63896 Safeguards and Sustainability Policies in a Changing World An Independent Evaluation of World Bank Group Experience IEG Study Series THE WORLD BANK GROUP WORKING FOR A WORLD FREE OF POVERTY The World Bank Group consists of five institutions—the International Bank for Reconstruction and Development (IBRD), the International Finance Corporation (IFC), the International Development Association (IDA), the Multilateral Investment Guarantee Agency (MIGA), and the International Centre for the Settlement of Investment Disputes (ICSID). Its mission is to fight poverty for lasting results and to help people help themselves and their environment by providing resources, sharing knowledge, building capacity, and forging partnerships in the public and private sectors. THE INDEPENDENT EVALUATION GROUP IMPROVING DEVELOPMENT RESULTS THROUGH EXCELLENCE IN EVALUATION The Independent Evaluation Group (IEG) is an independent, three-part unit within the World Bank Group. IEG-World Bank is charged with evaluating the activities of the IBRD and IDA (The World Bank), IEG-IFC focuses on assessment of IFC’s work toward private sector development, and IEG-MIGA evaluates the contributions of MIGA guarantee projects and services. IEG reports directly to the Bank’s Board of Directors through the Director-General, Evaluation. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank Group’s work, and to provide accountability in the achievement of its objectives. It also improves Bank Group work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. Safeguards and Sustainability Policies in a Changing World An Independent Evaluation of World Bank Group Experience 2010 The World Bank http://www.worldbank.org/ieg Washington, D.C. © 2010 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 13 12 11 10 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank Group. The findings, interpre- tations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the govern- ments they represent. The World Bank does not guarantee the accuracy of the data in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages a dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyrights.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H , Street, NW Washington, DC 20433, USA: fax:202-522-2422; email:pubrights@worldbank.org. Cover image: Women in India displaced by a development project. Photo courtesy of Reidar Kvam. ISBN-13: 978-1-60244-159-0 ISBN-10: 1-60244-159-6 Library of Congress Cataloging-in-Publication data have been applied for. World Bank InfoShop Independent Evaluation Group E-mail: pic@worldbank.org Communications, Strategy, and Learning Telephone: 202-458-5454 E-mail: ieg@worldbank.org Facsimile: 202-522-1500 Telephone: 202-458-4497 Facsimile: 202-522-3125 Printed on Recycled Paper Contents vii Abbreviations ix Acknowledgments xi Foreword xiii Executive Summary xxv Management Response xliii Chairperson’s Summary: Committee on Development Effectiveness (CODE) xlvii Statement of the External Advisory Panel 1 1 Evaluation Context 3 Introduction 4 Evaluation Design 7 Environmental and Social Policies at the World Bank Group 9 Roles and Responsibilities 10 Portfolio Trends 13 Organization of the Report 15 2 Process Implementation by the World Bank Group 17 Introduction 17 Evaluation of Process Implementation: World Bank 21 Evaluation of Process Implementation: IFC 25 Evaluation of Process Implementation: MIGA 27 Quality of WBG Supervision 35 Accountability Mechanisms 38 Summary 41 3 Environmental and Social Performance 43 Analytical Framework 44 Factoring Environmental and Social Performance in the Assessment of Project Development Outcomes 44 Self-Assessment of Safeguard and Sustainability Results 46 Stakeholder Perceptions of Safeguards Performance 47 Client implementation 48 Environmental and Social Performance in Bank Projects 54 Sustainability Performance in IFC Projects 58 Sustainability Performance in MIGA Projects 61 Summary of Main Findings 63 4 Risks, Benefits, and Costs of the WBG’s Safeguards and Sustainability Policies 65 Introduction 66 What Determines Safeguard Categorization? iii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 70 Estimating Benefits from Risks 71 Benefits and Costs of the WBG’s Safeguards and Sustainability Policies 73 Assessing Costs and Benefits 78 Stylized Model of Benefits and Costs 80 Summary of Main Findings 81 5 Safeguards and Sustainability Frameworks: Policy Issues 83 Relevance of the Safeguards and Performance Standards to the Current Portfolio 85 Lessons from the Use of Country Systems 87 MDB Safeguard and Sustainability Policies 93 Fitting Social Safeguard Policies to Context: Lessons from MDBs 96 WBG Paradigms for Environmental and Social Management 98 Implications for the World Bank Group 99 6 Conclusions and Recommendations 101 Conclusions 103 Recommendations 107 Appendixes 109 A: Safeguard and Sustainability Policies 113 B: Portfolio Evaluation Methodology 121 C: Survey, Interview, and Focus Group Methodologies 123 D: Feedback from Nongovernmental Organizations 125 E: Main Messages from Focus Group Discussions with World Bank Group Staff 127 Endnotes 135 References Boxes 20 2.1 Magnitude of Involuntary Resettlement in Bank Operations 27 2.2 Spotlight on the Post-2007 Disclosure of Environmental and Social Review Summaries (ESRS) 29 2.3 Adaptive Management on Safeguards in Project Restructuring 30 2.4 Local Implementation of an ESMF in a Community-Based Project in Africa 31 2.5 Staff Views on Resources for Safeguards 32 2.6 Incentives Are a Constraint on Safeguard Effectiveness in Bank Projects 33 2.7 IFC Staff Survey Feedback on Resources for Environmental and Social Work 34 2.8 IFC’s New Social Performance Standards 37 2.9 Manager Comments on the Inspection Panel 50 3.1 Structured Learning Process in the Chinese Transport Sector 51 3.2 Protecting Vulnerable Populations—Do Safeguards Ensure Adequate Coverage of Social Risks? 53 3.3 Concerns of Country Directors about Safeguards 67 4.1 Risks and Benefits Model 72 4.2 Feedback from Bank Clients and Managers 78 4.3 Quantitative Estimates of Costs and Benefits from Stylized IFC Projects iv CONTENTS 79 4.4 Quantitative Estimates of Costs and Benefits from Stylized Bank Projects 93 5.1 Country Director Views on Social Safeguard Policies 94 5.2 Sector Manager Comments on Bank Safeguards 95 5.3 Civil Society Comments on Bank Safeguards Tables 7 1.1 Comparison of WBG Safeguards and Performance Standards 18 2.1 Safeguards Preparation and Appraisal in Bank Projects (percent satisfactory) 19 2.2 Safeguards Category by Approval Year (percentage of projects) 22 2.3 Correlation of Environmental and Social Appraisal and Supervision Ratings with ESE Ratings in IFC Projects 23 2.4 IFC’s Quality at Preparation and Appraisal 23 2.5 Average Project Size at IFC by Category 26 2.6 IEG Assessment of Process Implementation in MIGA Guarantees 36 2.7 Inspection Panel Complaints Filed and Policy Violations, 1995–2009 56 3.1 Performance of FI Projects in IFC’s Portfolio 67 4.1 Indicators for Estimating Social and Environmental Risks 69 4.2 Distribution of Projects by Significance of Risks in Completed Bank Projects 69 4.3 Likelihood That a Project Is Category A (min-max) 75 4.4 Average and Median Costs for Safeguards (US$) 76 4.5 Distribution of Projects on Benefit-Cost Quadrant 88 5.1 Comparison of WBG Safeguards and Performance Standards with Other MDBs 97 5.2 Paradigms for Achieving Social and Environmental Results Figures 6 1.1 IEG Safeguards Evaluation Building Blocks 11 1.2 Bank Lending by Safeguard Category, Number, and Commitment (FY 1999–2010) 12 1.3 Trends in IFC’s Portfolio 13 1.4 Changes in MIGA’s Portfolio Composition (share of MIGA guarantee volume issued per sector) 21 2.1 IFC Net Commitments by Category and CESI Resources 28 2.2 Supervision of Safeguards in Bank-Financed Projects, by Region 29 2.3 Supervision of Safeguards in Bank-Financed Projects, by Safeguard Category 30 2.4 Projects with Policy Frameworks versus Projects with Mitigation Plans 33 2.5 IFC’s Environmental and Social Work Quality at Supervision for Pre-Performance Standard Projects 45 3.1 Adequacy of ICR Reporting on Safeguards in Completed Bank Projects, by Safeguard Category 47 3.2 Quality of Client Implementation in Bank Projects, FY 1999–2008 Approvals v S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 48 3.3 Safeguards Performance in Bank Projects 49 3.4 Bank Safeguards Performance by EA Category 49 3.5 Process Implementation and Performance for IDA/IBRD Projects 51 3.6 IEG Ratings for Implementation Performance in Bank Projects 52 3.7 Manager Feedback on Bank Safeguards Effectiveness 52 3.8 Task Team Leader Assessment of Mitigation Outcomes of Bank Safeguards, by Region 55 3.9 Implementation Performance of Safeguard Policies and Performance Standards (real sector) 56 3.10 Satisfactory Environmental and Social Performance in IFC’s Real Sector, Pre-Performance Standard Projects Appraised, FY 1999–2004 70 4.1 Predictive Probabilities for Category-A Projects 71 4.2 Cascade of Benefits from Safeguards and Sustainability Policies 73 4.3 Impact of Safeguard Policies on Lending, by Region 76 4.4 Bank Benefits and Costs for Environmental and Social Safeguards 77 4.5 Bank Client Benefits and Costs for Environmental and Social Safeguards 77 4.6 IFC Safeguard Benefits and Costs 84 5.1 Safeguards and Performance Standards in WBG Portfolio vi Abbreviations ADB Asian Development Bank AFR Africa Region AMR Annual Monitoring Report BP Bank Procedures CAO Compliance Advisor and Ombudsman CES Environment and Social Department (IFC) CESI CES Investment Support Group CODE Committee on Development Effectiveness DPL Development policy loans EA Environmental assessment EAP East Asia and Pacific Region EBRD European Bank for Reconstruction and Development ECA Europe and Central Asia Region EHS Environmental health and safety EIA Environmental Impact Assessment EMP Environmental Management Plan EPI Enhancing positive impacts ESE Environmental and Social Effects ESIA Environmental and Social Impact Assessment ESMF Environmental and Social Management Framework ESRD Environmental and Social Review Document ESRP Environmental and Social Review Procedure ESRS Environmental and Social Review Summary ESSD Environmentally and Socially Sustainable Development FI Financial intermediary FY Fiscal year(s) IBRD International Bank for Reconstruction and Development ICR Implementation Completion Report IDA International Development Association IDB Inter-American Development Bank IEG Independent Evaluation Group IFC International Finance Corporation IFI International financial institution IPN Inspection Panel IPP Indigenous Peoples Plan ISR Implementation Status and Results LCR Latin America and the Caribbean Region M&E Monitoring and evaluation MDB Multilateral development bank MIGA Multilateral Investment Guarantee Agency MNA Middle East and North Africa Region MNI Mitigating negative impacts vii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D NGO Nongovernmental organization OD Operational Directive OP Operational Policy OPCS Operations Policy and Country Services PPSSES Policy and Performance Standards on Social and Environmental Sustainability QACU Quality Assurance and Compliance Unit RAP Resettlement Action Plan SAR South Asia Region SCC Strengthening client capacity SDR Safeguards Diagnostic Reviews SEMS Social and environmental management system UCS Use of country systems WBG World Bank Group XPSR Expanded Project Supervision Report viii Acknowledgments The report was prepared by a team coordinated (Chad-Cameroon), Keith Oblitas (Lesotho), by Anis Dani (IEG-World Bank). The IFC part Anttie Talvitie and Andres Liebenthal (China), was led by Jouni Eerikainen (IEG-IFC) and the and Richard Carlos Worden (Brazil, Colombia, MIGA part by Stephan Wegner (IEG-MIGA). The the Philippines, and Nepal). team was supervised by Cheryl Gray (Director, IEG-World Bank), Marvin Taylor-Dormond Kris Hallberg and John Redwood prepared a (Director, IEG-IFC), Christine Wallich (Director, background paper on the use of country systems IEG-MIGA), Monika Huppi (Manager, IEGSE), and visited pilot projects in Romania, Tunisia, and Stoyan Tenev (Manager, IEG-IFC). and Egypt. Lawrence Salmen prepared a paper on social safeguards. Bernard Baratz and Julia Substantial inputs were provided by Peter Maldonaldo undertook background studies on Freeman (former coordinator, retired), Ramach- accountability mechanisms. andra Jammi, Andres Liebenthal, and Richard Carlos Worden. Analysis of the benefits and costs The report has benefited from the council of safeguards was undertaken by Ade Freeman, of the Advisory Panel, comprised of Mohan Anil Markandya, and Martyn Riddle, with other Munasinghe, Professor of Sustainable Develop- core team members, while the risk analysis was ment, University of Manchester, United Kingdom, carried out by Julia Maldonaldo and Andaleeb and Chairman of the Munasinghe Institute for Alam. Portfolio analysis was carried out by Development; Francis Vanclay, Professor in the Lopamudra Chakraborti, Payton Deeks, Unur Department of Cultural Geography, University Demberel, Gonzalo Griebenow, Kavita Mathur, of Groningen, The Netherlands; and Luiz Gabriel Bekir Onursal, Lisa Overbey, Chau Pham, Keith Todt de Azevedo, Sustainability Director in the Pitman, and Cameron Wilson. The portfolio Energy Vice-Presidency, Odebrecht SA, Brazil. database was set up by Maria J. Mar and the Peer reviewing took place at various stages in the e-survey undertaken by Matthew Petrie. Ethel project. We are indebted to J. Gabriel Campbell, Tarazona managed the IEG-MIGA component Kenneth Chomitz, Kirk Hamilton, Arthur Dennis during the initial phase of the evaluation. Long, Muthukumara Mani, Håkon Fonseca Nordang, and Jeremy Warford for their invaluable Field case study missions involved the late comments. Dan Aronson (Indonesia), Jouni Eerikainen (Brazil, Peru, South Africa, Tanzania, Ethiopia, We are grateful to Helen Chin for editing the the Russian Federation, and Nicaragua), Kris report for publication, to William Hurlbut for his Hallberg (Bulgaria), Lauren Kelly (Burkina Faso), editorial help on an earlier version of the report, Robert Lacey and Joy Hecht (India), Hernan Levy and to Romayne Pereira for unstinting adminis- (Brazil), Fernando Manibog and Gary Costello trative support. Vinod Thomas, Director-General, Evaluation Stoyan Tenev, Manager, IEG–IFC Cheryl W. Gray, Director, IEG–World Bank Anis A. Dani, Overall Task Coordinator, Marvin Taylor-Dormond, Director, IEG–IFC IEG–World Bank Christine Wallich, Director, IEG–MIGA Jouni Eerikainen, Task Manager, IEG–IFC Monika Huppi, Manager, IEGSE, Stephan R. Wegner, Task Manager, IEG–MIGA IEG–World Bank ix A social specialist discusses resettlement issues with villagers affected by a project in Ethiopia. Photo courtesy of Jouni Martti Eerikainen. x Foreword The global financial crisis and the BP oil spill improvement. Better coordination of supervi- illustrate dramatically the vital role regula- sion resources by the social and environmental tory regimes must play in enabling sustainable units in the regions can enable more strategic, development. The World Bank Group’s (WBG’s) risk-based supervision. While assigning respon- safeguards and sustainability policies were sibility for environmental and social monitoring enacted to prevent or mitigate adverse impacts of to clients can improve ownership, greater disclo- its projects on people and the environment. The sure of monitoring findings accompanied by effectiveness of the regulatory regime depends third-party verification are vital for accountability. not only on its targets but also the checks and balances provided by monitoring and evaluation, Safeguards and sustainability policies were disclosure of findings, and objective verification designed to address environmental and social of results. This evaluation seeks to enhance the impacts of projects at the micro level. An effectiveness of the WBG’s policies in this context unintended consequence has been a growing for achieving social and environmental results. separation between the work on safeguards and that on environmental and social sustainability. The evaluation finds that the safeguards and The importance of macro-level sustainability is sustainability policies have helped to avoid or highlighted by the growing significance of global mitigate large-scale social and environmental risks public goods, especially biodiversity and climate in the projects financed by the WBG during the change, and the growing portfolio of sectorwide, past decade. The quality of design and appraisal programmatic lending, such as Development has improved significantly since the mid-1990s. Policy Loans and Sectorwide Approaches, whose However, the implementation of the safeguard environmental and social effects also need to be policies has meant enforcing compliance with addressed. These micro-macro linkages between mandatory policies and procedures, without project safeguards and macro sustainability will engendering strong client ownership. And the need to be better handled. quality of environmental and social supervision has been deficient, particularly, but not only, in The WBG is a crucial player in promoting better World Bank-financed projects. Projects with high environmental and social outcomes worldwide. corporate risks have received adequate attention, This evaluation’s assessment is that there is a but there is insufficient differentiation of the need to improve thematic coverage of the Bank’s other projects by environmental and social risks. safeguard policies and to enhance disclosure and Results can be improved by targeting supervision independent verification of IFC monitoring and resources toward the relatively riskier projects. supervision reports to ensure accountability. The evaluation makes recommendations to maintain Adopting a systems approach—linking policy the objectives of safeguards and sustainability regulations to project design, supervision, policies; strengthen compliance, implementa- monitoring, evaluation, and disclosure—is tion, and accountability; and help ensure better essential for the effectiveness of safeguards and environmental and social results. sustainability policies. For stronger environmen- tal and social results, consistency in coverage of social and environmental impacts across the WBG is essential. Implementation and monitoring of Vinod Thomas environmental and social outcomes needs urgent Director-General, Evaluation xi Resettlement planning in Jamaica. Photo courtesy of Reidar Kvam. xii Executive Summary T he World Bank Group’s safeguards and sustainability policies were put in place to prevent or mitigate adverse impacts of its projects on people and the environment. These goals remain critical given current environmental and social trends. Recent global experience in the financial and environmental arenas demonstrates clearly the need to put in place and enforce regulatory frameworks that balance costs and benefits, both private and social. This evaluation looks, for the first time, at the full set of safeguards and sustainability policies used in the World Bank Group (WBG)—including the World Bank, the International Finance Corporation (IFC), and the Multi- lateral Investment Guarantee Agency (MIGA). The findings are intended to inform ongoing reviews of policies and strategies across the WBG, with an eye toward greater effectiveness in achieving environmental and social outcomes. Overall, the evaluation finds that the safeguards been lacking, and impacts on environmental and and sustainability policies have helped to avoid social outcomes are not yet known. or mitigate large-scale social and environmen- tal risks in the projects financed by the WBG The evaluation makes recommendations in during the review period, fiscal years 1999–2008. five areas: (1) policy frameworks to harmonize Categorization of risks has not been consistent thematic coverage across the WBG and enhance across the WBG, however, and supervision or their relevance to client needs; (2) client capacity, monitoring of results has not been thorough. responsibility, and ownership; (3) guidelines, Implementation, particularly in World Bank instruments, and incentives to strengthen projects, has meant enforcing compliance with supervision; (4) monitoring, evaluation, comple- mandatory policies and procedures, which has tion reporting, verification, and disclosure; and not engendered strong client ownership. The (5) systems and instruments for accountability Bank’s compliance-based approach is becoming and grievance redress. less effective as its portfolio moves beyond traditional investment projects (which now Although the policies that are the foundation constitute less than half of new lending across of today’s safeguards were promulgated in the the WBG). Greater emphasis on developing 1980s, they gained more prominence after the client ownership and systems is needed going Morse Commission’s 1992 report on the Sardar forward. Ownership among private sector clients Sarovar Dam highlighted significant failures in and business partners has improved with the enforcing social and environmental policies. To introduction of a new Performance Standards ensure that such errors were not repeated, the approach at IFC and MIGA, but verification, World Bank established the Inspection Panel, disclosure, and community ownership have a permanent body reporting to the Board of xiii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Directors to investigate complaints, and created reviews of the portfolio, background studies, a separate Quality Assurance and Compliance field studies, staff and nongovernmental organi- Unit in 1999 to provide additional oversight of zation surveys and focus group discussions, and safeguards quality in Bank projects. For IFC interviews with WBG clients and managers. and MIGA an accountability mechanism—the Compliance Advisor and Ombudsman (CAO)— Development policy lending is excluded from was created in 1998. The CAO, which reports the analysis because it is not subject to the same to the WBG president, acts as a mechanism for environmental and social requirements. Consid- grievance redress and, if cases are not resolved, ering the sizable share of development policy then investigates projects to verify if they are in lending, the nature and quality of its environ- compliance with relevant policies. mental and social impacts need to be looked at separately from the current evaluation. World Bank safeguards consist, today, of 10 separate policies—6 environmental, 2 social, This evaluation focuses on the relevance of the and 2 legal. IFC replaced the safeguards in 2006 safeguards and Performance Standards, the with a single policy on social and environmental effectiveness of the implementation process sustainability and eight Performance Standards during preparation and supervision, and the divided equally among social and environmental environmental and social performance at the standards; MIGA followed suit in 2007. The new project level. It applies a model to rank safeguards policy has clarified roles and responsibilities for risks and estimate benefits. The model also the private sector clients of IFC and MIGA and provides a means to assess the robustness of is accompanied by a range of advisory services the WBG’s project categorization system and intended to strengthen their clients’ institu- the efficiency of resource use by comparing the tions and capacities. Environmental and social benefits and costs of applying the safeguards skills among WBG staff have been enhanced in and Performance Standards. This analysis is varying degrees and, as evidenced by findings constrained by data limitations but provides from the Quality Assurance Group, the quality some useful insights in addition to serving as a of safeguards design and appraisal has improved guide for further work. significantly since the mid-1990s. However, substantial challenges in supervision, monitor- Context ing, and follow-up remain. The context in which the WBG operates has seen several significant shifts over the past The Scope of the Evaluation two decades. First, the WBG’s borrowers have This is the first comprehensive evaluation of all the diversified and now range from middle-income safeguard policies and Performance Standards of countries, with well-developed institutions and the World Bank Group. IEG previously evaluated capacities, to countries with weak institutions, individual safeguard policies, including involun- as well as fragile and conflict states. Second, tary resettlement (1998) and indigenous peoples the private sector has become an