Talking Points: Proposed World Bank Environmental and Social Policy (ESP) and Assessment and Management of Environmental and Social Risks and Impacts - ESS1 Bank Information Center (November 2014) Vince McElhinny (vmcelhinny@bicusa.org) The proposed Environment and Social Policy (ESP) defines the World Bank's safeguard responsibilities. As part of the overarching Environment and Social Framework (ESF), the ESP will replace the other safeguard Operational Policies and sets out the mandatory requirements of the World Bank. ESS1 establishes the responsibilities of the Borrower in the activities required for the Assessment and Management of Environmental and Social Risks. ESP and ESS1 reflect many joint requirements for E&S risk and impact management. The following preliminary observations are based on the July 30, 2014 version of the ESF, ESS1, including Annexes 1 & 2, and Information Notes 1 & 2, pending the disclosure of additional ESF elements. Massive ESP loopholes spell the end of safeguards as the World Bank's approach to protecting people and the planet: Safeguards were defined as a minimum, predictable safety net of protections for vulnerable people and the environment. IEG concluded that the existing Safeguards were working, but the implementation was poor. In contrast to IEG recommendations and the 2012 approach paper mandate to "review and update" the existing policies, the ESP reverses over thirty years of accumulated knowledge about safeguards as a concept. The ESP abandons clear definition and strict adherence to ex ante minimum procedural requirements in favor of unlimited discretion by Bank Management. The ESP introduces a proliferation of 'opt out' clauses without clear eligibility thresholds and advances a vague commitment to adaptive management during implementation. The extreme inconsistency in how the proposed Bank standards will likely apply in practice will result in a safety net with so many loopholes, uncertainties and unaccountable judgments that the original purpose of safeguards will be rendered ineffective. Unequivocal and significant dilution: The ESP reverses a thirty year progressive strengthening of safeguard protections. A number of the proposed changes are positive steps to bring the World Bank in line with international standards. However, some clear dilutions of current safeguard policy requirements are indisputable, regardless of the disclaimer about an outcomes orientation. Dilutions include the broadly expanded but weakly regulated deferral of Bank safeguard responsibility through multiple opt out clauses each of which is unaccompanied by clear thresholds (ESP 9-12, 23-26, 31); unlimited flexibility to defer appraisal and adopt open-ended compliance timeframes (ESP 7, 49); absence of explicit minimum procedural requirements particularly for consultation and disclosure (ESP 42-45); an opt out clause for the Indigenous Peoples Policy (ESP 33); the lack of similar disclosure and assessment requirements for substantial risk subprojects (ESP 34-35); the elimination of any attempt to define a project's area of influence, and the potential for many other dilutions that will become clear as the forthcoming sections of the ESF are disclosed. Multiple 'Opt Out' clauses for delegating core safeguard responsibility without clear thresholds: The ESP allows the Bank to delegate the risk management responsibility to a Borrower's framework or to third parties when these alternatives would "enable the project to achieve objectives materially consistent with the ESSs" ( 23). The ESP promotes deferral of Bank responsibility when the Borrower framework is consistent with the largely aspirational ESS objectives, which is a very low bar to meet and a dilution of current safeguard policies. Gaps and a lack of clarity regarding how the ESF objectives would be met leaves excessive room for subjective interpretation. For example, if an alternative framework under consideration lacks a resettlement policy, an established definition of a critical natural 1
habitat, consistent environmental assessment requirements for substantial risk activities, or any disclosure norms for monitoring and reporting on E&S risk management performance, would it matter how the ESS or ESS objectives would be met? The ESP and ESS1 contain little detail on when or how gap analysis will be done or disclosed (ESS1 Information Note 2). The ESP lacks minimum standards for ensuring Borrower frameworks are an acceptable equivalent to Bank safeguards or standards. The deferred responsibility option will make Bank enforcement of compliance even less certain under several other framework opt out provisions for transferring safeguard responsibility to a third party. The ESP proposes the adoption of a common approach agreement for jointly financed projects, financial intermediaries, projects in advanced state of construction, and associated facilities financed by other parties - where no material deviation from the objectives (rather than the already weakened requirements) of the ESSs is the new, lower standard ( 9-12). Exclusion of strict requirements for FI subprojects (ESS9 7), and unclear grounds for whether the Borrower has demonstrated that it has no control or influence over an Associated Facility (ESP 11), are just two examples of weak threshold criteria that can easily result in the use of loopholes to escape safeguard requirements. Proposed Revisions: list activities that can not be delegated, including all high risk and substantial risk activities; condition eligibility for use of alternative frameworks based on consistency with more operational, less aspirational ESS objectives; define criteria for acceptable threshold of Borrower capacity (such as track record) state requirement that regional safeguard advisors (RSAs) are to retain clearance authority on all framework or delegated risk