies, procedures and track record:

  • formal, tested policies and procedures for E&S risk management, including correct identification of Category A;

  • a track record of Category A projects handled to international standards (binding ESAPs, ESIAs with meaningful consultation and disclosure), ideally including examples from previous DFI or EBRD credit lines;

  • a dedicated grievance mechanism aligned with EBRD’s stakeholder engagement guidance.

• Transparency and accountability (plus general indicators):

  • web-published ESMS summary and description of the FI-level grievance mechanism;

  • ESIA disclosure practices in line with national law and EBRD’s views on good practice;

  • public sustainability reporting that covers key E&S aspects and, ideally, is subject to some level of third-party assurance;

  • no unresolved serious complaints at OECD National Contact Points;

  • no recent material E&S legal actions that raise concerns about the PFI’s approach;

  • no delisting from initiatives such as the Equator Principles for reasons other than mergers or acquisitions.

Equator Principles membership is explicitly noted as neither necessary nor sufficient. What matters is the underlying capability and governance.

For PFIs, this effectively becomes a self-diagnostic checklist for any strategy that contemplates Category A exposure in an EBRD-linked portfolio.

4. Capital market instruments and green bonds

In certain circumstances, investments in capital market instruments may involve Category A subprojects. The ESR 9 GN provides a concrete set of expectations for this.

In summary, those include:

• Scope.
  The main focus here is use-of-proceeds bonds issued by PFIs (green, social, sustainability or sustainability-linked), where:

  • EBRD’s investment is channelled via the bond; and

  • Category A subprojects are possible or probable within the eligible project universe.

• Due diligence on “possible or probable”.
  The EBRD will assess, on the basis of the bond documentation and the issuer’s sustainability framework and reporting, whether Category A exposure is likely. Indicators include:

  • focus on SME/small ticket lending vs. large project finance;

  • caps on individual loan size;

  • the mix of sectors in the eligible pool.

  Where information is limited, the note says the precautionary principle applies (the EBRD assumes higher E&S risk).

• ESMS expectations if Category A is on the table.
  If Category A is possible or probable, the same “advanced ESMS” thinking applies: the EBRD will look for:

  • mature policies and procedures,

  • relevant experience with high-risk projects,

  • the ability to identify Category A correctly and apply ESRs 1-8 and 10,

  • appropriate ESIA, consultation and disclosure standards at project level.

• Information flows and reporting.
  The GN encourages PFIs to:

  • reflect E&S risk management arrangements and the treatment of Category A in their bond frameworks and second-party opinions;

  • provide relevant summary E&S information and identification of any Category A subprojects in post-issuance allocation and impact reports, taking into account the International Capital Markets Association (ICMA) guidance on E&S risks in eligible categories;

  • use other available information channels where appropriate.

• Referral requirement and bonds.
  The note confirms that the Category A referral requirement does not apply in the same way for EBRD investments in capital market instruments, provided:

  • the EBRD has been informed in advance that Category A subprojects are possible or probable; and

  • due diligence has satisfied the Bank that the PFI’s ESMS is adequate for such exposure.

For FI issuers, these requirements give a fairly clear idea of what will make a green bond structurally investable for EBRD where large, higher-risk assets might sit under the hood.

5. Context-specific guidance and cross-cutting themes

The final chapter 6 consolidates a set of context-specific criteria that used to sit largely in separate briefing notes. This does not change the underlying ESR 9 obligations, but it does bring a few previously “background” expectations into the main guidance.

5.1. Banks and the Trade Facilitation Programme

For banks, the GN:

• points explicitly to the Principles for Responsible Banking and the Equator Principles as relevant reference frameworks; and

• clarifies that under the Trade Facilitation Programme:

  • the E&S Exclusion List applies to all relevant transactions;

  • additional due diligence may be needed for flows with elevated supply-chain risk (solar equipment, textiles, soft commodities such as palm and soy, etc.);

  • PFIs are expected to undertake their own E&S checks in addition to EBRD’s appraisal.