important (2003a). More recently, IEG undertook an evalua- development partner of the WBG, as business tion of environmental sustainability (2008b) that clients and financiers of development projects, covered the whole of the WBG. thereby increasing the importance of IFC and MIGA in the WBG’s overall portfolios. Third, the This evaluation aims to inform ongoing reviews WBG’s lending portfolio has steadily evolved of the environmental strategy of the World Bank from heavy reliance on stand-alone investment and Performance Standards of IFC to enhance the projects toward greater use of other financial effectiveness of future support. The evaluation instruments. Fourth, the services offered by also aims to inform the ongoing reform of World the WBG now include a wide range of analytical Bank investment lending. It covers the period and advisory services to build client institutions fiscal years 1999 to 2008 and is based on desk and capacities, such as Bank-supported Country xiv EXECUTIVE SUMMARY Environmental Analyses and technical assistance Equator Principles are voluntary standards for and IFC Advisory Services. The safeguards and determining, assessing, and managing social sustainability frameworks need to evolve to stay and environmental risk in project financing. relevant to this changing context. They were developed by private sector financial institutions and launched in 2003. Those institu- The volume of infrastructure and agricultural tions chose to model the Equator Principles on lending—the sectors with the most significant the safeguard policies of IFC. As of October 2009, environmental and social risks—fell signifi- 67 financial institutions had adopted the princi- cantly in the 1990s. Infrastructure and agricul- ples, which have become the de facto standard tural lending started growing again at the World for banks and investors on how to assess major Bank and IFC after 2003, and has increased even investment projects around the world. In July more rapidly during the recent crises, resulting 2006 the principles were revised after IFC’s new in renewed demand for safeguards expertise. sustainability policy was approved, increasing The World Bank’s portfolio has seen an even their scope and strengthening their processes to more rapid increase in types of lending to which match those of the Performance Standards. safeguards and Performance Standards are not well suited. Development policy lending for IFC’s Performance Standards were adopted institutional and policy reforms, programmatic by the European Bank for Reconstruction or sectorwide lending, and lending through and Development (EBRD) in 2008, with some financial intermediaries now comprise more changes, the most significant being the replace- than half the portfolio. Managing environmental ment of aspirational standards with performance and social effects is more challenging in financial requirements and the addition of a requirement intermediary and decentralized projects and for financial intermediaries. The growing share of in sectorwide and community-driven develop- support for the financial sector at IFC and MIGA ment programs, where use of proceeds are not suggests that more explicit guidance for financial fully identifiable at appraisal. IFC’s business intermediaries would benefit them too. has also evolved in recent years from project finance toward a growing portfolio of trade Effectiveness of World Bank Group finance and equity investments. IFC’s corporate Support of Safeguards and or equity investments in companies with several Sustainability Policies production facilities and various activities pose The WBG’s ability to mitigate social and environ- a substantial challenge for environmental and mental risks has improved significantly in the social appraisal, supervision, and evaluation. eyes of the public. About three-fourths of the MIGA’s portfolio composition has also shifted nongovernmental organizations responding to over time, with a significant increase in the an IEG survey rated WBG performance better share of guarantees for financial sector projects than in the 1990s, compared with 10 percent whose environmental and social effects are more that rated it worse. WBG clients interviewed by difficult to assess and supervise. These shifts in all IEG also acknowledged the contribution of WBG three portfolios present a challenge for the WBG environmental and social policies. to ensure continued relevance and effectiveness of the safeguards and sustainability policies. World Bank staff and management broadly support the objectives of the safeguards and Private Sector Ownership of Social and sustainability policies but, like clients, contest Environmental Sustainability some policy prescriptions more than others. The consistency of IFC’s Performance Standards The current social safeguard policies appear with the Equator Principles, which IFC helped to be more problematic than environmental create for private banks, appears to have increased policies because of the limited coverage of the their acceptance among the private sector clients social safeguards (Involuntary Resettlement and business partners of IFC and MIGA. The Policy, Indigenous Peoples Policy). Current xv S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D World Bank social safeguards do not provide disclosure, particularly to affected communi- adequate coverage of community impacts, labor ties, has not been adequate. The reliance on and working conditions, and health, safety, and self-assessment by business clients is a point of security issues at the project level, provisions vulnerability in the absence of full disclosure and that are integral to IFC and MIGA Performance independent verification. Due diligence by IFC Standards. The absence of an integrated approach for trade finance and projects with supply chains to social risks, combined with perceived rigidities to agribusiness—a highly sensitive area in terms in the application of the social safeguard policies of environmental outcomes—was found to be and continuing differences between the social inadequate. safeguards and national policies, impede broader dialogue with borrowers on a comprehensive The implementation of the Performance social policy. Standards needs a robust approach for identi- fying and addressing environmental and social Appraisal risks along the supply chain, particularly related Attention to safeguards and Performance to suppliers’ areas of influence on biodiversity, Standards is reasonably good during appraisal. forestry, and other environmental and social The aggregate quality of preparation and issues. IFC is following Environmental Health and appraisal was found to be 85 percent satisfac- Safety (EHS) guidelines in appraising projects, tory in World Bank projects approved in fiscal but EHS indicators have not been well integrated years 1999–2008. The quality of Environmental in monitoring and supervision instruments by Assessments, Environment Management Plans, IFC. Documentation on public disclosure and Resettlement Action Plans, and Social Assess- consultation emerged as one of the weaker areas ments was satisfactory in over 90 percent of the in IFC’s due diligence. projects reviewed. Underwriting of MIGA guarantees subject to The scale of involuntary resettlement induced the Performance Standards has improved, as by World Bank projects is quite substantial. The compared with projects underwritten subject total number of persons affected in the 10-year to the safeguards policies—in particular with portfolio sample was about 418,000, of which 41 respect to community consultations and the percent were physically displaced, with the rest assessment of clients’ social and environmental facing impacts on livelihoods. Extrapolating to the management systems (including, for the first full Bank portfolio, the resettlement induced each time, for financial sector projects) and labor and year by new projects affects an average of 166,500 working conditions. However, MIGA’s review of additional persons. Since the resettlement process projects with high or substantial risks is based on lasts several years, IEG estimates that at any given site visits by specialists in only a quarter of cases. time involuntary resettlement affects over 1 In addition, due diligence of financial sector million people, two-fifths of which are likely to be projects focuses on the social and environmen- physically displaced and three-fifths economically tal management systems at the level of corporate affected by active Bank-financed projects. policy of the parent banks, rather than at the subsidiaries supported by MIGA’s guarantee. In IFC projects, the appraisal process has been strengthened under the Performance Standards IEG also found that MIGA’s 2007 Policy on Disclo- framework by the improved Environmental sure of Information exempts projects under the and Social Review Document and a structured Small Investment Program (guarantees under approach to monitoring performance indicators, $10 million) from disclosure requirements. as recommended earlier by IEG. The quality of Since guarantees under this program are a large due diligence on safeguards and Performance proportion of MIGA’s post-2007 projects, this Standards during appraisal in IFC-supported weakness in its environmental and social assess- projects has generally been satisfactory. But ment framework must be addressed. xvi EXECUTIVE SUMMARY Categorization of projects based on environmen- and monitoring reports, third-party review, and tal and social risks differs across WBG and is not community monitoring instruments. based on use of objective criteria to assess social and environmental risks. Unclear guidance on There is no significant difference in the supervi- categorization is reflected in IEG’s finding that sion quality of IFC projects prepared under Perfor- several IFC and MIGA high-risk, category-B cases mance Standards from those prepared under (substantial impact) would have been classi- the previous safeguards policies—more than fied as category A (very high impact) projects a quarter having inadequate supervision. The using the World Bank’s screening system. Data quality of IFC’s supervision of financial interme- from the portfolio review show that the World diary projects approved before the introduc- Bank has rightly moved from an environmentally tion of Performance Standards has improved driven classification system toward one based on in recent years, though it remains significantly social and environmental risks. below appraisal quality. IFC’s supervision under Performance Standards is affected by the timeli- Within the World Bank, 9 percent of projects ness and quality of Annual Monitoring Reports were classified as category A. During the review prepared by clients. period, the proportion of category-A projects, on an annual basis, increased from 5 to 11 percent, MIGA has monitored category-A projects but only in keeping with the increase in the volume a third of category-B projects. The issues identi- and scale of infrastructure lending. Category-B fied through this limited monitoring suggest that projects increased from 37 to 51 percent, while category-B projects need to be more frequently category-C projects (low or no impact) dropped monitored to provide credible assurance that the from 40 to 18 percent. While this indicates Performance Standards are being met. greater caution during project preparation, the wide band of risks now subsumed under Performance on Safeguards and category B indicates lack of sufficient differentia- Sustainability tion among projects with substantial and more Environmental and social outcomes are a modest impacts. consequence of the risks assessed, the mitigation plans designed, and the quality of compliance Supervision and implementation by the client in partnership More than a third of World Bank projects had with the WBG. inadequate environmental and social supervi- sion, manifested mainly in unrealistic safeguards Supervision and monitoring deficiencies ratings and poor or absent monitoring and constrain the World Bank’s ability to evaluate evaluation. Results varied significantly by region, safeguards results. The World Bank does not with East Asia and Pacific having the best and have a clear framework to assess the perfor- Sub-Saharan Africa having the worst record. mance and impacts of its safeguard policies. Supervision quality was better in category-A Performance indicators are rarely specified projects, four-fifths of which were well supervised. and integrated in projects’ results frameworks, Staff incentives and unpredictability of resources and data for monitoring and evaluation are not for supervision constrain effectiveness. routinely collected or used. Completion reports for one-fifth of category-A and half of category- The increasing reliance in World Bank projects B projects lacked information on safeguards on policy frameworks is a cause for concern performance. This deficiency is more evident because these projects include multiple subproj- for environmental safeguards, whose impacts ects and are less well supervised than projects are more diverse and not as well tracked as for with project-specific risk assessments and mitiga- social safeguards. IEG was able to substantiate tion plans. Effective supervision would require mitigation of negative impacts in only two-thirds greater reliance on disclosure of supervision of the portfolio review sample. Other IEG studies xvii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D also reveal vital gaps in managing environmental Quality-at-entry and careful supervision are the risks induced by Bank-financed projects, in that WBG’s main mechanisms to ensure success- implementation and follow-up were deficient ful outcomes. The Inspection Panel found the even in projects for which the environmental majority of policy violations related to safeguard assessment had identified serious risks to nearby policies in World Bank projects originated natural habitats and biodiversity. in unresolved design issues, stemming from inadequate assessment of environmental or Environmental and social effects is one of four community impacts and inadequate consultation dimensions used to evaluate project develop- with affected people. One-fifth of the instances of ment outcomes of IFC and MIGA projects and noncompliance arose from inadequate attention an integral part of the Development Outcome to other implementation issues. Supervision Tracking System adopted by IFC. IFC has stream- provides an opportunity to deal with unantici- lined monitoring for its projects, which enables pated risks that arise during implementation. more systematic assessment of environmental Careful attention to complaints and resolution of and social performance, though the Perfor- genuine grievances can increase accountability. mance Standards are too recent to evaluate their environmental and social outcomes. Benefits and Costs The assessment of benefits and costs shows Investing in client capacity can lead to high that the WBG’s safeguards framework generates payoffs over the long term. Efforts to strengthen significant benefits for the mitigation of environ- social and environmental institutions in client mental and social risks of projects, even as these countries lag behind mitigation of immediate benefits need to be systematically measured risks within Bank-financed projects. IFC’s Perfor- or quantified. This study estimated benefits by mance Standards have provided a platform to extrapolating from the assessment of environ- balance their focus on mitigation with that on mental and social risks and comparing the sustainability by requiring an assessment of the results against available costs to analyze the client’s social and environmental management efficiency of resource use. IFC’s spending on system. While maintaining the focus on mitigat- its sustainability framework is being efficiently ing negative impacts, increased attention is being allocated toward projects with higher risks and paid to enhancing positive impacts and strength- benefits, but allocative efficiency is less evident ening client capacity under the Performance in World Bank spending on safeguards, particu- Standards. However, there is a need for greater larly among category-B projects. The delega- disclosure and verification of monitoring reports tion of responsibility for safeguards supervision by third parties and communities to ensure that to sector management units has contributed to desired environmental and social outcomes are supervision deficiencies. achieved. Costs incurred by World Bank clients on The Performance Standards have led to greater safeguards are estimated at about 5 percent focus on MIGA clients’ social and environmental of World Bank financing and 3 percent of total management systems. To address client capacity project cost. World Bank clients tend to allocate challenges, MIGA has established the Trust Fund resources efficiently in meeting safeguards to Address Environmental and Social Challenges requirements, but results cannot be established in MIGA-guaranteed projects in Africa; this is not for IFC clients because IFC does not collect client currently available to clients in other regions. cost data. Although there have been notable improve- ments, there is still a substantial gap in MIGA’s Benefit-cost analysis can provide useful insights ability to monitor implementation performance into environmental and social performance. and provide assurance that the objectives of the Benefit-cost analysis of two stylized models of Performance Standards are being met. World Bank and IFC projects illustrates that, on xviii EXECUTIVE SUMMARY their own, the benefits of safeguards outweigh using policy frameworks has been worse than the costs. However, the benefits of World Bank other projects due to weak follow-up during projects are more muted due to the narrow scope implementation. of the current social safeguards. The potential of IFC projects is enhanced by the additional The Performance Standards paradigm is based benefits from attention to labor conditions on a commitment by the private sector client and community impacts. Better monitoring, or partner to the principles of the policy and documentation, and reporting of environmental the Performance Standards to be achieved, and social impacts are needed to improve the with covenanted remedies if the standards are quality of benefit-cost analysis. not met. IFC places greater responsibility for implementation and monitoring of performance Paradigms for Achieving Environmental indicators specified by IFC on the business client, and Social Results and envisages supporting this with supervision The WBG is using two paradigms to address and documentation of performance. By the environmental and social risks: the safeguards same token, the implementation and report- paradigm of the World Bank, largely for the ing on environmental and social effects are the public sector, and the Performance Standards responsibility of the private sector client. The paradigm of IFC and MIGA for the private sector. crucial question is whether this self-assessment The two share similar objectives and have differ- by the private sector captures public concerns, ent strengths and weaknesses. The Bank seeks which in the final analysis can only be judged by “to avoid, mitigate, or minimize adverse environ- the environmental and social results achieved. mental and social impacts of projects supported Greater disclosure of environmental and social by the Bank” and to ensure that they are “environ- information, including to local communities, mentally sound and sustainable.” IFC applies the and verification of results are needed to capture Performance Standards “to manage social and these public good concerns. environmental risks and impacts and to enhance development opportunities in its private sector The introduction of the Performance Standards financing.” is too recent to compare their results with those under the safeguard policies. However, in certain The safeguards paradigm contains mandatory respects there is evidence that the instruments requirements, with specific mitigation and practices introduced by IFC, along with measures designed before project approval, the Performance Standards, have advantages just as traditional projects are fully designed at over those of the World Bank. IFC’s systems appraisal. If national regulations differ from the include balanced thematic coverage of social and Bank’s safeguard policies, clients need to accept environmental risks, including more relevant the higher standards of the Bank’s policies. In Performance Standards on labor conditions and practice, supervision has focused largely on community impacts. Compliance with standards compliance with the mitigation plan rather than is covenanted by IFC in most legal agreements, on monitoring outcomes. In high-risk projects, while for World Bank safeguards this is more responsibility for clearance and oversight lies prevalent among category-A projects. Annual with the safeguards advisors and environmen- monitoring reports with performance indica- tal and social units in the regions. Responsibility tors are required from clients. However, this for processing and supervision of other projects self-assessment by the client is not currently is delegated to sector units managing the verified independently to ensure compliance projects. Rules have been modified for financial and results. IFC undertakes supervision and intermediary and other decentralized projects, performance reviews. And IFC and IEG evaluate which are now allowed to replace risk assess- environmental and social effects as one of ments with policy frameworks. However, the four dimensions of the project’s development performance of delegated projects and projects objectives. xix S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D The World Bank’s systems are front-loaded verification, and disclosure. Without investing and have more specific requirements. They in these elements in the right sequence, that is, emphasize screening, risk assessment, and ensuring enhanced supervision and monitoring appraisal of the proposed mitigation plans to with client capacity, disclosure, and verification ensure compliance with the requirements. systems in place, the Performance Standards These elements have served well in safeguarding approach would be a riskier option with respect high-risk (category-A) projects. Supervision and to environmental and social outcomes. monitoring receive far less attention, especially in category-B projects, affecting implementation Country-Level Work quality and leading to highly uneven results. The limitations of the safeguards paradigm become visible in the World Bank pilots on the The coverage of the environmental policies across use of country systems. The country systems the WBG is similar in scope, but the World Bank’s pilots were an attempt to adapt to the changing social safeguards are more limited than those of context, but the requirements spelled out in IFC and MIGA. The frequency with which the the operational policy (OP 4.00) governing the safeguard policies and Performance Standards pilots were overly prescriptive and excessively are triggered is an indication of their relevance focused at the project level. The design to the portfolio. About two-thirds of World constraints governing the pilots have prevented Bank investment projects approved since 1999 their application at the country level. The only triggered environmental assessment. One-fourth large-scale successes have been with parastat- triggered the involuntary resettlement policy; als that agreed to adopt the Bank’s safeguard other safeguards occurred more rarely. IFC’s suite policies, but these are not country systems. The of Performance Standards was found to be more pilots have not yet been effective in integrating relevant to the portfolio: four of them—including social safeguards at the country level, and the the standards on labor and working conditions, piecemeal approach to safeguards in the pilots and on community health, safety, and security— has reduced the likelihood that any borrower applied to half of IFC’s total portfolio and to 90 will be able to adopt the entire suite of safeguard percent of all projects with high or substantial risk. policies or that the country systems approach can be scaled up. In contrast, the uptake of A paradigm that is based on more relevant environmental development policy loans, which thematic coverage, procedural flexibility (but is always underpinned by country analytic work, without compromising on the integrity of suggests that the