management projects require use of community monitoring and/or independent 3rd party monitoring or audits for all delegation of safeguard responsibility. Deferred appraisal, open-ended compliance and questionable ESCP leverage: The ESP expands the flexibility to defer appraisal and when to establish compliance with other requirements in the ESSs. Although a measure of flexibility on the time of appraisal and compliance exists with the current safeguard policies, the ESP and ESS1 provide unlimited flexibility to defer appraisal of all activities and subjects all compliance deadlines to an open-ended timetable. The Borrower must now, "meet the requirements of the ESSs in a manner and timeframe acceptable to the Bank." (ESP 7, 13, 16; ESS1 6, 9) With this option of deferred appraisal, many more projects will go to the Board with frameworks for guiding future risk management, rather than with concrete and budgeted assessments and mitigation plans. The ESP introduces an Environmental and Social Commitment Plan (ESCP) as part of the legal agreement that will outline the deferred actions, deadlines and monitoring indicators. Although a potentially positive step for clarifying commitments, the ESP removes any limitation on the type of activities for which appraisal can now be deferred (ESP 39). But as an instrument that is both flexible and legally binding, the ESP does not clarify how Bank leverage will function for non-compliance during implementation by simply depending on an environment and social commitment plan (ESCP). Among the weaknesses of the ESCP explanation in ESS 1 (Annex 2): failure to explain how an ESCP can be both flexible through adaptive management and legally binding. failure to specify how conditionality will be triggered in instances of non-compliance. failure to define the range of non-compliance remedy, recourse or sanction (i.e. relation to disbursements). lack of clear budget or complete timetable for all action items. 2
does not identify the Bank person or entity responsible for supervising the enforcement of the agreed upon commitment plan. The ESS1 simply suggests that when compliance is postponed, "the Borrower will not carry out any activities in relation to the project that may cause material or significant adverse environmental or social risks or impacts until the relevant plans or actions have been completed and (where necessary) implemented to the satisfaction or the Bank." (ESS1 36). Such a protection may prove to be wishful thinking, as has too often been the case with existing ESMPs, if the proposed deferral of biodiversity, indigenous people's consultation or cumulative impact management plans (all indicated action examples in Annex 2) call for design changes that are no longer possible or "technically or financially feasible," possibly with irreversible impacts. The lack of clarity is also confusing for borrowers, as it is unclear what the potentially required assessment capacity will be or the consequences will be for noncompliance. See Talking Points on ESS1- ESCP for more detail. exclude and list high risk and substantial risk activities from any means of deferred appraisal ( 7, 13, 16; ESS1 6, 9); clarify threshold criteria for deferral ineligibility; link ESCP to disbursements, among the other remedy, recourse and sanction options and clarify their use in the ESP; define minimum ESCP disclosure and consultation criteria, minimum duration, waiver process,; require 3rd party and community monitoring for deferred appraisal. New risk classification system could reduce routine safeguard requirements: The ESP introduces a new risk rating system that replaces the old Category A,B and C for high, moderate and low environment and social risks, with a new 4 tier high, substantial, moderate and low risk classification system (ESP 20-22). The proposed 4 tier system is a potentially positive step to improve the precision of environmental and social risk signaling. While the Bank retains the responsibility to assign the risk classification, an information note that attempts to explain how the new risk classification system will work, leaves more questions than answers. Without greater clarification, including a mapping of a sample of Bank's current portfolio under the proposed classification system, there is little confidence that ESP requirements for "substantial" risk projects or subprojects will be consistently followed. Several concerns with the new risk rating system stand out: 1. Cost savings is not the goal. The motivation for the change seems sound - to assure a better allocation of safeguard resources to ensure the highest risk activities receive the resources needed to properly manage impacts on people or the environment. However, it would be erroneous to conflate this objective with reducing safeguard costs, when the Bank has accumulated a significant deficit after years of underinvestment in safeguard implementation and significant Borrower capacity strengthening needs remain. 2. Unclear definition and requirements for substantial risk projects. The introduction of a "substantial" risk category (ESP 20) is a potential improvement. By better distinguishing, capturing and treating the "high B" projects as "high risk", the Bank would be addressing a major cause of safeguard implementation problems associated with misclassification of risk. Criteria for "high" risk projects are strengthened. However, the precautionary nature of safeguards should weigh against making such professional judgments overly complex (Info note 1, 9). The Bank has declined to adopt a simpler "significant risk" threshold or include EBRD's indicative 3