5.2. Microfinance institutions

For microfinance, the note sets a “minimum viable ESMS”:

• basic processes for Exclusion List checks and verification that any micro-enterprise requiring an E&S-related permit under national law has it or is in the process;

• at least one staff member responsible for ESMS implementation;

• simplified E&S reporting; and

• alignment with recognised Client Protection Principles, including an accessible customer grievance mechanism.

5.3. Leasing companies

For leasing, the core E&S issues are defined as:

• the nature of the asset itself;

• the use to which it is put; and

• the lessee’s capability to use and maintain the asset safely and responsibly.

The note then directs PFIs to the more detailed leasing-specific briefing note.

5.4. Funds

For funds (private equity, venture capital, infrastructure, private debt), the GN:

• suggests aligning with the Principles for Responsible Investment (PRI) and other recognised frameworks cited in the funds briefing note;

• notes that many such funds operate under AIFMD/UCITS and therefore fall under EU sustainable finance rules (Taxonomy, SFDR, CSRD); and

• signals EBRD’s intent to seek interoperability between its own E&S requirements and EU disclosure frameworks, particularly around the equivalence of reporting and key indicators.

Funds that in turn finance PFIs are expected to mirror ESR 9-type policies and procedures in their own operations.

5.5. Cross-cutting issues: SEAH and supply chains

Finally, the note anchors two themes that have increasingly come to the surface in FI discussions:

• Sexual exploitation, abuse and harassment (SEAH).
  All PFIs are expected to integrate SEAH considerations into E&S risk assessment and management of subprojects financed from EBRD proceeds, drawing on the dedicated SEAH briefing note.

• Supply chain E&S risks.
  PFIs are expected to consider supply chain risks, especially for:

  • commercial and utility-scale renewables (wind, solar);

  • energy storage;

  • soft commodities with high risks of modern slavery and biodiversity loss (cocoa, cotton, palm oil, soy, and similar).

Again, these themes are not new to EBRD’s broader work; what changes is that their relevance to ESR 9 is now explicitly written into the main FI Guidance Note.

6. The Hamburg Score view

Viewed as a whole, this updated note does not fundamentally alter what the EBRD expects from PFIs. The ESMS architecture, the proportionality principle and the core obligations are continuous with the 2019 framework.

What it does do is:

• lock in the new ESR 9 elements (disclosure, grievance mechanisms, potential Category A via bonds);

• provide a much more concrete playbook for Category A FI business; and

• pull a range of contextual expectations (Paris alignment, SEAH, supply chains, EU sustainable finance rules) into a single, reasonably coherent document.

For practitioners, the practical question is more “how far does our existing ESMS already match this picture – and where are the gaps in governance, disclosure and Category A readiness that will be visible when someone reads this GN carefully?”

Subscribe now

Over the past few months I’ve been posting the odd reflection on LinkedIn about how development banks write (and actually use) their environmental and social standards. I assumed these would sink without trace, but quite a few of you seemed to recognise the familiar quirks, silences and small absurdities I was pointing to. Several even asked for something more permanent than a LinkedIn thread.

So this is it.

The name comes from Viktor Shklovsky’s “Hamburg score”: the ranking of boxers done quietly in Hamburg, away from the circus of reputation and polite applause. In Hamburg, he said, fighters were judged as they genuinely were, not as they were meant to be.

It struck me as an uncomfortably accurate metaphor for the world of development finance. On paper, most institutions have immaculate frameworks. In practice… well, the story tends to be rather more textured. The interesting part is rarely the grand declaration, but the footnote, the omission or the quiet exception tucked into the back pages.

The Hamburg Score is simply a place to examine these things with a clearer eye, a bit more space and the occasional raised eyebrow. I plan to publish:

• Monthly Briefings: longer pieces that pick apart particular frameworks or trends;

• Dispatches: shorter notes on whatever development, contradiction or curiosity happens to surface;

• And now and then, a brief reflection from “behind the scorecard”.

This isn’t a grand venture but rather a way of gathering the thinking I’ve already been doing, giving it a better home than LinkedIn’s disappearing feed, and keeping track of how E&S standards shift across the development finance universe.

If any of this sparks discussion, all the better.

More soon.

Val