parallel work on country-level standards), and client responsibility and capacity assessments and environmental development for monitoring seems to lead to more client policy lending has been well received, particu- ownership. However, the quality of implemen- larly in the Latin America and Caribbean and the tation and monitoring, which depends on client South Asia Regions. capacity and commitment, needs to be adequate, and checks and balances must be in place to At the country level, the broad nature of the ensure that intended social and environmental environmental assessment policy has provided outcomes are achieved. The prior existence of the the Bank with a vehicle to engage in country- safeguard policies provided a critical benchmark level policy dialogue to help countries put in for Performance Standards. In this context, four place economywide policies and institutions elements that build on the strengths of both are for environmental sustainability. This alternate vital: investment in the clients’ social and environ- modality of client engagement has created a mental management system; integration of parallel stream of work, with greater ownership adequate environmental and social performance among clients and environmental specialists indicators in the project’s results framework; than the work on safeguards. There is a need to effective instruments for monitoring by the client; bridge this divide between the Bank’s work on and regular supervision, performance review, safeguards and environmental sustainability. xx EXECUTIVE SUMMARY Neither the environmental dialogue at the policies; strengthen compliance, implementa- country level nor the country systems pilots tion and accountability; and improve clients’ and have included social safeguards. Though the the WBG’s ability to promote positive social and social safeguards for resettlement and indige- environmental results: nous peoples help the World Bank mitigate risks of impoverishment arising from unintended 1. Revise the policy frameworks to har- consequences of Bank-financed projects, the monize thematic coverage and guid- restrictive and prescriptive nature of the current ance across the WBG and enhance the social safeguard policies limit the prospects for relevance of those frameworks to client systemic dialogue at the country level. With the needs. exception of the country dialogue on indigenous peoples in some Latin American countries and a • IFC, MIGA, and the World Bank should jointly few attempts to broaden the scope of resettle- adopt and use a shared set of objective crite- ment, as in the incomplete effort on resettle- ria to assess social and environmental risks ment in India, social safeguards have focused to ensure adequacy and consistency in proj- primarily at the project level, resulting in missed ect categorization across the WBG, using the development opportunities for the very people more inclusive criteria for category A, and re- who need it most. IFC and EBRD have overcome fining the categorization system to address the local resistance by assessing impacts on indige- bunching of higher- and lower-risk projects nous peoples as an integral part of community within the current category B. impacts, but this alternative does not exist under the Bank’s current framework. The World Bank should: Recommendations • Ensure adequate coverage of social effects— The WBG’s safeguards and Performance integrating community and gender impacts, Standards play a critical role in ensuring adequate labor and working conditions, and health, attention to environmental and social outcomes. safety, and security issues not currently cov- Given the changing nature of its clients and ered by its safeguard policies—by consolidat- portfolios, the challenge is to ensure the contin- ing existing social safeguards with other WBG ued relevance and effectiveness of the WBG’s policies on social risks as requirements under environmental and social policies while comple- one umbrella policy on social sustainability. menting the emphasis on compliance with • Consolidate the environmental policies as effective implementation. The evaluation points requirements under one umbrella policy on to the need for a systems approach, balancing environmental sustainability. up-front risk assessment with implementation • Revise the current approach to safeguards pi- support to increase effectiveness; policy consoli- lots on use of country systems to focus on dation with more comprehensive, balanced strengthening country institutions and systems thematic coverage to ensure adequate up-front to manage environmental and social risks. regulations while providing for better supervi- sion, monitoring and evaluation, verification, IFC should: and disclosure; and partnership with clients, third parties, and local communities to enhance • Strengthen the provisions on sustainability ownership and results, integrating elements of to address emerging issues, notably climate the Bank’s safeguards with some of the practices change and supply chains and their commod- under IFC’s Performance Standards. ity certification. • Develop more robust approaches to the im- The following recommendations to the World plementation of the Performance Standards in Bank, IFC, and MIGA are made to help maintain financial intermediary projects, listed equities, the objectives of safeguards and sustainability and trade finance. xxi S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D • Strengthen policies and practices on disclo- • Use advisory services to build social and en- sure, including at the local levels. vironmental management systems and imple- mentation capacity, especially among small MIGA should: and medium enterprises, financial intermedi- aries, and clients in countries and sectors with • Increase the capacity of the Environmental weak environmental and social management. and Social Unit to the level needed to provide • Mobilize resources at appraisal for energy and credible assurance on performance against the clean production audits, using auditors with standards for every project. Should MIGA be relevant sector knowledge. unable to increase its resources devoted to • Define areas of influence and requirements to implementation of Performance Standards, better address supply chain risks and opportu- it should revise its Policy on Social and Envi- nities, particularly ones related to biodiversity ronmental Sustainability to disclaim any re- and forestry, and expanding the application of sponsibility for monitoring the projects’ social material biodiversity along the supply chain and environmental performance and ensuring for suppliers. that they comply with the standards. Under this option, MIGA’s role would be limited to MIGA should: reviewing the client’s assessment of the proj- ect’s environmental and social risks against • Focus the due diligence reviews of financial the standards, identifying corrective actions as sector projects on the Social and Environ- needed, and securing the client’s commitment mental Management Systems of developing- to implement these actions. country subsidiaries the project supports, • Require that category-B, Small Investment Pro- rather than the corporate policies of the par- gram projects follow the same disclosure re- ent banks. quirements as for regular category-B projects. • Expand the size and eligibility of the Trust Fund for Addressing Environmental and Social 2. Enhance client capacity, responsibility, Challenges to all low-capacity clients on the and ownership. basis of need. The World Bank should: 3. Revise guidelines, instruments, and in- centives to strengthen supervision ar- • Increase the synergies between safeguards rangements. work and broader Bank engagement on en- vironmental and social sustainability by in- The World Bank should: vesting in upstream analytical work, technical assistance, and lending to strengthen country • Assign responsibility and budget for safeguards and sector institutions and capacities in client oversight and reporting to environmental and countries. social units in each operational Region—in • Require regular reporting by the borrower on line with IFC practice—in place of the delega- implementation and outcomes of safeguards tion of safeguards processing and supervision in Bank-supported projects, and work with to sector management units. clients to develop instruments and indicators • Introduce a certification program to expand the to help in such monitoring. pool of staff qualified to undertake social and environmental preparation and supervision IFC should: while ensuring quality and consistency, and provide orientation training on environmental • Develop incentives for investment officers and social sustainability to all task team leaders. to share ownership of the Performance Stan- • Develop and implement an action plan to en- dards and mainstream their implementation. sure regular supervision of financial intermedi- xxii EXECUTIVE SUMMARY ary projects and investment projects that use sion when assessing achievement of the proj- social and environmental policy frameworks ect’s development objective, as has already through third-party or community monitor- been done for IFC and MIGA. ing for higher-risk projects, and disclosure of monitoring and supervision reports. IFC should: IFC should: • Disclose project-level environmental and so- cial information from monitoring and supervi- • Enhance the supervision of financial interme- sion reports. diaries at the subproject level by developing • Make use of independent, third-party, or com- clear guidelines for applying the Performance munity monitoring and evaluation for its proj- Standards at the subproject level and by adopt- ects, particularly for projects with involuntary ing a systematic approach to environmental resettlement and higher-risk financial interme- and social specialists’ site visits to selected diary and agribusiness projects. subprojects. • Use loan covenants, including Conditions of MIGA should: Disbursement to enforce compliance with environmental and social requirements and • Disclose project-level environmental and so- reporting if the clients lack commitment and cial information from supervision reports. are continuously out of compliance. • Develop a credible mechanism to ensure that Performance Standards are adhered to by fi- 4. Strengthen safeguards monitoring, evalu- nancial sector projects. ation, and completion reporting. 5. Improve systems and instruments for The World Bank should: accountability and grievance redress. • Include performance indicators on environ- IFC, MIGA, and the World Bank should: mental and social outcomes in project results frameworks and ensure systematic collection • Seek greater symmetry in the structure of of data to monitor and evaluate safeguards WBG accountability and grievance redress performance. mechanisms. For the World Bank this would • Ensure that Implementation Completion Re- entail creation of a grievance redress and con- ports and IEG reviews of those reports rate flict resolution mechanism to complement and report effectively on the outcomes of safe- the Inspection Panel. For IFC and MIGA this guards and, for all projects with significant would entail a more independent compliance environmental and social effects, ensure the review process, ensuring that the CAO submits results are incorporated as an essential dimen- its audits directly to the Board. xxiii Construction of a pipeline in Peru. Photo courtesy of Jouni Martti Eerikainen. xxiv Management Response Introduction Management’s specific responses to IEG’s Management welcomes this evaluation of the recommendations, with which it broadly agrees, World Bank Group’s (WBG) environmental and are noted in the attached Management Action social safeguard frameworks, covering the period Record. fiscal years 1999–2008, by the Independent Evaluation Group (IEG). This evaluation is timely Management Observations – IBRD/IDA as the WBG moves to completion of an updated Overview Environment Strategy, and as the International Bank Management agrees broadly with many of Finance Corporation (IFC) undertakes a review the key findings and recommendations of the and update of its Sustainability Framework. It IEG regarding the need for: greater emphasis on also is relevant as the Bank has begun an exercise the use of safeguard polices to support environ- in Investment Lending Reform, and the WBG mentally and socially sustainable development; faces emerging circumstances in the post-2008 greater emphasis on assessment of a wider range financial crisis that require new and innovative of potential social impacts and risks; improve- financial instruments to meet client needs and ments in supervision; staff certification/accredi- restore economic growth at a global as well as tation; more efficient and effective approaches to national level. monitoring, evaluation, and completion report- ing, including the enhanced use of indicators Management welcomes the overall positive and the value of creating a grievance redress findings of the evaluation. IEG’s findings show mechanism. Bank Management particularly sees that the WBG has made significant progress value in considering further and pursuing more during the evaluation period in improving consistently its role in encouraging borrowers its efforts to safeguard the environment and and project implementing entities to report to vulnerable communities from unintended local communities and stakeholders on how they consequences of the development process. It are managing environmental and social impacts also confirms the importance of linking sustain- and risks as part of Bank-financed projects. At able development with safeguard policies the same time, Bank Management agrees broadly and performance standards to promote the that it should take into consideration risk factors broad objectives of the WBG. Among the arising from project design objectives, with many important findings of this evaluation is greater attention to both the project’s context the recognition by many stakeholders that the and impact and risk assessment. Bank Manage- WBG’s performance on environmental and ment also notes that the IEG found that the signif- social impacts has improved during the evalua- icant investments during the 1990s led to success tion period, especially with respect to high-risk in developing legislative frameworks and client projects. Management also notes IEG’s finding on capacity, which provide in many countries the the clear benefits of developing and implement- foundation for country ownership as endorsed in ing a more holistic approach to environmental the Paris Declaration1 and the Accra Agenda for and social impact assessment and risk manage- Action.2 It also appreciates the IEG’s finding that ment that is appropriately aligned with interna- the current approach to Use of Country Systems tionally recognized good practice relevant for (UCS) in the pilot program has been rigorous in the work of each of the WBG units. its approach, which leads Bank Management to xxv S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D propose that it may soon be time to scale-up and will support Bank Management’s commitment to broaden the approach as suggested by the Board reinforcing and enhancing the effectiveness of discussion of the Pilot Program in January 2008. current policies and increasing the emphasis on activities that lead to beneficial and sustainable Updating and Consolidation of Safeguard outcomes. We believe an updated policy, comple- Policies mented by support for institution and capacity Bank Management recognizes, as does IEG, the building, will help manage environmental and importance of undertaking a comprehensive social impacts and risks in Bank-supported invest- updating and consolidation of the safeguard ment projects, especially in low-capacity and policies. This should include a rigorous consid- fragile countries. Alternatives will, of course, be eration of those aspects of the IFC Perfor- subject to cost estimates. mance Standards (PS) for the private sector that represent a modernization in approach toward The joint team will be tasked with defining the environmental and social standards possibly specific objectives of the review, outlining the lacking in parts of the World Bank’s current process (including consultation and engagement suite of safeguard policies. In this context, with internal and external stakeholders), develop- Bank Management plans to use a joint team ing a timeline, and preparing recommendations led by the Sustainable Development Network to enhance the development effectiveness of the (SDN), Operations Policy and Country Services Bank’s safeguard policies. This review of global Network (OPCS) and the Legal Vice Presidency good practices will provide the opportunity for a (LEG), to engage during the next 24 months in a discussion of current and emerging practices in learning and consultative process with a diversity both Part I and Part II countries at the national of shareholders and stakeholders on global and subnational level. Consultations would be good practice and integrate this dialogue into undertaken with governments at various levels, an update of the Bank’s approach to safeguard private sector representatives, academic and policies in its projects. At the conclusion of applied research institutions, professional associa- this process, Bank Management will report to tions, civil society organizations, nongovernmen- CODE and the Board how it intends to further tal organizations and other stakeholders. The strengthen environmental and social sustain- Bank will explore, as part of this process, policy ability in its projects, including presentation, for and regulatory instruments that can be used their consideration, of a policy paper setting out in addition to environmental and social impact its updated and consolidated approach. assessments, to both mitigate adverse impacts and also enhance support for broader environ- The review process will focus on developments mental and social sustainability in projects. that allow the Bank to achieve outcomes with greater environmental and social sustainabil- Use of Country Systems ity, and on helping clients build institutions that Bank Management agrees that the approach to can effectively pursue such outcomes. The IEG UCS for Environmental and Social Safeguards evaluation will be used as a reference point for to date has been relatively narrow, particularly the consideration of various issues and options. A with regard to the provisions of Operational period of 24 months has been proposed to provide Policy (OP) 4.00 concerning the determina- adequate time to undertake an interactive review tion of equivalency as approved by the Board in process, develop a draft umbrella safeguard policy, March 2005. Bank Management also notes that prepare translations, and conduct consultations if a change in the approach toward UCS is to within and outside the WBG. Periodic briefings occur, the Bank needs to endorse a more flexible are planned to be held on a regular basis with application of OP 4.00. Bank Management representatives of CODE and the Board to keep believes that although the current approach has them apprised of developments and to seek their been appropriate for a pilot-stage program, a guidance to the joint team as needed. This process more robust approach is warranted for broader xxvi MANAGEMENT RESPONSE application of UCS in the future. At the same While the review on global good practice is time, Bank Management fully supports increased under way, Bank Management will address this emphasis on strengthening country institutions concern on an interim basis by having SDN, and systems to manage environmental and social OPCS, and LEG prepare and issue guidance by impacts and risks especially in low-capacity and the third quarter of fiscal 2011 on the scope fragile countries. and coverage of social issues in the context of the preparation and implementation of en- Bank Management believes that, based on vironmental assessments. This guidance will the results of the pilot program, there is now ensure a more balanced approach between sufficient experience to begin a mainstreamed environmental and social risks and impacts, as program with a revised version of OP 4.00 well as the identification of actions to support that moves beyond the pilot stage and takes more sustainable social benefits from Bank- Bank practice closer to the vision of the Accra supported projects. Agenda for Action, in which UCS increasingly • Occupational Health and Safety. Occu- becomes the default approach. Bank Manage- pational health and safety, working condi- ment is currently reviewing the implementa- tions, and security are recognized as issues tion of OP 4.00 and plans to expand this work that deserve more explicit recognition in the in the context of the global good-practice review Bank’s safeguard work. SDN, OPCS, and LEG in order to develop proposed revisions to the will expand their work with the Regions to policy that scale up, broaden, and mainstream ensure a heightened awareness of the need its use. As part of this review, consideration will to address these concerns in the context of be given to several options, including the default project preparation, appraisal, and supervi- use of country systems for European Union and sion. OPCS has already incorporated these advanced European Union-accession countries, topics more explicitly in its safeguard train- as well as countries whose environmental systems ing program, which entails a rotating series and performance are conditioned by interna- of courses on various safeguard topics on a tional trade agreements. At the end of the global weekly basis. These topics have been incor- good-practice review period, Bank Management porated in three of the courses offered by plans to recommend revisions to OP 4.00 to the the OPCS-sponsored training program: (a) Board as part of its overall approach. Overview of Safeguards; OP 4.01, (b) Envi- ronmental Assessment; and (c) Guidance to Selected Areas for Action Staff Working on Joint Bank-IFC projects. In Bank Management would like to highlight several the past year, OPCS has also developed and areas for action in response to the findings of the offered a new course on the WBG Environ- IEG report: mental, Health, and Safety Guidelines, which includes many aspects of occupational health • Coverage of Social Issues. Bank Manage- and safety and working conditions. Priority is ment agrees with the recommendation that a also being given to providing selected staff in more comprehensive and balanced approach the Regions with more intensive specialized to social issues would be useful in supporting training on these issues from external sources. the broad development objectives of the Bank Bank Management will undertake an outreach and bring greater consistency between the and training program on these issues during Bank and IFC, while recognizing the difference fiscal 2011 and will more broadly disseminate between public and private sector clients. A the WBG Guidelines to both staff and borrow- more balanced coverage of those social issues ers as part of this process. relevant in the public sector context would • Monitoring, Evaluation and Use of Indi- allow a stronger emphasis on social opportuni- cators. Bank Management agrees with the ties, impacts and risks not currently specifically need to strengthen monitoring and evalua- covered by safeguard policies and guidance. tion arrangements. To address this issue, SDN, xxvii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D OPCS and LEG will collaborate on developing numbers of people requiring resettlement or guidelines on monitoring and evaluating safe- who are at risk for potential impacts on their liveli- guard performance for selected types of proj- hoods. Bank Management recognizes that this ects by the third quarter of fiscal 2011. These may have been at the expense of applying staff guidelines will focus on more systematically time and budget resources to supervising those measuring outcomes, including through the projects considered less risky. Bank Management use of core environmental and social perfor- agrees with IEG that there is a need to strengthen mance monitoring indicators, and on evaluat- supervision for medium- and low-risk projects. ing impacts. Reporting will be integrated in How this will be done may need to differ from the Implementation Status and Results Report region to region, depending on country capacity (ISR), building on the new Operational Risk and project type and mix. As of July 1, 2010, Assessment Framework under the Investment enhanced