activity risk list, two measures which could reduce the complexity. Of greater concern is that neither the ESP nor ESS1 adequately clarifies the dividing line between high and substantial risk projects, and consequently leaves unclear how risk management requirements would differ. The Bank began to streamline safeguards for medium and low risk projects several years ago, but what is the point of a new classification system that fails to define clear requirements for substantial risk? 3. Substantial risk subprojects not adequately covered by ESS. ESS1 and other ESS requirements apply to only high risk subprojects (ESP 34-35). Over half of the Bank's portfolio has some type of subprojects, a share that will only increase. High risk and substantial risk subprojects should be subject to the same risk management requirements. 4. Borrower commitment. The classification of environmental and social risk now places a larger emphasis on the "capacity and commitment of the Borrower to manage such risks and impacts in a manner consistent with the ESSs" (Info note 1, 6). Reference is also made to the Borrower's "track record of past project implementation." Clarification of how both concepts will be assessed is required. 5. Adjustment of the risk classification. Another significant change in the new system is the possibility that risk classification can be adjusted by the Bank (Info note 1, 5). This flexibility is already a part of current safeguard practice during preparation. The gain would seem to be in having greater ability to adjust a risk classification during implementation. A periodic review of the validity of an initial risk classification decision can be useful means for assessing and reporting on safeguard performance. However, the ESP does not explain how ex ante risk classifications would change during implementation. Again, references to assessing the Borrower's track record on risk management, or changes to the context, among other factors, do not seem like sufficiently clear or consequential enough changes to justify a lowering of environmental and social risk during the construction and early operation phase of a project (which is presumably the ESCP duration). As with the ESS1 Annex 2 example of an ESCP, the Bank must clarify the circumstances in which a high or substantial risk project might be revised to moderate or low risk, as well as the resource allocation implications. Similarly, the Bank should indicate how a review would result in increased risk classification, which would trigger more frequent or intense monitoring among other changes. 6. How does the SORT risk rating inform the E/S classification adjustment? In September, 2014, the Bank approved a standardized risk rating tool (SORT), but has not explained how environmental or social risk management requirements will be compatible or subsumed within SORT project risk ratings (ESP Information Note 1, 1), or how either will determine appropriate levels of resources to properly manage risk. simplify and expand the definition of high risk based on precautionary approach adopt indicative high risk and substantial risk activity list (EBRD); clarify how Borrower commitment or "track record" for risk management would be assessed. specify ESS1 requirements for substantial risk projects; high risk and substantial risk subprojects should be subject to the same risk management requirements. clarify with examples how adjustment in risk classification would be triggered, decided and reported in the ESCP. 4
require long-term supervision of the impacts of any project initially classified as high or substantial risk, independent of change in risk classification; define an activity exclusion list Flexible opt out clauses and excessive discretion: Flexibility is appropriate in any safeguard policy and certain conditional clauses can be found both in existing safeguards as well as the policies of other IFIs. In fact, considerable flexibility can already be found in the existing safeguards. However, the ESP removes minimum bottom-line Bank requirements related to the application of the mitigation hierarchy. In the absence of certain, predictable, bottom-line requirements, the increased flexibility transforms the entire ESF into a free floating set of aspirations with extremely limited accountability. Combined, these measures constitute dilution with respect to the current safeguards. The exaggerated dependence on 'opt out' clauses provides an limitless space for interpreting or justifying risk management decisions behind a veil of unaccountability. The phrase, " in a manner and timeframe acceptable to the Bank" establishes unbounded discretion to postpone, assess, respond to and report on ESS compliance of any high or low risk activity by an unidentified Bank official. Restore predictable, certain, bottom-line minimum requirements, particularly on disclosure and consultation. See prior suggestions Accountability is impaired. The Bank is sending mixed signals on safeguard accountability. The Bank's Executive Directors encourage trust in the Bank s professional judgment, but also hedge their own trust by reserving authority over final decisions when the various 'opt out' clauses are triggered. In the end, they say, it will be the Board that approves any 'opt out' decision. However, the proliferation of conditional clauses without clear minimum requirements provides excessive discretion without equivalent accountability. Interpretation of strict compliance by the Board or the Inspection Panel will be much harder for investment lending using frameworks, borrower systems, deferred appraisal and many other new flexible clauses in the ESF. The excessive discretion left to Bank Management and the Borrowers will ensure that only in instances of gross oversight will the Inspection Panel be able to determine if policies were violated and what action was necessary. If in just the overarching framework, such a labyrinth of escape clauses are provided, the forthcoming mandatory annexes and directives are likely to expand accountability loopholes. New preventative