implementation support has been Lending Reform which is already enhancing launched as one of the key pillars of Investment the monitoring of environmental and social Lending Reform, including the launch of the new risk mitigation measures in ISRs. The proposed ISR template. A major component of the reform guidelines on monitoring and evaluation will is improved monitoring and reporting on risk further emphasize the need for the Implemen- and on measuring progress in the implemen- tation Completion Report (ICR) to evaluate the tation of risk mitigation measures, including achievement of the safeguard objectives and those related to social and environmental risks. identify lessons for future projects. The information will be used to better align implementation support budgets with risk. Inclusion of Safeguards in Legal Agreements Bank Management plans to undertake a review Bank Management notes that the IEG found by the second quarter of fiscal 2011 concerning that inclusion of safeguard-related provisions current practices with respect to responsibility, in IFC legal agreements under the IFC Perfor- accountability, incentives, staffing, and budget- mance Standards is more prevalent than in Bank ing for safeguard processing and supervision. projects, other than for World Bank Category- Based on this review, practices will be updated A projects. Bank Management notes that this with the objective of enhancing effectiveness and finding was limited in scope to the Category-B efficiency and maximizing the synergies between projects in the sample survey and is not conclu- safeguard work and broader Bank engagement sive as to the percentage of World Bank projects on environmental and social sustainability. that use safeguard covenants. However, Bank Management takes this finding seriously, and Projects Using Frameworks and Third Party will undertake a review to evaluate the matter Monitoring further to determine whether there is any gap Bank Management recognizes that the use in coverage in Category-B safeguard-related of various safeguard framework instruments covenants. As part of the review, Bank Manage- (Environmental and Social Management ment will discuss with IFC counterparts IEG’s Framework, Resettlement Policy Framework, approach toward covenants, including differ- Indigenous Peoples Plan Framework), initially ences that might be related to private sector designed to be used as an appropriate instrument versus public sector projects. In addition, the as part of Financial Intermediary (FI) lending, Bank’s Legal Vice Presidency has already set up has become more widespread as the nature of a program for training for all LEG lawyers to Bank lending moves increasingly to projects that become better versed in safeguard application. use a programmatic approach and support the use of subprojects. Bank Management agrees Supervision that improvements are needed in supervision of IEG found that the Bank has done well in address- projects that rely on frameworks as the appropri- ing environmental and social issues in the most ate safeguard instrument for such projects. Bank risky projects, especially those involving large Management is currently engaged in a Bank-wide xxviii MANAGEMENT RESPONSE review of the use of frameworks that will examine to the IFC Compliance Advisor/Ombudsman these types of projects and identify good (CAO) or other multilateral financial institutions. practices. The review will include an examina- Bank Management will present the results of this tion of a variety of means to strengthen monitor- study to the Board to ensure that any decisions ing of such projects including, in appropriate emerging from the study will be consistent with situations, the use of third-party or community the Board Resolution and related Board decisions monitoring for selected projects. The review is concerning the Inspection Panel, and in a manner expected to be completed by the third quarter of that takes fully into account the current require- fiscal 2011 and will provide the basis for guidance ments and experiences with project-based to be issued by SDN, OPCS, and LEG for use by grievance mechanisms (including as required Bank staff and borrowers by the fourth quarter under OP 4.12, Involuntary Resettlement, and of fiscal 2011. OP 4.10, Indigenous Peoples). This study will be coordinated among Bank units with considerable Grievance Redress and Conflict Resolution experience in this field to ensure institutional Mechanism coherence and efficiency. Based on this study and Bank Management agrees with IEG that there is the aforementioned consultations with the Board, value in creating a grievance redress mechanism subject to cost considerations, Bank Management for which Bank Management will take respon- will establish a grievance mechanism by the first sibility and which is complementary to, but quarter of fiscal 2012, and provide to the Board separate from, the Inspection Panel. Because a detailed report on the initial operation of the this mechanism would seek to resolve grievances grievance mechanism by the end of fiscal 2012. without examining the issue of compliance with Bank policies, use of this mechanism would not Management Observations – IFC be a precondition to review of requests by the Overview Inspection Panel. Management takes note that IFC Management welcomes the evaluation by the similar complementary systems are now in place IEG and appreciates IEG’s endorsement of the at the IFC, and at other multilateral financial policy and implementation direction IFC has set in institutions, such as the Asian Development recent years. All IEG recommendations and sugges- Bank. A grievance mechanism can also help to tions will be considered in the ongoing review complement the Bank’s current accountability and update of IFC’s Sustainability Framework, system by promptly responding to concerns, consisting of the Sustainability Policy, Performance including through the use of local mediators and Standards and Disclosure Policy, as well as in the facilitators. Management wishes to underscore context of IFC’s ongoing supervision of its portfo- that establishing this mechanism would not alter lio projects. IFC Management concurs with many the responsibility of borrowers and recipients for of the aspects of IEG’s findings and recommenda- implementing projects, and that in many cases, tions, such as the recommendation to harmonize the grievances are not necessarily with the Bank, thematic coverage and guidance across the WBG, but between our clients and project-affected including categorization. IFC Management has also people. Nevertheless, these grievances are often identified a number of implementation challenges, brought for resolution to the Bank. which are described below. Specific action-based responses to IEG’s recommendations are provided Therefore, by the end of the third quarter in the attached MAR. of fiscal 2011, Bank Management intends to complete a survey and review of a wide range of Observations concerning Recommended potentially analogous existing grievance-redress Actions mechanisms as a basis for designing one for In a limited set of cases the recommended the Bank. The study will include a review of the actions proposed by IEG need to be adapted to cost implications and potential cost savings that the IFC business model. This applies particularly could be engendered by using a system similar to the areas listed below: xxix S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D • Supply Chains. IFC’s approach to supply risk categorization for FIs: high, medium, and chains has been to focus client actions on the low, with risk-appropriate due diligence re- most immediate and serious risks in their sup- quirements. With regard to listed equities, IFC ply chains—such as child labor, forced labor, considers legal and regulatory frameworks in and potential clearing of critical habitats. In the development of options that are tailored addition, IFC proposes to: (i) strengthen its to the respective framework in which the in- supply chain assessment methodology as part vestee companies operate. This informs the of appraisal; (ii) make changes to the Perfor- way that IFC can include Performance Stan- mance Standards by adding significant safety dards—either in Shareholders Agreements, issues as a new risk factor to be considered Policy Agreements directly with the Company, in the supply chain assessment; and (iii) con- or adopted by the Company’s Board of Direc- tinue supporting certification schemes, both tors (and/or incorporated into the Company’s through investment projects and advisory Charter) before IFC invests. The approach services, including engagement in a number used by IFC for the existing trade finance pro- of global commodity roundtables. However, grams (e.g., Global Trade Finance Program, IFC Management notes that the number of Global Trade Liquidity Program) is to apply the credible certification schemes is still limited, exclusion list and undertake regular reviews of and that these schemes could disadvantage this application. IFC will review this approach small-scale producers and suppliers, particu- from a corporate risk perspective and as ad- larly those in emerging markets. In addition, ditional trade finance products are developed IFC clients have varying degrees of control of and launched. or influence over their supply chains. • Local Disclosure. IEG recommends IFC to • Financial Intermediaries, Listed Equi- improve its disclosure practices, including ties, and Trade Finance. IFC welcomes local disclosure to ensure access to informa- IEG’s finding that the quality of Performance tion by affected communities and other key Standards implementation in FI projects has stakeholder groups. IFC Management agrees improved considerably, with a high quality with this recommendation. The existing IFC of appraisal (IEG nonetheless noted that Disclosure Policy takes a hybrid approach, the quality of supervision is still lower than specifying the types of information IFC will that of appraisal), and that this is starting to disclose, subject to a list of exceptions (which translate into outcomes. There are unique is a standard approach among private sector challenges associated with financial interme- financial institutions). The Policy emphasizes diaries. IFC’s development impact is achieved disclosure of information up to the point of through the emergence of robust FIs, able to Board approval. Management is currently con- manage financial and nonfinancial risks well. sidering a number of revisions to its Disclosure IFC has no contractual relationships with the Policy in order to enable more disclosure of FI’s subprojects. IFC’s due diligence process information throughout the IFC investment therefore involves a risk-based approach, tak- lifecycle, including disclosure of development ing into account the nature of IFC’s exposure impact during project implementation. How- to environmental or social risks in the FIs’ ever, this will not include disclosure of all en- portfolio, and focusing on their capacity and vironmental and social reporting from clients effectiveness in developing and implement- since there has to be a balance between client ing a Social and Environmental Management confidentiality and disclosure. IFC will con- System (SEMS). In the case of FIs with higher- tinue its practice of holding its clients respon- risk portfolios, IFC ensures a right to review sible for reporting to the local community. IFC select subprojects before they are approved. understands that not all communities have In the proposed revisions to the Sustainability access to the information that it discloses, and Policy, IFC is further refining its risk-based ap- will be reviewing translation requirements in proach with the introduction of a three-tiered line with the WBG Translation Framework. xxx MANAGEMENT RESPONSE Supervision a risk-based approach, and IFC will conduct The IEG report points out that the quality of its supervision of subprojects on a selective supervision has improved. Many of the challenges basis. The methodology and selection criteria to implementation and supervision specified by for this will be developed as part of the review IEG are consistent with those identified through and update of the Sustainability Framework. IFC’s ongoing review and update of its Sustain- ability Framework. IFC Management agrees with Accountability and Grievance Redress IEG’s identification of the key challenges, and has IFC Management recognizes that as a private the following observations on the recommenda- sector development institution it is key to have tions: an independent office with effective compliance and mediation functions. IFC Management notes • Capacity of Clients. The capacity of clients, that the recommendation with respect to the especially in higher-risk country contexts or CAO will not require any management action. At industries, can present obstacles to imple- the same time, IFC Management has not noted mentation of Performance Standards. IFC’s any concerns with the effectiveness of the current environmental and social specialists routinely grievance and mediation function. IFC Manage- provide support to its clients during appraisal ment is awaiting with interest the outcome of and supervision. However, there are situations the Board’s ongoing reviews of oversight and where clients would benefit from additional or accountability mechanisms. more targeted support, and use of advisory services is an option that IFC can use for this Management Observations – MIGA purpose, especially for low-capacity clients in Overview high-risk environments. IFC is currently work- The Multilateral Investment Guarantee Agency ing to define how best to target and prioritize (MIGA) management thanks the IEG for its this type of support. evaluation, and would like to comment on • Third Party Monitoring. IFC Management several points raised that specifically address agrees with some aspects of IEG recom- MIGA’s capacity and performance. mendations on third-party monitoring. IFC proposes to selectively make greater use of • Improved Performance. MIGA Manage- third-party monitoring, including participa- ment welcomes the IEG finding that MIGA’s tory monitoring where practical, particularly social and environmental preparation and ap- in higher-risk situations. A key priority for IFC praisal under the Performance Standards have is to strengthen client capacity and owner- improved compared with projects prepared ship for environmental and social issues, and under previous safeguard policies; and that all third-party monitoring should be viewed in the projects approved under the Performance that context, not just in terms of independent Standards were found to be satisfactory by verification. IFC Management agrees that there IEG in terms of identification and screening, should be independent grievance mechanisms disclosure and consultation, and preparation where possible, and this should be combined and social and environmental appraisal. with regular engagement and outreach con- • Environmental and Social Trust Fund. ducted by the client. MIGA Management appreciates and agrees • Subproject Level Supervision. IFC Man- with the recommendation that the Japan- agement agrees that supervision of financial supported Environmental and Social Trust intermediaries should focus on the overall Fund for Africa should be expanded in terms environmental and social management system of size and eligibility, to allow projects in other of the financial intermediary and include some regions to be eligible for Trust Fund support level of subproject oversight consistent with as well. MIGA believes that the first three pilot a risk-based approach. Monitoring of all sub- years showed that the Trust Fund has been projects is not appropriate or required under valuable and the evaluation notes that MIGA’s xxxi S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D follow-up client surveys indicate that client keeping its current risk-based approach, MIGA satisfaction with these activities has ranged Management will consider whether more re- from very good to excellent. MIGA has started sources can and should be allocated to project taking steps to extend the Trust Fund, includ- monitoring to allow for more frequent visits to ing initiating contact with several potential complex projects. The IEG alternative of MIGA donors. MIGA Management sees considerable declaring it will not monitor projects’ environ- value in maintaining and extending this facility, mental and social performance is not consis- but at the same time notes that the ability to tent with MIGA’s mandate as a development comply with this recommendation will largely institution, and MIGA Management would like be driven by the willingness and capacity of to know the rationale for this alternative. MIGA external donors to be involved. Management instead recommends continuing • Small Investment Program. MIGA Man- with its risk-based approach, but increasing agement partially agrees with IEG’s recom- the number of site visits for those projects mendation that Category-B Small Investment where issues are most likely. At the same time, Program (SIP) projects follow the same dis- MIGA would maintain routine contact with closure requirements as regular Category-B those projects where the risks are viewed to projects. MIGA will review its Disclosure Policy be less. after IFC completes its own review in fiscal • Financial Sector Guarantees. Finally, in the 2011 and will make changes as warranted at case of financial sector guarantees, IEG rec- that time. In the meantime, MIGA will continue ommends MIGA focus on the SEMS specifi- to post the Summary of Proposed Guarantee cally of the project enterprises, rather than the (SPG) for all SIP projects, which may include corporate policies of the parent banks. MIGA a more detailed explanation of environmental Management appreciates this comment, but and social issues, and will attach the project wishes to add a clarification. In the case of Environment and Social Impact Assessment guarantees provided in support of shareholder (ESIA) if warranted by the nature of the proj- loans from a parent bank to a subsidiary, MIGA ect. MIGA will still require its clients to disclose looks to the SEMS that the parent company the project’s social and environmental impacts imposes as a matter of corporate policy on to local communities. itself and its subsidiaries (at local project en- • Capacity of the Environmental and Social terprise level), which the subsidiaries are ex- Unit. The evaluation recommends increasing pected to follow. These are normally part of the the capacity of the Environmental and Social company’s credit policies and standards, and Unit to the level needed to provide credible the parent company provides guidance, train- assurance on performance against the stan- ing, and enforcement of these policies. MIGA dards for every project guaranteed by MIGA, Management plans to start examining how the and should MIGA be unable to increase its client (i.e., corporate parent) implements its resources devoted to implementation of the policies at the local project enterprise level Performance Standards, it should revise its during the guarantee period, but will conduct Policy on Social and Environmental Sustain- this monitoring exercise on a selective basis. ability to disclaim any responsibility for moni- toring the project’s social and environmental Management Observations – World Bank performance. While past performance in this Group area may have left room for improvement, Project Categorization the situation today is different and improving. Management recognizes that two key challenges MIGA has taken a number of important steps exist in establishing consistent approaches to that squarely address this issue, including the categorization between the Bank, IFC, and strengthening of MIGA’s Environmental and MIGA. The first challenge is in the definition: Social Team, which has increased significantly the current definitions of Category A and C in terms of staff size since fiscal 2006. While are relatively clear to most project teams when xxxii MANAGEMENT RESPONSE projects occur at the far “ends” of the bell curve mental and social specialists from the World Bank, that characterizes the distribution of projects at IFC, and MIGA to discuss how their respective various levels of project risks and impacts. The units’ policies on categorization can be improved definition of Category-A has been particularly in practice and what approaches to a shared set useful in focusing attention on the relatively of objective criteria are possible. There is ongoing small number of high-risk projects. The current work in this regard. IFC is proposing changes definition, and interpretation, of Category B to its approach to environmental and social covers a wide spectrum of risk levels, however, categorization as part of the review and update with no clear distinction regarding the location of its Sustainability Framework. The proposed or width of the “threshold” between Category A changes, if endorsed by the Board, will bring and B or Category B and C. A second challenge IFC categorization more in line with the World is in the significant difference in business models Bank approach, while allowing for adaptations with respect to the client’s project cycle: whereas that are important for IFC, such as a three-tiered the Bank frequently engages in a very early stage risk approach for Category FI. IFC will consider of project concept and makes initial determi- these recommendations in its revised Sustain- nations of categorization based on “potential” ability Framework, which will be presented to impacts for a project that is still at a conceptual the Board. IFC Management will provide internal level, the IFC and MIGA most frequently find guidance to staff regarding categorization as themselves becoming involved with a potential part of the update of Environmental and Social client either in a clearly defined project or an Review Procedures, which will be finalized at the existing operation, where risks and impacts may completion of the ongoing process of updating be well defined and mitigation measures already the Sustainability Framework. The recommenda- built into project design or operations. tions of this review will be factored into Manage- ment’s review of global good practice, which will During the first half of fiscal 2011, Management be carried out in preparation of an update of will convene a small group of senior-level environ- Bank safeguards. xxxiii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Management Action Record RECOMMENDATION MANAGEMENT RESPONSE 1. Revise the policy frameworks to harmonize thematic coverage and guidance across the Bank Group and enhance the relevance of those frameworks to client needs IFC, MIGA, and the World Bank should World Bank, IFC, and MIGA: Agreed. Bank, IFC, and MIGA jointly adopt and use a shared set of objective Management will convene within the first half of fiscal 2011 a small criteria to assess social and environmental group of senior-level environmental and social specialists to discuss risks to ensure adequacy and consistency in approaches to either a shared set of objective criteria or alternative project categorization across the WBG, using approaches to categorization that are more refined in scope and the more inclusive criteria for category A, and clearer to teams. The recommendations of this review will be factored refining the categorization system to address into Bank Management’s review of global good practice, which will the bunching of higher- and lower-risk projects be carried out in preparation of an overall update of Bank policies within the current category B. on project safeguards. IFC will consider these recommendations in its revised Sustainability Framework, which will be presented to the Board. IFC Management will provide internal guidance to staff regarding categorization as part of the update of its Environmental and Social Review Procedures, which will be finalized at the completion of the ongoing process of updating IFC’s Sustainability Framework. MIGA will review its Policy on Social and Environmental Sustainability to make necessary changes and bring its categorization more in line with IFC and the Bank, after IFC revises its Sustainability Policy and proposed changes are endorsed by the Board. Timeline: In parallel with the update of Bank safeguards (see below) and following Board approval of the updated IFC Sustainability Policy and Performance Standards. The World Bank should: Disagreed. While Bank Management recognizes, as does IEG, the Ensure adequate coverage of social effects— importance of undertaking a comprehensive updating and consolidation integrating community and gender impacts, of its safeguard policies, it is not yet ready to agree in this detail labor and working conditions, and health, on the final outcome of that process. Instead, taking into account safety, and security issues not currently IEG’s analysis and consideration of IFC’s Performance Standards for covered by its safeguard policies—by its private sector support in the context of the Bank’s public sector consolidating existing social safeguards with support, Bank Management plans to engage in a learning and other World Bank Group policies on social risks consultative process with a diversity of shareholders and stakeholders as requirements under one umbrella policy on on global good practice (in developing countries as well as industrial social sustainability. countries). Bank Management plans to complete this process in the next 24 months and then report to CODE and the Board on how it Consolidate the environmental policies as intends to further strengthen environmental and social sustainability in requirements under one umbrella policy on its projects, including presentation, for their consideration, of a policy environmental sustainability. paper setting out its updated and consolidated approach. Revise the current approach to safeguards A period of 24 months has been proposed to provide adequate time pilots on use of country systems to focus on to undertake an interactive review process, develop a draft umbrella strengthening country institutions and systems safeguard policy, prepare translations, and conduct consultations to manage environmental and social risks. within and outside the WBG. Periodic briefings are planned to be held on a regular basis with representatives of CODE and the Board to keep them apprised of developments and to seek their guidance to the joint team as needed. Timeline: 24 months. During this process, on an interim basis, Bank Management will address concerns related to the balance between environmental and social issues by preparing and issuing guidance on the scope and coverage of social issues in the context of the preparation and implementation of environmental assessments. Timeline: Guidance issued by the end of the third quarter of fiscal 2011. (continued on next page) xxxiv MANAGEMENT RESPONSE Management Action Record RECOMMENDATION MANAGEMENT RESPONSE IFC should: Agreed. IFC has proposed changes through the ongoing review Strengthen the provisions on sustainability and update of its Sustainability Policy and Performance Standards, to address emerging issues, notably climate which will address climate change, supply chains, and biodiversity, change and supply chains and their commodity among others. Management feels that the proposals put forward with certification. regard to climate change—including the consideration of low-carbon technology options and resource efficiency, and to strengthen reporting on greenhouse gas emissions—are consistent with good business, and with good international industry practice. Timeline: Following Board approval of the updated Sustainability Policy and Performance Standards. Proposed changes to the Performance Standards include extending supply chain considerations to significant safety issues. Regarding certification, IFC is generally supportive of these schemes and has an active and ongoing engagement in a number of global commodity roundtables. Timeline: Following Board approval of the updated Sustainability Policy and Performance Standards. Disagreed: Management notes that the number of credible certification schemes is still limited, and that these schemes could disadvantage small-scale producers and suppliers, particularly those in developing countries. In addition, IFC clients have varying degrees of control of or influence over their supply chains. Develop more robust approaches to the Agreed/Ongoing. With regard to FIs, IFC has been implementing implementation of the Performance Standards a number of measures to strengthen its environmental and social in financial intermediary projects, listed risk management approach and implementation of the Performance equities, and trade finance. Standards. A risk-based approach was developed and adopted as the basis for managing risk in FI operations, supported by a global team of staff and consultants that are specialized in this line of business. This approach has led IFC to review select due diligence work undertaken by FIs as a standard approach to supervision. In the proposed revisions to the Sustainability Policy, IFC is further refining its risk-based approach with the introduction of a three-tiered risk categorization for FIs: high, medium, and low, with risk-appropriate due diligence requirements. For higher-risk operations, IFC has established the practice of a right to review before investments are made (Funds), in accordance with the provisions set out in the legal agreement. The approach used by IFC for the existing trade finance programs (e.g., Global Trade Finance Program, Global Trade Liquidity Program) is to apply the exclusion list and undertake regular reviews of this application. IFC will review this approach from a corporate risk perspective and as additional trade finance products are developed and launched. Strengthen policies and practices on disclosure, Agreed. IFC’s current Disclosure Policy is a hybrid approach, specifying including at the local levels. what information it will disclose, subject to a list of exceptions, which is the standard approach among private sector financial institutions. IFC is currently reviewing proposals that will move disclosure to a process that spans the investment lifecycle. IFC is proposing revisions to the Disclosure Policy to provide stakeholders with updated information regarding development impact during project implementation. Timeline: Following Board approval of the updated IFC Disclosure Policy. (continued on next page) xxxv S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Management Action Record RECOMMENDATION MANAGEMENT RESPONSE MIGA should: Agreed/Ongoing (Improvement of Monitoring). While MIGA Increase the capacity of the Environmental Management agrees that past performance in this area may have and Social Unit to the level needed to provide left room for improvement, the situation today is different and credible assurance on performance against still improving. MIGA has taken a number of important steps that the standards for every project. Should MIGA squarely address this issue, including the strengthening of MIGA’s be unable to increase its resources devoted Environmental and Social Team, which has increased significantly to implementation of Performance Standards, in terms of staff size since fiscal 2006, and the introduction of new it should revise its Policy on Social and safeguard (Performance Standards) and disclosure policies. While Environmental Sustainability to disclaim any keeping its current risk-based approach, MIGA Management will responsibility for monitoring the projects’ review whether more resources should be allocated to monitoring, and social and environmental performance and recommends continuing with its risk-based approach, but increasing ensuring that they comply with the standards. the number of site visits for those projects where issues are most Under this option, MIGA’s role would be limited likely, and regular monitoring reports are received from investors, while to reviewing the client’s assessment of the maintaining routine contact with those projects where the risks are project’s environmental and social risks against viewed to be less. MIGA’s Environmental and Social Team has begun the standards, identifying corrective actions as development of a monitoring strategy to this end. needed, and securing the client’s commitment Agreed/Completed (Increasing Department Capacity). MIGA’s to implement these actions. Environmental and Social Team has increased significantly since the early years of the review period. Over the last two fiscal years MIGA Management has doubled this department, bringing it to its current size of six staff members. The department is now operating with sufficient capacity. Disagreed (Disclaiming Responsibility for Monitoring). The IEG alternative of MIGA declaring it will not monitor projects’ environmental and social performance is not consistent with MIGA’s mandate as a development institution. Timeline: MIGA’s Environmental and Social Team has started developing a monitoring strategy, and plans to start implementation in the second quarter of fiscal 2011, after the program has been reviewed by the MIGA Senior Management Team. Require that category-B Small Investment Disagreed. At this stage this recommendation is too early to be Program projects follow the same disclosure acted on. MIGA Management will review its Disclosure Policy after requirements as for regular category-B projects. IFC completes its own review. MIGA will still post its SPG for all SIP projects, which may include a more detailed explanation of environmental and social issues, and will attach the project ESIA if warranted by the nature of the project. It should be noted that SIPs are subject to all requirements of PS1, which require MIGA clients to undertake local disclosure (including any ESIA) and community consultations as warranted by the nature of the project. Timeline: More detailed environmental and social information in SPGs of SIPs and ESIA disclosure (if warranted): Work is ongoing. 2. Enhance client capacity, responsibility, and ownership The World Bank should: Agreed. Bank Management agrees and will work among SDN, OPCS, Increase the synergies between safeguards LEG, and the Regions to promote this approach. This issue will also be work and broader Bank engagement on an element of the global good practice review discussed above. For environmental and social sustainability by example, as part of the updated Environment Strategy process, SDN is investing in upstream analytical work, technical developing guidelines on how to incentivize analytical work, technical assistance, and lending to strengthen country assistance, and lending that strengthens environmental governance, and sector institutions and capacities in client institutions, and capacity in client countries. countries. (continued on next page) xxxvi MANAGEMENT RESPONSE Management Action Record RECOMMENDATION MANAGEMENT RESPONSE While Bank Management agrees, it suggests that it not be included in future Management Action Records for monitoring because there is no clear way of demonstrating its implementation. Require regular reporting by the borrower on Not Agreed. Instead, this issue will be included in the process implementation and outcomes of safeguards in outlined above in response to Recommendation 1. Bank-supported projects, and work with clients to develop instruments and indicators to help in such monitoring. IFC should: Agreed/Ongoing. With regard to staff incentives, these are Develop incentives for investment officers to constantly evolving and incentives related to environmental and social share ownership of the Performance Standards issues have become more prominent in recent times, especially in and mainstream their implementation. areas where environmental and social performance is a core aspect of project sustainability. IFC Management has reinforced environmental and social issues as a shared and core agenda and will hold staff accountable for this. Environmental and social due diligence is required for success in approval of investments. IFC will consider opportunities to include environmental and social aspects in the performance management process. Use advisory services to build social and Agreed. With regard to advisory services, IFC will work to environmental management systems and strengthen the capacity of select clients to develop and manage their implementation capacity, especially among environmental and social management systems through a mix of tools small and medium enterprises, financial and approaches, including the selective and strategic use of advisory intermediaries, and clients in countries and services. IFC will also use other approaches, as appropriate, including sectors with weak environmental and social the use of environmental and social specialists who engage with management. clients in developing action plans and use supervision to verify and support implementation of environmental and social standards; and country-based expertise, especially in middle-income countries. Timeline: Work with advisory services has been initiated and a strategic approach is expected before the end of fiscal 2011. Mobilize resources at appraisal for energy and Agreed/Ongoing (energy and cleaner production audits). IFC clean production audits, using auditors with includes energy/cleaner production audits as part of appraisal or relevant sector knowledge. ongoing improvement of clients’ operations when deemed useful and appropriate. Disagreed (resource mobilization approach). IFC disagrees with the proposed resource mobilization approach. Resources are mobilized through different avenues not linked to timing of appraisal and may include funding of audits directly by clients. Define areas of influence and requirements Agreed. IFC requirements on supply chains apply to all sectors with to better address supply chain risks and a focus on the highest risks, such as child labor, forced labor, and opportunities, particularly related to clearing of critical habitats. The proposed changes to the Performance biodiversity and forestry, expanding the Standards expand this to include significant safety issues. IFC has application of material biodiversity along the included provisions under PS1, PS2, and PS6 to ensure an adequate supply chain for suppliers. assessment of supply chains is undertaken as part of appraisal. Timeline: Following Board approval of the updated Sustainability Policy and Performance Standards. (continued on next page) xxxvii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Management Action Record RECOMMENDATION MANAGEMENT RESPONSE MIGA should: Agreed. MIGA Management wishes to clarify that in the case of Focus the due diligence reviews of financial guarantees provided in support of shareholder loans from a parent bank sector projects on the Social and Environmental to a subsidiary, it looks to the SEMS that the parent company imposes Management Systems of developing-country as a matter of corporate policy on itself and its subsidiaries (at local subsidiaries the project supports, rather than project enterprise level), which the subsidiaries are expected to follow. the corporate policies of the parent banks. MIGA Management plans to start examining how the client (i.e., corporate parent) implements its policies at the local project enterprise level during the guarantee period, but will conduct this monitoring exercise on a selective basis. Timeline: Will start in second quarter of fiscal 2011. Expand the size and eligibility of the Trust Agreed/Ongoing. MIGA Management has started taking steps Fund for Addressing Environmental and Social to extend the Trust Fund, including initiating contact with several Challenges to all low-capacity clients on the potential donors. The first three pilot years showed that the Trust fund basis of need. is a very valuable tool and resource for MIGA. MIGA Management sees considerable value in maintaining and extending this facility, but at the same time notes that the ability to comply with this recommendation will largely be driven by the willingness and capacity of external donors to be involved. 3. Revise guidelines, instruments, and incentives to strengthen supervision arrangements The World Bank should: Disagreed. Bank Management agrees with IEG that there is a need to Assign responsibility and budget for safeguards strengthen supervision of medium- and low-risk projects. How this will oversight and reporting to environmental and be done may need to differ from region to region, depending on country social units in each operational Region—in line capacity and project type and mix. with IFC practice—in place of the delegation Bank Management does not agree with the specific recommendation of safeguards processing and supervision to on giving the responsibility and budget for safeguard oversight and sector management units. reporting to environmental and social units in each operational Region and this will need to be dropped from further monitoring by IEG. Bank Management plans to undertake a review by the second quarter of fiscal 2011 concerning current practices with respect to responsibility, accountability, incentives, staffing, and budgeting for safeguard processing and supervision. This review will also cover the issue of financial intermediary projects and projects that use environmental and social policy frameworks (see below). Based on this review, practices will be updated with the objective of enhancing effectiveness and efficiency and maximizing the synergies between safeguard work and broader Bank engagement on environmental and social sustainability. Timeline: Bank Management action, based on the review, by the third quarter of fiscal 2011. Bank Management notes that, as part of Investment Lending Reform process, it has actions ongoing to enhance the effectiveness and efficiency of implementation support. These include: (a) the assignment of staff and budget in line with the level of risk associated with an operation, using the new risk assessment and management procedures; and (b) the embedding of grievance redress mechanisms more broadly into projects. (continued on next page) xxxviii MANAGEMENT RESPONSE Management Action Record RECOMMENDATION MANAGEMENT RESPONSE Introduce a certification program to expand Agreed/Ongoing. OPCS is developing, in coordination with SDN and the pool of staff qualified to undertake social LEG, a mandatory Operational Core Course for task team leaders which and environmental preparation and supervision includes modules on safeguard policies and their implementation. while ensuring quality and consistency, and Bank Management also has several ongoing and planned initiatives to provide orientation training on environmental expand the pool of qualified environmental and social staff that can and social sustainability to all task team provide support on safeguards and sustainability issues. leaders. Bank Management supports the initiation of a certification/ accreditation program for environmental and social staff working on sustainability and safeguard issues starting in fiscal 2011. SDN is working on the design of a core environmental and social sustainability and safeguards course, which will act as a mentoring and certification/ accreditation program for environmental and social staff, selected staff of other sectors, and safeguard consultants. The certification/ accreditation program will commence by the end of fiscal 2011. SDN also has launched several complementary initiatives to improve the staffing and skills mix for sustainability and safeguards, and to align incentives with the mainstreaming of environmental and social sustainability throughout the portfolio. These include: (a) a Bank-wide analysis of staffing for environmental and social sustainability and safeguards; (b) the development of competencies that emphasize skills in sustainability and safeguards, on both the environment and social issues; (c) consistent management signaling regarding the importance of working on sustainability and safeguards; and (d) the organization of field-based training sessions on sustainability and safeguards. Timeline: Processes in place (subject to cost considerations) by the beginning of fiscal 2012. Develop and implement an action plan Agreed/Ongoing. Bank Management is currently engaged in a Bank- to ensure regular supervision of financial wide review of the use of frameworks that will examine these types intermediary projects and investment projects of projects and identify good practices. The review will include an that use social and environmental policy examination of a variety of means to strengthen monitoring of such frameworks through third-party or community projects, including, in appropriate situations, the use of third-party or monitoring for higher-risk projects, and community monitoring for selected higher-risk projects. The review is disclosure of monitoring and supervision expected to be completed by the third quarter of fiscal 2011 and will reports. provide the basis for guidance to be issued for use by Bank staff and borrowers by the fourth quarter of fiscal 2011. Disagreed. See above on supervision. To be clear, Bank Management does not agree and will not be held accountable in future Management Action Records for asking clients to implement third-party or community monitoring. Timeline: Action completed by the end of fiscal 2011. IFC should: Agreed/Ongoing. IFC has been strengthening the oversight of its Enhance the supervision of financial investments in and through financial intermediaries at the portfolio, intermediaries at the subproject level by company, and subproject levels. This approach has been developed developing clear guidelines for applying the and is being implemented in accordance with a risk-based approach, Performance Standards at the subproject level which is intended to deploy resources efficiently where the risk is and by adopting a systematic approach to highest and/or performance is poorest. IFC provides guidance on the environmental and social specialists’ site visits application of Performance Standards at the subproject level through to selected subprojects. its ongoing engagement with clients. (continued on next page) xxxix S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Management Action Record RECOMMENDATION MANAGEMENT RESPONSE There is an ongoing multiyear effort to ensure that IFC’s approach continues to be suitable. The approach to the subproject level supervision will be codified in IFC’s Environmental and Social Review Procedure. Use loan covenants, including Conditions of Agreed/Ongoing. The use of loan covenants to support compliance Disbursement to enforce compliance with with environmental and social requirements is a standing practice. IFC environmental and social requirements and has several instruments to support client compliance. These include reporting if the clients lack commitment and specific provisions in the loan agreement or legal documentation are continuously out of compliance. and action items linked to disbursement and specific deadlines. There are covenants in legal documentation through which IFC monitors compliance, including a “policy put” in some cases, whereby noncompliance of policy provisions would trigger the option for IFC to sell its shares. 