grievance redress measures are introduced to the ESP ( 50-51), referred to as a grievance redress service, which also could delay justifiable complaints going straight to the Panel, and which could pose additional obstacles to accountability. These combined policy dilutions in the ESF and increased pressure on the Panel itself will impair safeguard accountability. ESP downgrades safeguard coverage to less than 50% of the Bank's business: Contrary to IEG recommendations, ESP does not harmonize risk management across the World Bank Group portfolio. Policy and programmatic lending, which constitute a 40% (and growing) share of Bank business (defined by lending volume), are excluded from the ESP ( 1). Many loans will also fall outside the scope of the ESP through loopholes that will allow project risks to be governed by other, possibly lesser standards. Indiscriminate use of frameworks, including for public-private partnerships (PPPs), options for delegating safeguard responsibility or fast tracking lower risk projects will reduce to a small fraction the share of Bank lending to which safeguards explicitly apply. The ESP therefore fails to raise the environmental and social risk management standards to a common higher level across the Bank 5
portfolio. Instead, World Bank environmental and social risk management will rest on a fragmented and inconsistent policy framework that incentivizes safeguard avoidance by the Bank largest borrowers. Expanded scope of social assessment, but for fewer projects and without detail commitments: ESP rightly expands consultation and assessment to include children and people with disabilities, a nonbinding mention of human rights and the principle of non-discrimination, inclusion of a first time labor standard, community health and safety, a broadened stakeholder engagement plan - all significant, positive steps that are unfortunately overshadowed by the broader policy dilution. The vague requirement of differentiated measures (ESS1 27; ESS10 16) for consulting and assessing risks of vulnerable peoples leaves excessive discretion for Bank projects to lump all vulnerable groups into the same basket, and avoid using appropriate tools to investigate and address the risks and impacts pertinent to the unique vulnerability or interests of each group. On the benefits sharing side, there is also a troubling lack of clarity. The Bank argues that the ESF is not required to do more than identify if disparities rooted in exclusion may prevent disadvantaged or vulnerable groups to share in a fair distribution of benefits from the project. While the elimination of such disparity and exclusion may not be a reasonable requirement for the ESF, the Bank is left to decide what is a reasonable effort to reduce identified disparities or patterns of exclusion (i.e. in terms of time, sustainability of outcomes and methods for assessing such risks). Within an ESS1 directive on vulnerable groups, clarify mandatory requirements Require impacts on each vulnerable group be assessed and monitored on the basis of disaggregated baseline, risk and outcome data. Clarify Bank role in defining outcome indicators, monitoring methods and reporting requirements. Associated facility loopholes & weak requirements for indirect and cumulative risk assessment: The existence of associated facilities with significant risks usually triggers a cumulative impact assessment. The concept of associated facilities is introduced, but defined more narrowly than the IFC (ESP 11-12). For the Borrower ESIA to consider an associated facility, the related investment must be necessary for the Bank funded project to be viable, and in turn would not have been constructed or expanded if the Bank project did not exist ( 12 fn 13). Bank responsibility for associated facilities is proportionate to Borrower control or influence over it (ESS1 30). The Borrower is required to demonstrate to the satisfaction of the Bank, the extent to which it cannot exercise control or influence over the associated facilities by providing details of the relevant considerations, which may include legal, regulatory, and institutional factors (fn 26). No detail is provided for how the sequence of decisions regarding associated facilities will be made (i.e. before or at concept review). So, the first hurdle to any meaningful requirement on cumulative impact is the broad discretion provided for determining whether a related investment constitutes an associated facility: (1) it must meet the two way financial dependence test; 2) Borrower must fail to show it has no control or influence; 3) unclear criteria for Bank satisfaction are met regarding either 1 or 2). While all projects must consider indirect and cumulative impacts (ESS1 21 fn 18), and the ESS1 notes the likely focus of regional or sectoral impact assessments on cumulative impacts (ESS1, Annex 1, 5g, h) such an indirect or cumulative impact assessment is not explicitly required in any specific instance (say for transboundary or global impacts or hydroelectric dams in a river basin facing competing water 6