4. Strengthen safeguards monitoring, evaluation, and completion reporting The World Bank should: Partially Agreed/Ongoing. Bank Management agrees with the need Include performance indicators on to strengthen monitoring and evaluation arrangements. To address environmental and social outcomes in project this issue, the Bank will collaborate on developing guidelines on results frameworks and ensure systematic monitoring and evaluating safeguard performance by the third quarter collection of data to monitor and evaluate of fiscal 2011. These guidelines will focus on more systematically safeguards performance. measuring outcomes, including through the use of core environmental and social performance monitoring indicators, and on evaluating impacts. Reporting will be integrated in the ISR, building on the new risk framework under the Investment Lending Reform, which is already enhancing the monitoring of environmental and social risk mitigation measures in ISRs. The proposed guidelines on monitoring and evaluation will further emphasize the need for the ICR to evaluate the achievement of the safeguard-related objectives and identify lessons for future projects. Timeline: Guidelines issued by the end of the third quarter of fiscal 2011. Ensure that Implementation Completion Not Agreed. Bank Management does not agree and will not be held Reports and IEG reviews of those reports rate accountable in future Management Action Records for asking clients to and report effectively on the outcomes of use performance indicators on environmental and social outcomes in safeguards and, for all projects with significant all project results frameworks. environmental and social effects, ensure the results are incorporated as an essential dimension when assessing achievement of the project’s development objective, as has already been done for IFC and MIGA. IFC should: Agreed (disclosure of some project-level information). IFC Disclose project-level environmental and social is reviewing its Disclosure Policy to determine where it is most information from monitoring and supervision appropriate to make modifications to the policy and to practices reports. throughout the project life cycle. Disagreed (disclosure of all information). However, this will not include disclosure of all environmental and social reporting from clients since there has to be a balance between client confidentiality and disclosure. IFC will continue its practice of holding its clients responsible for reporting to the local community. Timeline: Following Board approval of the updated Disclosure Policy. (continued on next page) xl MANAGEMENT RESPONSE Management Action Record RECOMMENDATION MANAGEMENT RESPONSE Make use of independent/third-party or Agreed (community and select independent/third-party community monitoring and evaluation for monitoring and evaluation). IFC Management will explore how to its projects, particularly for projects with strengthen community engagement, participatory monitoring and how, involuntary resettlement and higher-risk in selected high-risk cases, third party monitoring or advice can be financial intermediary and agribusiness incorporated. projects. Timeline: Following Board approval of the updated Sustainability Policy and Performance Standard. Disagreed (independent/third party monitoring across the board). IFC Management does not see third-party monitoring as an approach that should be appropriate for all projects, but rather one that may be considered in selected higher-risk situations. Since a key priority is to strengthen client capacity and ownership, third- party monitoring should be seen in that context, not just in terms of independent verification. MIGA should: Disagreed. At this stage this recommendation is too early to be Disclose project-level environmental and social acted on. MIGA Management will review its Disclosure Policy after information from supervision reports. IFC completes its own review and any modifications are endorsed by its Board. MIGA will continue its practice of it being the client’s responsibility to report to the local community. Timeline: fiscal 2012, following Board approval of IFC’s disclosure policy. Develop a credible mechanism to ensure that Agreed/Ongoing (Project Enterprise SEMS). The term “credible Performance Standards are adhered to by mechanism” is not entirely clear to MIGA Management. If this means financial sector projects. that MIGA should ensure that for financial sector projects the project enterprise has a Social and Environmental Management System (SEMS) consistent with MIGA’s Policy and Performance Standards, then MIGA Management agrees, notes that this is ongoing, and wishes to echo the clarification made above. In the case of guarantees provided in support of shareholder loans from a parent bank to a subsidiary, MIGA looks to the SEMS that the parent company imposes as a matter of corporate policy on itself and its subsidiaries (at the local project enterprise level), which the subsidiaries are expected to follow. MIGA Management plans to start examining how the client (i.e., corporate parent) implements its policies at the local project enterprise level during the guarantee period, but will conduct this monitoring exercise on a selective basis. Disagreed (Third Party Monitoring). If “credible mechanism” means third-party monitoring, as with IFC, then MIGA Management disagrees that third-party monitoring is needed or cost effective for all projects. MIGA Management suggests that this be prioritized to focus on high-risk situations. It should be noted that it is MIGA’s current practice to require independent (third-party) assessments as warranted by the nature of the project. (continued on next page) xli S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Management Action Record RECOMMENDATION MANAGEMENT RESPONSE 5. Improve systems and instruments for accountability and grievance redress IFC, MIGA, and the World Bank should: IFC: IFC management recognizes that as a private sector development Seek greater symmetry in the structure of Bank institution it is key to have an independent office with effective Group accountability and grievance redress compliance and mediation functions. IFC Management notes that mechanisms. For the World Bank this would the recommendation with respect to the CAO will not require any entail creation of a grievance redress and management action. At the same time, IFC Management has not conflict resolution mechanism to complement noted any concerns with the effectiveness of the current grievance the Inspection Panel. For IFC and MIGA this and mediation function. IFC Management is awaiting with interest the would entail a more independent compliance outcome of the Board’s ongoing reviews of oversight and accountability review process, ensuring that the CAO submits mechanisms. its audits directly to the Board. MIGA: Like IFC, MIGA Management recognizes that having an independent office with effective compliance and mediation functions is key to a private sector development institution. MIGA Management notes that the recommendation with respect to the CAO will not require any management action. At the same time, MIGA Management has not noted any concerns with the effectiveness of the current grievance and mediation function. MIGA Management is awaiting with interest the outcome of the Board’s ongoing reviews of oversight and accountability mechanisms. World Bank: Agreed/Ongoing. Bank Management agrees with IEG that there is value in creating a grievance redress mechanism for which Bank Management will take responsibility that is complementary to, but separate from, the Inspection Panel. Bank Management wishes to underscore that establishing this mechanism would not alter the responsibility of borrowers and recipients for implementing projects, and that in many cases, the grievances are not necessarily with the Bank, but between our clients and project-related stakeholders. Nevertheless, these grievances are often brought for resolution to the Bank. Therefore, by the end of the third quarter of fiscal 2011, Bank Management intends to complete a survey and review of a wide range of potentially analogous existing grievance redress mechanisms as a basis for designing one for the Bank. The study will include a review of the cost implications and potential cost savings that could be engendered by using a system similar to the IFC CAO or other multilateral financial institutions. Bank Management will present the results of this study to the Board to ensure that any decisions emerging from the study will be consistent with the Board Resolution and related Board decisions concerning the Inspection Panel, and in a manner which takes fully into account the current requirements and experiences with project-based grievance mechanisms (including as required under OP 4.12, Involuntary Resettlement, and OP 4.10, Indigenous Peoples). This study will be coordinated among Bank units with considerable experience in this field to ensure institutional coherence and efficiency. Timeline: Bank Management will (subject to cost considerations) establish a grievance mechanism by the first quarter of fiscal 2012, and provide to the Board a detailed report on the initial operation of the grievance mechanism by the end of fiscal 2012. xlii Chairperson’s Summary: Committee on Development Effectiveness (CODE) O n July 28, 2010, the Committee on Development Effectiveness (CODE) considered the report Safeguards and Sustainability Poli- cies in a Changing World: An Independent Evaluation of World Bank Group Experience, prepared by the Independent Evaluation Group (IEG), and the Draft Management Response. The Committee commended IEG for its first and clients. The need to give Bank Management comprehensive evaluation of the full set of sufficient time for the review of global good safeguards and sustainability policies used by practice to update the Bank’s safeguard policies the World Bank Group (WBG). It welcomed the was raised, although some speakers suggested timely discussion as the WBG is completing its shortening the proposed timeline. Members new Environment Strategy; IFC is reviewing and emphasized the importance of engaging client updating its sustainability framework; and the countries in the review of E&S safeguards. Bank is working on the Investment Lending (IL) Reform. The Committee noted one of IEG’s main Members agreed with IEG on the importance findings regarding the benefits of environmental of effective implementation of safeguard and social (E&S) safeguards and performance policies and strengthened supervision; and on standards for sustainable development and the the checks and balances provided by monitor- overall positive impact in client countries and for ing and evaluation, disclosure of findings, and the private sector, although challenges remain verification of results. They also concurred in effective supervision and monitoring of E&S on the need to strengthen client capacity and outcomes. enhance responsibility and ownership. Members commented on the challenges of expanding the Members stressed the importance of carrying out a Bank’s pilot use of country systems, address- comprehensive update of Bank safeguard policies, ing safeguard concerns at the subproject level, and harmonizing the categorization of projects particularly at the IFC, and budget and incentives across the WBG. Some members remarked that issues. Greater symmetry in the structure of a common policy framework could promote a WBG accountability was encouraged, and many better understanding by external stakeholders members agreed with Bank Management on of the WBG approach, although others observed the importance of considering the creation the need for some differentiation by each WBG of a grievance redress and conflict resolution institution given their specific business nature mechanism. xliii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Recommendations and Next Steps across the WBG, and noting the need to Management will revise its draft response, taking balance safeguards (do not harm) and perfor- into account the main issues raised at the meeting, mance standard management (management including to further clarify sections in which it of risks), cautioned against unified safeguard expressed “partial” agreement/ disagreement standards. They noted the need for differenti- with IEG. The Committee recommended a full ated approaches for the specific nature of each Board discussion of the IEG report and a revised WBG institution. Other interventions focused on Management Response given the relevance of the need to prioritize clients’ capacity building, safeguards and performance standards for the distinguishing between countries and private WBG activities. clients, and considering the different stages of development. The Bank, IFC, and MIGA Management will convene within the first half of FY11 a small Speakers raised questions about the impact of group of senior-level environmental and social implementing IEG’s recommendations on the development specialists to discuss approaches to cost of doing business for clients and for the adopt and use a shared set of objective criteria to WBG; how to prioritize these recommenda- assess E&S impacts and risks to ensure adequacy tions; and the main purpose of the consultation and consistency in project categorization across process planned by the Bank team led by OPCS. the WBG. A recommendation was made to include field- based managers in the proposed Bank team. Bank Management committed to a review of A member underlined that the WBG is one of global good practice that will integrate the update many development players, and that it should of safeguard policies. A team led by OPCS, with consider the following: (i) be invited to support participation by SDN and Legal, will be formed and do business in a developing or transition to engage in a learning and consultative process country and provide support that adds value and with diverse shareholders and stakeholders at the is consistent with the country’s strategy; (ii) staff national and subnational level during the next 24 should be encouraged to deal with risks and with months. At the conclusion of this process, Bank complex projects; and (iii) “perfect” safeguards Management will report to CODE/Board on how and performance standards are difficult to apply. the Bank intends to strengthen E&S sustainabil- Another member welcomed IEG cost-benefit ity in projects, including the possibility of a more analysis of safeguards. A few members encour- consolidated policy framework. aged Management to enhance communica- tion, including on the benefits of managing and IFC Management will take into account IEG mitigating impacts and risks in spite of the initial recommendations and CODE comments in the costly and time-consuming steps. Responding to ongoing review of IFC’s Sustainability Framework. a few speakers’ interest in a review of safeguards It also indicated that Phase 2 of the consultation for development policy lending (DPL), Bank process on the proposed draft changes to the Management noted the different nature of DPLs Sustainability Framework will be extended to 90 and that these requirements and practices will be days (initially proposed for 60 days at the CODE reviewed in the context of DPL retrospective. meeting on May 5, 2010). Environmental and Social Safeguards Main Issues Discussed Speakers noted IEG’s recommendation that a Updating the Safeguard Policies more comprehensive and balanced approach Some members encouraged Bank Management to social issues would be useful in supporting to review the current policies in the context of the Bank’s broad development objectives and the ongoing work on IL Reform. A few members, bringing greater consistency between the Bank while welcoming efforts to consolidate policy and IFC. A member preferred having separate frameworks and harmonize thematic coverage E&S “umbrella policies” to give them equal xliv C H A I R P E R S O N ’ S S U M M A R Y: C O M M I T T E E O N D E V E L O P M E N T E F F E C T I V E N E S S ( C O D E ) visibility, while two speakers cautioned against a performance. A member, however, cautioned stark division between environmental and social that the Bank should not request additional safeguards because they are interrelated in many reports that may not be in line with what the instances. Some members suggested that the country needs or what is being prepared for WBG should consider an integrated approach to internal reporting. social impacts and risks that takes into account specific national realities. As noted in its response, Bank Management committed to (i) review in Q2:FY11 on current Human Rights practices with respect to responsibility, account- A few members felt the WBG institutions must ability, incentives, staffing and budgeting for avoid adverse human rights impacts and ensure safeguard processing and supervision, which that a project does not infringe on government’s will serve to update the current practices and obligations under international and national enhance effectiveness; (ii) review by Q3:FY11, human rights law. For this purpose, they stressed ways to strengthen monitoring of projects the need to identify and fill the existing gaps through financial intermediaries lending and within the WBG, which may differ for public and projects that use a programmatic approach and private sector projects. Two members expressed support the use of subprojects; and (iii) develop the view that there should be a clear distinction guidelines for M&E safeguard performance by between the Bank’s involvement, limited to its Q3:FY11. project interventions and associated fiduciary responsibilities, and the country’s institutions Use of Country Systems (UCS) including domestic regulations and the judicial Members noted IEG’s findings that the current system. They also noted the need for the Bank to approach to pilot UCS has been rigid, and asked keep to its mandate. Bank Management to further explore the reason for limited progress, and whether client’s capacity Implementation and Supervision has been inadequate. In this context, some Speakers supported further investment to speakers emphasized support to strengthen strengthen E&S management systems including country institutions and systems, and integrate supervision, performance indicators, and data technical assistance to lending instruments. collection; and encouraged the Bank to include Bank Management indicated that if a change in E&S outcomes in the Implementation Comple- approach is to occur, a more flexible application tion Report. Moreover, a member noted that of the existing OP 4.00 will be needed. A few strengthening and using country institutions speakers cautioned that country standards and to monitor implementation of performance capacity needs should precede the UCS, and standards should be linked to the use of country necessary changes should happen not only at systems initiative. the project level but also in the broader system. Noting the Bank Management’s disagreement Grievance Mechanism with IEG’s recommendation to shift responsibil- Some members endorsed Bank Manage- ity and budget to monitor the implementation of ment’s proposal to complete a survey and E&S safeguards from Sector Management Units review grievance mechanisms in IFC and other to Environment and Social Units, a member international institutions by end-FY11, and to asked about this different approach from that establish a mechanism by FY12. A few speakers for procurement which is the responsibility of recommended that Bank Management consult Procurement Managers. Another member felt the Inspection Panel, given its broad experience Bank Management should be more flexible and with safeguard issues. Bank Management agreed suggested the shift may be consistent with the IL with IEG on the value of establishing a grievance Reform. Others encouraged the Bank to require mechanism that complements, but is separate systematic reporting by its clients on their E&S from the Inspection Panel. A few members xlv S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D proposed that the issue of grievance mechanism made on the need for IFC to review supply chain be addressed in the context of the “5 Is” discus- issues, and the challenges of doing business sion by COGAM. Two speakers favored expanding through financial intermediaries lending, and the mandate of the Inspection Panel, although a listed equities. A member felt that IEG could member felt that this was unnecessary given that have compared IFC performance standards with the grievance mechanism would formalize what those applied by other MDB institutions focused already exists. on private sector. Regarding MIGA, a member felt it should not revise its policy to disclaim IFC and MIGA Sustainability Policy responsibility for monitoring a project’s E&S Some speakers agreed with IEG that IFC and performance, and it should increase its capacity MIGA should adopt third-party verification to ensure compliance with the Performance more broadly in its oversight practice. They also Standards. encouraged the full and timely disclosure of monitoring reports on E&S performance, as well as more effective disclosure of project sponsors to local stakeholders. Some comments were Carolina Renteria, Acting Chairperson xlvi Statement of the External Advisory Panel T he external Advisory Panel welcomes this report on the World Bank Group’s (WBG) safeguard and sustainability policies. We concur with the findings of the evaluation and strongly endorse the five recom- mendations presented in the report. The Advisory Panel provided IEG’s evaluation team with preliminary comments based on a reading of an earlier draft of this report in Washington, DC, on April 22–23, 2010. Noting that most of that advice has been incorporated into this final version of the report, this final Panel Statement is brief. After more than 10 years of operation of It is the Advisory Panel’s view that while the the WBG’s safeguard policies, and with the Bank’s safeguard policies have been an appropri- subsequent introduction of the Policy and Perfor- ate mechanism “to prevent or mitigate adverse mance Standards in IFC and MIGA, it clearly was impacts of its projects on people and the environ- time for a thorough review of their effectiveness ment” in the past, an approach based solely on and ongoing appropriateness. We note that the “do no harm” is no longer good enough. The report finds that the safeguards and sustain- world has changed in the corresponding time ability policies have helped to avoid or mitigate period. Local communities and international large-scale social and environmental risks in stakeholders now expect more from develop- WBG-financed projects. More importantly, ment projects. Communities expect positive however, the report identifies a number of issues benefits to flow from projects and they expect relating to the implementation of the safeguard opportunities to be provided to them within policies that potentially has reduced their projects to ensure that they are beneficiaries. effectiveness, especially in terms of contributing Around the world, a Corporate Social Respon- to development. We note the concerns about the sibility culture has been developing partly due appropriateness of the Bank’s compliance-based to consumer and community pressures, as well focus as its portfolio moves away from traditional as the activities of the World Business Council investment lending, toward a much wider array for Sustainable Development and industry-led of lending instruments. We note the need for initiatives. The use of continuous improvement greater emphasis on developing client ownership standards like ISO (International Standards and country systems. While the report indicates Organization) 14001 for Environmental Manage- that it is too early to compare the results of ment Systems is increasing in the commer- IFC’s Performance Standards against the Bank’s cial sector, in parallel with the use of Strategic safeguard policies, the Advisory Panel has no Environmental Assessment at government and doubt that there is much merit in this approach sector levels. Over 5,000 businesses have signed taken by IFC and MIGA. We would encourage the up to the UN Global Compact and many others adoption of the latter approach throughout the are signing up to similar industry-level initiatives. whole WBG. While the existing safeguards approach has also xlvii S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D helped to encourage such developments, the The Advisory Panel feels that a comprehen- WBG now needs to move forward, to encourage sive and well-balanced Performance Standards further movement toward social and environ- approach, as implemented by the IFC and mental sustainability. MIGA, has considerable merit and is superior to the current safeguards approach. We therefore One of the most obvious things wrong with the recommend that there be a monitoring of the current safeguards approach in the Bank is the results of the Performance Standards over time. lack of consideration of the full range of social The Panel notes the advantages of the Perfor- issues. There is a wide range of social issues that mance Standards in giving attention to the full should be considered, many of which are not range of social issues. We therefore recommend, adequately addressed in the Bank’s safeguard as is implicit in the report’s first recommenda- policies. Based on our personal knowledge of tion, that there be a harmonization of thematic the Bank as well as our review of the evaluation coverage across the WBG and a major revision of report, the Panel considers that the operation- the way the social and environmental issues are alization of the safeguard policies in the Bank, addressed, especially in the World Bank. While especially the associated compliance culture that we endorse the recommendations as presented has developed, is detrimental to the achievement in the last chapter of the report, we would like to of key development objectives. give particular emphasis to some aspects of them. It could be argued that the safeguards (do no • We consider that expansion of the scope of harm) approach is basically focused on protect- social policies to harmonize thematic coverage ing the reputation of the Bank. The Advisory across the WBG is essential. Panel suggests that it is time to change the • We believe that consolidation of the social