demands). The ESP no longer attempts to even define a project area of influence, although several references are made to this undefined concept (ESS1, Annex 1, 17C). The likelihood of cumulative or indirect impact assessments being ignored or done ineffectively is heightened by the multiple loopholes for defining associated facilities, the nonexistent trigger and the lack of any procedural requirements (i.e. defining spatial and temporal boundaries of analysis based on the valued environmental or social component, VEC, rather than the project). The final loophole for avoiding the management of cumulative or indirect impacts should they be identified is the "financial and technical feasibility" criteria in the mitigation hierarchy clause that provides additional, broad discretion to rule out, defer or minimize the necessary actions to address these types of harms (ESS1, 25, fn 21). The multiple hurdles for triggering and acting appropriately to avoid, minimize or mitigate indirect or cumulative impacts amounts to an additional opt out clause. 1. Associated facilities definition should be irrespective of the source of financing and not necessarily essential for a Bank financed project to be viable. 2. Require cumulative assessments for any project with associated facilities, transborder or global impacts 3. Clarify minimum required procedures for cumulative impact assessment and management of impacts. ESP eliminates the 'red flags': Despite an expansion of stakeholder engagement, there are no longer any disclosure or participation minimums (such as those defined in BP 4.01 now). In addition, the ESP and ESSs do not provide explicit minimum requirements for the draft disclosure and content requirements, such as for the ISDS and ESCP under conditions of deferred appraisal or framework lending, or criteria for gap filling and use of borrower frameworks or common approach for jointly financed projects, safeguard performance indicators, an exclusion list, or a full implementation plan. Leaving unclear these requirements will risk taking away the few, predictable public indicators (red flags) during project preparation that safeguards are being followed. See BIC talking points on ESS10 for suggested revisions. ESP is weakened by unclear reporting requirements, underfunded safeguard capacity and no public implementation plan. The core trade-off in the ESP is between relaxed front end (ex ante) safeguard requirements in favor of enhanced supervision ("implementation support"). Despite the widely acknowledged underfunding and understaffing of past safeguard supervision and the resulting knowledge gaps, the ESP introduces few new concrete project monitoring, feedback and adjustment measures besides an annual monitoring report and the poorly defined ESCP (ESP 49-55). The Bank's role in defining what safeguard outcome indicators to monitor, where third party or community monitoring is required, and how to mitigate recurrent candor gaps in project monitoring are all left unclear. Neither does the Stakeholder Engagement Plan outlined in ESS10 explain how affected communities can participate in supervision or how any consultation feedback mechanisms would work. The ESP supervision mandate defers to OP 10.0 ( 46), which involves prior dilutions of Bank supervision policy requirements in terms of ensuring safeguard outcome indicators are included in project results frameworks, supervision frequency, skill mix or properly sequenced capacity building actions, among others. The internal resource needs for proper implementation are unknown despite an internal safeguard system in evident disarray and large gaps in supervision reporting. Beyond a list of generic implementation issues in the non-binding overview statement (pg 14), the ESF contains no commitment to consult on a detailed safeguard corrective implementation plan. For such a premium to 7
be placed on better safeguard implementation, remarkably little has changed to instill confidence in the Bank's capacity to deliver on adaptive management feedback loops that this approach would require. Disclose draft World Bank safeguard implementation plan Disclose baseline safeguard budget (FY14) and significant (100%) increases in safeguard resources for (FY15) Reinforce Regional Safeguard Advisor oversight and clearance responsibility for all high and substantial risk projects, and any projects using frameworks. Enhance incentives for safeguard specialists Increase dedicated funding for Safeguard capacity building Require and clarify Bank role in defining E & S performance indicators, reporting and evaluation requirements, and means of verification; Require 3rd party or community monitoring and E & S outcome impact assessment for certain projects. Missed opportunities for meeting highest international standards: The ESP/ESF introduces some significant expansions of safeguard application (non-discrimination, community health and safety, risks to a wider range of vulnerable peoples, FPIC, supply chains, and stakeholder engagement, among others). The ESP arrests momentum toward upward harmonization of social and environmental standards and abdicates the World Bank role as standard setter for safeguard policy by ignoring much international best practice. Some examples include the broader scope of safeguard coverage and associated requirements for policy lending (ADB, AfDB); greater uniformity of coverage between public and private sector risk management (IDB, AsDB); decision making tools, such as indicative lists for defining highest risks (EBRD), the lack of obligation to comply with the International Labour Organization's core labour standards (IFC, EBRD, ADB, AfDB), more stringent definition of associated facilities (IFC); greater resource and sequencing commitment to strengthening and use of borrower systems (AsDB); a more explicit requirement for upstream risk assessment (Arhaus convention, EU Commission); World Bank corporate reporting on environmental and social risk management falls short of the annual IDB Sustainability Report, which includes an overview of portfolio risk distribution, safeguard capacity and updates on high risk project management. While the World Bank claims that the proposed ESF increases harmonization with other MDBs, in fact the ESF is selective in the areas where the highest global standard is adopted, and ignores the many additional higher standards that are missed. Disclose full benchmarking comparison of the ESF against MDB safeguard policies, including the World Bank's existing safeguard policies. Explain where higher international standards were adopted or not, with justification. 8