emphasis to one based on risk management and and environmental policies at the World Bank sustainability. The Fourth Assessment Report of under one overarching and internally con- the Intergovernmental Panel on Climate Change sistent policy is likely to have considerable strongly made this point in 2007. It stressed that benefit providing a better balance between the most effective way to address the climate issue environmental and social components, and is to integrate climate change policies into an we urge the Bank to give due consideration overall, proactive sustainable development, risk to this suggestion, perhaps by undertaking management strategy, rather than implementing a feasibility study or options analysis. In any piecemeal (safeguard-type) approaches—that case, some strengthening of effort in relation is, isolated, reactive adaptation and mitiga- to social issues is absolutely necessary. We note tion measures. As noted in IEG’s report, one of that given many governments are weaker on the consequences of the safeguards approach addressing social risks, this will have resource is a heavy investment up front in ensuring implications for the Bank in terms of building sign-off of projects, but this is at the expense human capacity among clients to implement of ongoing supervision and adaptive manage- the new safeguard policies. ment. The changing blend in the categorization • We argue that the integration of environmental of Bank-funded projects (screening) is evidence and social dimensions is important to a proj- of the way project staff respond to the operation ect’s development outcome, and this should of the safeguard policies. A shift in focus away be built into project reporting, as it is in IFC from a legalistic safeguards-based approach to a within their Development Outcome Tracking process that relies on ongoing risk management System. is likely to enhance development outcomes. • We believe that increased transparency and a Potentially the categorization of projects should greater use of independent, third-party moni- depend not only on environmental risks, but also toring and/or community monitoring and eval- on a wider range of social issues, as well as the uation (as appropriate) would assist in shifting local capacity to address those issues. the focus from compliance to outcomes. xlviii S TAT E M E N T O F T H E E X T E R N A L A D V I S O R Y PA N E L As mentioned in chapter 5 of the report, the Bank the bar around the world. The Bank must now was innovative and a leader in the 1990s when correspondingly update its approach, and this the safeguard policies were being formulated. IEG report is an important first step in that However, the world has changed since then and process. We look forward to seeing action by the expectations are now higher, partly because the Bank in response to the recommendations in the Bank has succeeded in contributing to raising report. Luiz Gabriel Todt de Azevedo, Sustainability Director, Odebrecht Energy, Brazil Mohan Munasinghe, Chairman of the Munasinghe Institute for Development, Sri Lanka, and Professor of Sustainable Development, University of Manchester, United Kingdom Frank Vanclay, Professor in the Department of Cultural Geography, University of Groningen, The Netherlands; former Leader of the Rural Social Research Group, University of Tasmania, Australia xlix Chapter 1 Evaluation Essentials • This is the first comprehensive evaluation of the WBG’s safeguard policies since they were formulated in 1989. • The evaluation examines how ef- fective the WBG’s safeguards and sustainability frameworks have been in preventing and mitigating adverse environmental and social impacts. • It covers WBG safeguard policies and IFC and MIGA Performance Standards in projects approved from fiscal year 1999 to 2008. A community consultation in Paraguay. Photo courtesy of Reidar Kvam. Evaluation Context Introduction countries with growing institutional capacity, Environmental and social safeguard policies and to fragile and conflict states. In reform-minded Performance Standards are a cornerstone of the middle-income and low-income countries, the World Bank Group’s (WBG)1 support for sustain- nature of Bank lending has evolved from invest- able development and poverty reduction.2 The ment projects dominated by infrastructure objectives of these policies, to which it has and agriculture toward a growing portfolio of committed and is publicly accountable, are to development policy loans (DPLs) for institutional improve the quality of investments and guaran- and policy reforms, and programmatic lending tee operations and to prevent or mitigate undue for social sector, financial sector, and governance harm to people and the environment in the operations. DPLs are governed by a different set development process. Over time the focus has of environmental and social requirements from shifted from mandatory compliance with do-no- those of the safeguard policies. Safeguard policies harm policies and procedures toward doing apply to all investment projects but are more good through greater focus on sustainability difficult to implement in sectorwide investment and the management of associated risks. Similar programs, financial intermediary (FI) projects, policies are now widely used internationally as a community-driven development projects, and fundamental aspect of sound business manage- other forms of decentralized projects. Traditional ment practice and development effectiveness.3 investment lending is not well suited to these They have been adopted in various forms by portfolio trends. The Bank is respond- most major financial institutions lending to the ing to this changing context by reform- The context in which public and private sectors. ing its investment lending policies the WBG operates has and instruments promoting use of changed significantly The context in which the WBG operates has risk-based approaches and placing since the introduction of changed in many ways since the introduction greater emphasis on implementa- the safeguard policies, of the safeguard policies, particularly in the tion support. Safeguard policies will particularly in the nature of WBG clients and in the nature of the consequently require significant nature of WBG clients lending portfolio. WBG clients have diversified adaptation to ensure their continued and in the nature of with greater differentiation among countries relevance. the lending portfolio. and the growing significance of private sector and subnational clients. The World Bank’s public The International Finance Corporation (IFC) sector clients now range from middle-income and the Multilateral Investment Guarantee countries, many with well-developed regulations Agency (MIGA) support private sector clients and institutions, to rapidly reforming low-income whose role in development continues to grow 3 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D and whose portfolio continues to diversify. WBG assistance for the environment over 15 IFC and MIGA recognized the need to better years, but it was not intended to evaluate the distinguish clients’ responsibilities from their Bank’s safeguard polices. For IFC and MIGA, the own, and manage private sector environmen- environmental sustainability review did consider tal and social aspects that were not covered in early results of the environmental Performance the safeguards framework, by transforming the Standards, but it was conducted a year after they WBG’s safeguards policies into a new policy were introduced and, in any case, did not cover framework with Performance Standards for their the social Standards. The main purpose of the clients. This transformation shifted the emphasis current evaluation, therefore, is to address this from prescriptive procedures to a more explicit gap taking into account the rapidly changing focus on the client’s social and environmental business environment, new lending modalities management systems (SEMS). Further evolution and financing instruments, as well as evolving of the two agencies’ portfolios in recent years best practices and client needs.7 continues to pose challenges: IFC’s business has evolved from project finance toward corporate Discussions between IEG and WBG operational finance, trade finance, and equity investments, staff, including the World Bank’s Sustainable and MIGA’s portfolio has seen a substantial Development Network Council, at the concept increase in guarantees for the financial sector. stage of this evaluation revealed an interest in examining whether the current Operational Before considering the evaluation findings it Policies remain fully relevant to today’s issues is essential to understand the context within and challenges, given that client interests and which the WBG’s safeguard and sustainability capacities as well as the lending portfolio have policies are operating. This chapter describes altered substantially from the time when these the rationale, approach, and methodology of the policies were first developed. evaluation and then presents three aspects of the context: (i) the WBG’s safeguard policies and The Bank recently initiated a process to reform the newer Policy and Performance Standards on investment lending.8 The current model uses Social and Environmental Sustainability (PPSSES) the project cycle concept in which technical adopted more recently by IFC and MIGA; (ii) the and financial viability and feasibility of detailed relevance of these policies to the previous portfo- engineering plans developed during prepara- lio; and (iii) the evolution of the lending portfolio tion are carefully assessed during appraisal, and to understand the emerging challenges faced by supervision monitors performance against the the safeguard and sustainability policies. original plan, budget, and implementation targets. The portfolio changes described above have led Evaluation Design to a rethinking of the conventional project model. Evaluation rationale In programmatic lending, the country, policy, and There has not been a comprehensive evalua- reputational risks matter as much or more than tion of the WBG’s safeguard policies4 since technical and economic risks. Good project design they were first formulated in 1989.5 Previous needs to be complemented by adjustments during IEG evaluations assessed the effectiveness of implementation. Reform is aimed at consolidat- individual safeguard policies and included the ing existing investment lending policies into a 1998 report “Recent Experience with Involuntary more concise, integrated policy and operational Resettlement”6 (IEG 1998) and “Implementa- framework that differentiates projects by risk tion of Operational Directive 4.20 on to adjust project processing. This is expected There has not been a Indigenous Peoples: An Evaluation of to increase flexibility according to the risks and comprehensive evaluation Results”(IEG 2003a). The recent IEG needs of different operations and complement the of the WBG’s safeguard evaluation Environmental Sustain- emphasis of intensive effort at appraisal-tailored policies since they ability: An Evaluation of World Bank implementation support (see “Moving Ahead on were formulated. Group Support (2008b) examined Investment Lending Reform: Risk Framework and 4 E VA L U AT I O N C O N T E X T Implementation Support,” World Bank 2009d). A this report goes beyond the core question to relevant question for this evaluation is the extent explore some underlying questions: To what extent to which this retooling of the project model will have the safeguards and sustainability policies led necessitate rethinking of the model currently in to improved environmental and social perfor- place for the safeguard policies. mance and impacts at the project and sector level? How successful is the WBG in helping clients build The IFC’s and MIGA’s adoption of Performance sufficient capacity to implement these environ- Standards (in 2006 and 2007, respectively) and mental and social policy frameworks? Has the their adoption by private financial institutions9 introduction of the new Policy and Performance and the European Bank for Reconstruction and Standards led to improved environmental and Development (EBRD) provide an opportunity social appraisal and supervision at IFC and MIGA to compare the strengths and weaknesses of compared with their previous approach? What are alternative policy and implementation modalities the benefits and costs of safeguards and Perfor- for addressing environmental and social effects mance Standards? How can the WBG improve the of operations. The impact of this new direction efficiency and the development effectiveness of is evaluated to the extent possible, taking into safeguard policy frameworks? A corollary, which account ongoing efforts to develop common emerged from the portfolio challenges found approaches by other leading international by IEG, is how the safeguards and sustainability financial institutions (IFIs).10 frameworks can be adapted to maintain their relevance to the WBG’s operational portfolio. In the past, nongovernmental organizations (NGOs) have been vocal about their views on Scope of the evaluation safeguard policies and have expressed misgiv- This evaluation covers safeguards and environ- ings about the interpretation, application, and mental and social Performance Standards in the effectiveness of safeguard and sustainability WBG for projects approved in fiscal years 1999 policies.11 Each policy revision or innovation through 2008.15 Since safeguard policies do not leads to concern about a potential watering down apply to DPLs financed by the World Bank, which of such policies. On the other hand, WBG clients are governed by the Operational Policy/Bank have called for greater flexibility to suit local Procedure (OP/BP) 8.60, DPLs are excluded from conditions and capacity.12 The Bank is currently this evaluation. Evaluation of IFC’s and MIGA’s implementing a pilot program13 to test the performance distinguishes projects prepared feasibility of relying on client country systems for before and after introduction of the Perfor- implementation of safeguard policies. The pilots mance Standards. IFC projects approved after on use of country systems (UCS) for safeguards April 2006 (and MIGA projects after October are governed by the provisions of a new policy 2007) use the new Performance Standards, but (Operational Policy 4.00) approved by the Board such projects are not yet sufficiently mature for of Directors in 2004. This evaluation will also a robust ex-post evaluation of environmental and assess the UCS experience on safeguards and its social results. Consequently, this evaluation puts potential for replication. more emphasis on comparing their differences at appraisal and during implementation. Evaluation questions The evaluation’s overarching purpose is to As part of the discussion on WBG performance assess: How effective have the WBG’s safeguards during appraisal and supervision, the report also and sustainability frameworks14 been in prevent- discusses findings and lessons related to the ing and mitigating adverse environmental and WBG’s accountability mechanisms—the Inspec- social impacts? tion Panel (IPN) for the World Bank, and the Compliance Advisor and Ombudsman (CAO) Bearing in mind the evolving context since the for IFC and MIGA. This evaluation does not safeguard policies were introduced in the 1980s, have a mandate and is not designed to assess 5 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D the performance of the IPN and CAO. However, and environmental and social specialists; focus given the impact of these mechanisms on the group discussions with WBG staff; and consulta- WBG, the evaluation includes a brief review of tion with NGOs (figure 1.1). Detailed results and their activities and explores how their efficiency examples of some of the instruments used are and effectiveness can be enhanced. shown in appendix C, while projects sampled for this evaluation are listed in annex 4. (Annexes The discourse on safeguards and Performance are available on the website for this report at Standards in the WBG has been devoid of http://worldbank.org/ieg) considerations of costs and benefits, with the notable exception of a review of the cost of For the portfolio review, a random sample of 252 doing business conducted in 2001, which sought category A, B, and FI projects16 (18 percent) was ways of increasing the efficiency of fiduciary and selected from the IEG-World Bank universe17 of safeguards work. The evaluation seeks to fill this all 2,495 operations approved in fiscal 1999–2008, gap by analyzing available data and by present- giving a confidence interval of ±5.6 percent ing alternative ways of assessing risks, benefits, Bank-wide18 at 95 percent confidence level (see costs, and cost-effectiveness. Given the absence sampling details in appendix table B1). of relevant data for much of the portfolio, this is mainly an analytical contribution with a prototype For IFC, a sample of 63 projects, including of benefit-cost analysis. category A, B, and FI projects (39 non-FI and 24 FI projects, including 23 from before and 40 after Evaluation methodology the Performance Standards) was selected from In addition to a literature review and commis- the population of 403 pre-Performance Standard sioned background papers, evidence for this projects and 220 post-Performance Standard evaluation comes from desk reviews of a projects for the portfolio review, yielding a representative sample of the portfolio and field confidence interval of ±11.7 percent at 95 percent visits to purposively selected projects from all confidence level. The sample was stratified to three WBG entities; semistructured interviews mimic the population based on region, industry with clients and WBG managers; staff surveys of sector, and environmental category (A, B, FI).19 WBG task team leaders and investment officers The stratified sample of 23 pre-Performance Standard projects was drawn from the randomly sampled Expanded Project Supervision Reports (XPSRs),20 and additional performance indicators Figure 1.1: IEG Safeguards Evaluation Building were sourced from additional IEG reviews of the Blocks XPSRs. In addition, for IFC, results from IEG’s evaluation database on 394 XPSRs and Environ- Background papers Surveys, interviews, mental and Social Review Reports up to 2009 focus groups, field visits i) Use of country systems ii) Social safeguards were used when appropriate. The confidence (WBG staff, clients, NGOs, etc) iii) Accountability mechanisms (Inspection interval for this expanded dataset was 3.3 percent Panel & CAO) iv) Benefits and costs from the population of 700 projects. IEG-MIGA undertook a portfolio review of a IEG safeguards stratified sample of 35 MIGA projects approved evaluation during fiscal 2000–09.21 The sample included all 14 projects (which account for 40 percent of the Literature review Desk review portfolio review sample)22 underwritten subject Publications and WBG Random sample and reports portfolio analysis to the 2007 Policy and Performance Standards up to the third quarter of fiscal 2009 to facilitate Source: IEG. findings of MIGA’s current implementation of its policies and standards. 6 E VA L U AT I O N C O N T E X T Environmental and Social Policies for Reconstruction and Development (IBRD) at the World Bank Group and International Development Association World Bank safeguard policies (IDA), although, as will be discussed in chapter In 1989 the World Bank introduced Operational 5, some have since customized and expanded Policies and Bank Procedures for environmen- these policies.23 tal assessment of Bank-financed projects, which were updated as Operational Directive 4.01 in Policy conversion involved minor revisions, from 1991. The Bank adopted an involuntary resettle- Operational Directives (OD) and Operational ment policy as an Operational Manual Statement Manual Statements (OMS) into Operational in 1980, which was revised as OD 4.30 in 1990. Policies (OP) and Bank Procedures (BP). The Other environmental and social policies were first OP on Pest Management was approved added over time to address individual environ- in 1998. The Environmental Assessment OD mental and social risks. was replaced by OP and BP 4.01 in 1999. The Involuntary Resettlement policy was converted In 1997 the Bank identified 10 policies as its suite to OP/BP format in December 2001. 24 The of safeguard policies, labeled them “do no harm” conversion process continued until 2006. Each policies, and started a process of policy conver- policy had a different set of stakeholders, so the sions for individual policies. The safeguard policy conversion was piecemeal and, accord- policies (see table 1.1) consist of six environmen- ing to the staff involved in this process, involved tal, two social, and two legal policies. Many other protracted discussions with a wide range of multilateral development banks (MDBs) initially stakeholders, leading to a lengthy process which based their own safeguard policies for public for the Indigenous Peoples Policy lasted seven sector lending on those of the International Bank years. The policy on International Waterways Table 1.1: Comparison of WBG Safeguards and Performance Standards IFC/MIGA Policy and Performance Standards on Social and Environmental Sustainability Bank Safeguard Operational Policiesa (2006/2007) Environmental and social PS 1: Social and Environmental Assessment and Management System Environmental 4.01 Environmental Assessment (1999) PS 6: Biodiversity Conservation and Sustainable Natural 4.04 Natural Habitats (2001) Resource Management 4.36 Forests (2002) PS 3: Pollution Prevention and Abatement 4.09 Pest Management (1998) PS 8: Cultural Heritage 4.11 Physical Cultural Resources (2006) 4.37 Safety of Dams (2001) Social 4.12 Involuntary Resettlement (2001) PS 5: Land Acquisition and Involuntary Resettlement 4.10 Indigenous Peoples (2005) PS 7: Indigenous Peoples PS 2: Labor and Working Conditions PS 4: Community Health, Safety and Security Legal 7.50 International Waterways (2001) 7.60 Disputed Areas (2001) Source: World Bank Group. Note: PS = Performance Standard. a. Except for pest management, all World Bank Operational Policies (OP) have accompanying Bank Procedures (BP). Consultation and disclosure processes are integral to the WBG safeguard and sustainability policies. 7 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D is currently being revised, a process now in its • Better balance in thematic coverage of envi- second year. The procedure for policy revisions, ronmental and social issues with the addition even small ones, has proved to be so cumber- of new issues relevant to the private sector some and time consuming that there is great • Complementing procedural compliance with reluctance to revise and improve the policies accountability for environmental and social even when the lessons of experience suggest performance, but with gaps in verification and that this would be beneficial. disclosure. Policy revision has proved When the safeguard policies were The PPSSES is an integrated policy framework to be so cumbersome labeled “do no harm” policies, the with an umbrella policy on environmental and and time-consuming that Bank’s senior management made social sustainability and relatively well-balanced there is great reluctance public commitments to enforce treatment of environmental and social effects. to revise and improve the compliance with these mandatory IFC added Performance Standards on Labor and policies even when this requirements, leading to significant Working Conditions (Performance Standard 2), would be beneficial. improvement in environmental and and Community Health, Safety, and Security social performance compared with (Performance Standard 4) to the two Perfor- the 1990s. However, the Bank’s list of safeguard mance Standards derived from the Bank’s policies was restricted to existing policies social safeguards. IFC does not have Perfor- designed to mitigate adverse environmental mance Standards on dam safety or on gender. and social impacts, effectively freezing policy However, the guidance notes for Performance development in the state that existed at that time. Standard 1 describe a Gender Impact Assess- Existing policies on Sociological Appraisal, which ment that should include measures to ensure is a part of the Bank’s policy on Project Appraisal that one gender is not disadvantaged relative (OMS 2.20), and Gender and Development (OP to the other in the context of the project. Some 4.20) were excluded from the safeguards suite. IFC projects integrate gender impacts within their community impact study for Performance The existence of an umbrella policy for Environ- Standard 4, and the assessment of impacts on mental Assessment provided an open-ended indigenous peoples can also be combined with mandate for engaging with borrowers and clients the community impact assessment. The WBG on the environmental agenda. By contrast, the revised its environmental, health, and safety restriction of social safeguards at the Bank to two (EHS) guidelines in the Pollution Prevention prescriptive policies focused attention on these and Abatement Handbook 1998 with a new set two effects but narrowed their relevance to a of industry-specific EHS guidelines, and EHS much smaller segment of the portfolio. Social general guidelines (April 2007). IFC has used risks subsequently addressed by IFC and MIGA applicable EHS guidelines, earlier safeguard have also not been integrated into the Bank’s policies, and present Performance Standards, safeguard policies. together with project-specific environmental and social requirements and Environmental and IFC and MIGA Performance Standards Social Action Plans, as covenants in its invest- The role of policy innovator within the WBG has ment projects. shifted from the Bank to IFC, whose Policy and Performance Standards on Social and Environ- IFC’s PPSSES have been emulated by other mental Sustainability, approved in 2006, has since financing organizations. MIGA adopted the been emulated by others (see table 1.1).25 The PPSSES in 2007, and in a somewhat modified PPSSES framework involves:26 form the European Bank for Reconstruction and Development adopted a similar policy in 2008. • Clearer roles and responsibilities for IFC and Over 60 private sector banks have voluntarily MIGA and the client in project preparation and adopted a set of Equator Principles, which now implementation include the Performance Standards approach, as 8 E VA L U AT I O N C O N T E X T a framework to address environmental and social reputational and social safeguard risks. However, issues in project finance. Both private sector responsibility for project processing and supervi- lenders and clients thus appear to be buying sion of lower-risk projects is delegated to the into the Performance Standards approach, appropriate sector management unit. although the short time since their introduction has prevented robust evaluation of outcomes Project implementation is the responsibility of and impacts. Global evidence also suggests that the borrower, while the Bank is responsible for voluntary adoption of safeguards by private supervision. Requirements vary depending on sector clients is often inadequate for mitigating the number and nature of safeguards policies social risks. It is, however, feasible to compare triggered by the project. the relevance of the Performance Standards to that of the safeguard policies. International Finance Corporation (IFC) IFC’s business model and project cycle are Roles and Responsibilities adapted to private sector clients and differ from World Bank those of the Bank. After IFC’s business develop- Since 1999 the Quality Assurance and Compli- ment officers have identified an investment ance Unit (QACU) and the Environmental and opportunity, an investment officer prepares a International Law Unit of the Legal Department project description in the Project Data Sheet— have provided central guidance on all matters Early Review 29 for IFC senior management relating to safeguards. All investment lending authorization of project appraisal, if warranted. operations follow a set of regular safeguard The investment team (which includes an procedures throughout the project life cycle.27 environmental and social specialist), during the (See appendix A for details of each safeguard.) appraisal (or due diligence) phase, assesses in detail the business potential and risks, includ- In 2006 the Bank consolidated two key ing environmental risks, and determines the networks—the Environmentally and Socially final categorization and action plans needed to Sustainable Development (ESSD) Network and comply with IFC’s detailed environmental and the Infrastructure Network—into the Sustainable social, disclosure, and consultation require- Development Network under one vice president, ments.30 With the client’s approval, the Environ- bringing the environmental and social staffs and mental and Social Review Summary (ESRS) and their internal clients from the infrastructure the Environmental and Social Action Plans are and agricultural sectors under one umbrella. At posted on the IFC website before being submit- the time of that merger, QACU and its counter- ted for Board approval.31 After adoption of the parts—the Regional Safeguards Advisors—in Performance Standards in 2006, IFC developed the Regions were transferred from ESSD to the an internal online Environmental and Social Operations Services group, to ensure that project Review Document (ESRD) system to identify, clearances were not unduly influenced by being rate, and monitor performance indicators. As housed within the same Network, to offset the with Bank projects, project implementation is perception of conflict of interest. the client’s responsibility, while IFC is respon- sible for supervision. Bank safeguards specialists provide guidance to task teams on applicability of safeguard policies, The building blocks of IFC’s new sustainabil- on the assessments and consultations to be ity framework consist of IFC’s 2006 PPSSES, undertaken and mitigation plans prepared by the the Guidance Notes and Policy on Disclo- client, and on the appraisal and disclosure require- sure of Information, and the newly revised ments to be met prior to project approval. 28 The EHS Guidelines, and Environmental and Regional Safeguard Advisor retains oversight Social Review Procedure (ESRP). Implemen- responsibility for all category-A projects and tation success depends equally on relevance category-B and -FI projects with potentially high and coverage of the sustainability framework, 9 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D proper IFC staffing, capacity, and resources, as disclosed the Environmental and Social Impact well as client commitment, skills, and capacity, Assessment (ESIA) for category-A projects, and available funds for environmental and social which would continue to be disclosed. investments. Real sector clients of category- • Examination of social and environmen- A and -B projects are obligated to provide tal management systems of financial in- an Annual Monitoring Report (AMR), and FI termediaries to verify that the FIs’ systems clients an Annual Environmental Performance are sound and appropriate for the specific Report. The environmental and social specialist cases, given the nature of their business. This reviews the annual report and prepares a formal includes an examination of the SEMS of the Review Report, which provides information on parent banks and of how it is applied to their data quality, compliance status, feedback to subsidiaries, including an initial assessment of the client, and the Environmental and Social local capacity and social and environmental Risk Rating.32 IEG has evaluated IFC project’s risks in the portfolio. Environmental and Social Effects since 1996 as a • Technical assistance to clients to meet part of the validation of the XPSRs prepared by the Performance Standards. MIGA has, in the project teams. 33 the past, not been able to provide technical expertise or financial support to its clients to Multilateral Investment Guarantee Agency help ensure that they meet its environmental (MIGA) and social standards. This changed in a lim- MIGA’s mandate, since it was established in ited way with the establishment of the Trust 1988, has been to encourage the flow of private Fund to Address Environmental and Social investment to WBG clients by offering politi- Challenges in MIGA-guaranteed projects in cal risk guarantees. MIGA policy requires all Africa. With the support of the government projects it supports to comply with applicable of Japan, this initiative launched a three-year MIGA environmental policies and guidelines. test of whether such technical assistance can Its work with clients focuses on environmental be provided and will be helpful, in the context assessment and monitoring of project compli- of an insurance provider rather than a lender ance with environmental and social guidelines or equity investor. and safeguards. MIGA followed applicable Bank • Local Community Development Effec- policies and used IFC staff for the environmental tiveness Reporting. This initiative was de- and social review of its operations during much signed to address concerns about the possible of the 1990s but established its own environ- impact of certain projects on the local commu- mental office in 1998. Since then, the role and nity, in particular when these impacts might composition of MIGA’s environmental and social be negative. MIGA therefore proposed that it unit has evolved and expanded to include two would regularly report on the local community social specialists. MIGA’s Environmental Assess- impacts of a small number of projects where ment and Disclosure Policies were approved by such impacts may be significant. the Board in 1999, and its issue-specific safeguard policies were approved on an interim basis in Portfolio Trends 2002. Following IFC, MIGA adopted the PPSSES The safeguards and sustainability policies were in October 2007. originally conceived for investment projects. They are more difficult to apply to other forms In tandem with new PPSSES, MIGA proposed and of lending, including programmatic lending, adopted four related initiatives:34 sectorwide lending, and decentralized projects at the World Bank; trade finance and equity invest- • Preparation and disclosure of ESRS for ments at IFC; and financial sector lending at all category-A and -B projects, together with MIGA. All three portfolios appear to be growing a summary of proposed guarantee similar to in precisely those segments where these policies the IFC process. Previously, MIGA had only face their greatest challenges. 10 E VA L U AT I O N C O N T E X T Portfolio trends at the World Bank Variations in environmental and The proportion of Bank The proportion of projects classified as cate- social risk within the portfolio are projects classified as gory B increased by a third, while those clas- affected by the nature of project category B increased sified as category C decreased by half during lending. Among the regions, East by a third while the period reviewed, reflecting greater cau- Asia and Pacific (EAP) has the highest category C decreased tion during project preparation. At the World proportion (23 percent) of category- by half, reflecting Bank a total of 2,495 lending operations were A projects, driven by infrastructure greater caution during approved during fiscal 1999–2008, of which projects, while Latin America and project preparation. 1,133 (45 percent) had been completed; the the Caribbean (LCR) has the lowest rest were still active. The distribution of proj- (4 percent). Europe and Central ects by safeguard category is depicted in fig- Asia (ECA) relies the most (13 percent) on FI ure 1.2. Over the 10-year period, 9 percent lending and has relatively fewer category-A of the universe was classified as category A and -B projects. The proportion of category- (very high impact), 44 percent as category B A projects increases with lending size while (substantial impact), 29 percent as category category-C projects are most prevalent among C (low impact), and 4 percent as category FI, smaller projects. FI projects are evenly distrib- but the distribution has changed substantially uted across different loan sizes. over time. During the review period, the pro- portion of category A increased from 5 to 11 Portfolio trends at IFC percent, with the increase in the volume and Trends in IFC’s portfolio are depicted in figure scale of infrastructure lending. Category B in- 1.3. The share of category-A projects in numbers creased from 37 to 51 percent, while catego- has declined since introduction of the PPSSES but ry C dropped from 40 to 18 percent. IEG was remains at the same level as earlier in commit- unable to detect any substantial change in the ment amount. Financial intermediary projects portfolio to explain the substantial increase in are about 32 percent by number of projects and category-B projects. slightly less by commitment amount. Figure 1.2: Bank Lending by Safeguard Category, Number, and Commitment (FY1999–2010) 100% 100% 100% 30,000 Percentage of investment commitments 25,000 Commitment amount (US $m) 80% 80% 80% Percentage of projects Number of projects 20,000 60% 60% 60% 15,000 40% 40% 40% 10,000 20% 20% 20% 5,000 0% 0% 0% 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Category: A – No. of projects B – No. of projects Investment – No. of projects DPL – No. of projects C – No. of projects FI – No. of projects A – Commitments Investment – Commitments DPL – Commitments Source: World Bank database (as of April 12, 2010). 11 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 1.3: Trends in IFC’s Portfolio 60% 50% 40% 30% 20% 10% 0% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Committment year (FY) A (by volume) B (by volume) FI (by volume) C (by volume) A (by number) B (by number) FI (by number) C (by number) Source: IFC database. In the past decade, IFC’s business has shifted ment) as well as IFC’s subscription for shares away from project finance toward financial in a company. IFC’s leverage from a minority intermediary, corporate, equity, and trade equity investment in a company that includes finance projects. Though IFC’s environmental a wide range of operations is more restricted procedures were created for a project finance compared with traditional project finance, but institution, by fiscal 2006 only 28 percent of the scope of IFC’s environmental and social IFC business was in project finance. With IFC’s review is limited to the countries or facilities move to wider portfolio risk management, the where IFC financing is directed. IEG interviews environmental and social risks have extended with 21 managers and local environmental and beyond the project’s area of influence to the social specialists revealed that IFC staff regard client’s business and environmental manage- the Performance Standard framework as fully ment as a whole. This development makes it feasible for project finance and corporate all the more imperative to develop the client’s loans with identified use of proceeds, but social and environmental management system, much less feasible for trade finance and equity and ensure adequacy of its implementation. investments in listed companies, which are not obligated to report annually to individual IFC’s corporate or equity investments in shareholders without compromising the legal companies with several production facili- rights of other shareholders. This may be ties and various activities pose a substan- mitigated by the fact that large internationally tial challenge for environmental and social listed companies often possess a sound social appraisal, supervision, and evaluation. In and environmental management system with corporate finance, use of proceeds is not good reporting practices and publicly available limited to specific assets, but are intended for Corporate Sustainability Reports, which, if corporate activities (restructuring, long-term transparent and complete, may serve as an strategic support, corporatewide invest- adequate reporting platform. 12 E VA L U AT I O N C O N T E X T Portfolio Trends at MIGA 1.4). In addition, MIGA has experienced a decline MIGA’s Convention and Operational Regula- in the number of new projects supported each tions, requiring MIGA to support projects that year, which decreased from 33 (fiscal 2005) to are consistent with host-country laws, regula- 20 (fiscal 2009). The increasing concentration tions, and development objectives, provide on financial sector projects has implications the institutional basis for the agencies sustain- for the implementation of MIGA’s sustainability ability framework. Its policies and guidelines framework. require that each project for which MIGA issues a guarantee is carried out in an environmen- Organization of the Report tally responsible manner in accordance with its The report is organized into six chapters. Chapter sustainability policy (PPSSES) and new policy 1 provides the evaluation context, objectives, and on Disclosure of Information. Its sustainability rationale, an introduction to the safeguard and framework also includes ensuring compliance sustainability policies (see details in appendixes A with IFC’s Environmental, Health, and Safety and E), the scope of the evaluation, and an outline Guidelines and relevant IFC industry and sector of the methodology (detailed in appendix B). guidelines. Chapter 2 examines the effectiveness of the WBG in complying with policy requirements, including MIGA’s portfolio composition has shifted over the quality of preparation and appraisal, supervi- time: the share of guarantees for financial sector sion, and monitoring, and includes the IPN and projects increased significantly during the past CAO findings. For IFC and MIGA it also compares decade. The amount of MIGA guarantees issued the findings for projects prepared before introduc- averaged $1.5 billion annually between fiscal 2000 tion of the PPSSES (pre-Performance Standards) and 2009, with considerable variation from year with projects appraised since their introduction to year. The financial sector now represents the (post-Performance Standards). Chapter 3 evaluates largest business segment in MIGA’s portfolio. At environmental and social performance of the the same time, the importance of the infrastruc- sample portfolio against the objectives mapped out ture and agribusiness, manufacturing, and in the respective assessments of relevant risks. The services sectors has shrunk significantly (figure chapter assesses the quality of client implementa- Figure 1.4: Changes in MIGA’s Portfolio Composition (share of MIGA guarantee volume issued per sector) FY00–04 FY05–09 Infrastructure Infrastructure 44% Oil, Gas 29% and Mining 10% Oil, Gas Agribusiness, and Mining Manufacturing 9% and Services 16% Agribusiness, Financial Sector Manufacturing Financial Sector 53% and Services 30% 9% Source: IFC database. 13 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D tion and the effectiveness of the safeguards and draws on the findings from the previous chapters sustainability frameworks in mitigating adverse to reconsider the relevance of the safeguards impacts, strengthening client capacity, and enhanc- and sustainability policies, summarizes the main ing positive impacts to promote development findings on the Bank country systems pilots, and effectiveness. Chapter 4 examines the robustness examines how the WBG can improve efficiency of the categorization system in use to classify of safeguards policy frameworks and strengthen projects by comparing results with those obtained their benefits. It also compares the WBG from application of a risk model to the portfolio. safeguards frameworks with those of major IFIs The risk model is also used to estimate benefits, and evaluates the Bank’s experience with adoption which are then compared with available data on of country systems for safeguard policies. Chapter costs to assess the efficiency of resource allocation 6 summarizes the conclusions and puts forward by the Bank, IFC, and country clients. Chapter 5 recommendations for the WBG. 14 Chapter 2 Evaluation Essentials • The World Bank Group (WBG) gives much better attention to safeguards and Performance Standards in proj- ect preparation and appraisal than during supervision. • The criteria for categorization of projects based on environmental and social risks differs across the WBG, with IFC and MIGA using a different approach than the Bank. • Several high-risk, category B cases in IFC would have likely been cat- egorized as category A projects using the Bank's screening system. • Bank supervision varies consider- ably by region, particularly between high and low performers. • Performance indicators for safe- guards are rarely specified and in- tegrated in the results framework, and data for monitoring and evalu- ation are not routinely collected or used by the Bank and MIGA. • IFC has improved monitoring with explicit client responsibility for an- nual monitoring using specified per- formance indicators. • IFC supervision quality is affected by the timeliness and quality of An- nual Monitoring Reports prepared by clients. Construction workers excavating a canal-bed beneath a Hanoi bridge. Photo by Tran Thi Hoa, courtesy of the World Bank Photo Library. Process Implementation by the World Bank Group Introduction on lessons from the IPN and CAO on process This chapter examines the quality of prepara- implementation. tion and appraisal, supervision, and monitor- ing of safeguards and Performance Standards in The findings from multiple sources of data WBG-financed operations. The specific questions indicate much better attention to safeguards evaluated are: and Performance Standards in project prepara- tion and appraisal than during supervision. This • How effective was WBG due diligence during is partly a function of the front-loaded nature project preparation and appraisal, and quality of the policy frameworks, which have more of WBG supervision, monitoring and evalua- detailed instructions and explicit standards for tion? compliance during project preparation than • Has the introduction of the Policy and Perfor- during supervision. As a result, Bank and MIGA mance Standards led to improved environ- management earmarks funds for safeguards mental and social appraisal and supervision work with designated teams of specialists to at IFC/MIGA compared with their previous ensure compliance during project prepara- approach? tion but not during supervision. Deficiencies in supervision are more acute in the Bank and The data for the assessment come from portfo- MIGA but are also found in IFC. lio reviews of a sample of category-A, -B, and -FI projects approved during fiscal 1999–2008 Evaluation of Process Implementation: by the Bank, IFC, and MIGA. Portfolio review World Bank results1 were triangulated with data from other Preparation and appraisal instruments, including surveys of all task team The aggregate quality of due diligence during leaders at the Bank and investment officers preparation and appraisal was found to be 85 at IFC, and environmental and social staff at percent satisfactory (table 2.1). The evalua- both organizations, 2 client surveys for Bank tion of environmental and social due diligence and IFC projects, interviews with managers,3 a covers the safeguard identification and screen- survey of NGOs,4 and focus group discussions ing process based on significance of environ- with environmental and social specialists at the mental and social risks, the quality of due Bank, IFC, and MIGA. The analysis also includes diligence evident from appraisal documen- information obtained from papers commis- tation, and compliance with disclosure and sioned for the study on social safeguards and consultation requirements. Screening for Performance Standards at the Bank and IFC and environmental and social risks is followed by 17 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D preparation of environmental assessments ments5 to be undertaken by the client during (EA) and assessments of social risks arising project preparation. Category-A projects with from involuntary resettlement or impacts on high environmental risks require prepara- indigenous peoples. Projects with significant tion of an Environmental Impact Assessment environmental and social risks are expected to (EIA) and an Environmental Management Plan consult with relevant stakeholders and disclose (EMP). Category-A projects with high social the assessments and action plans before project risks require the preparation of a Resettlement approval. Performance under each of the three Action Plan (RAP) or an Indigenous Peoples Plan elements—screening, risk assessment, and (IPP), as appropriate. Assessments for category- consultation—was better than 90 percent, but B projects with limited, site-specific environ- 15 percent of projects had deficien- mental and social risks that can be addressed The aggregate quality cies in one or more of these elements, more easily can have a narrower scope. For of due diligence during and relevant environmental or social projects with multiple subprojects and limited preparation and expertise was lacking in 11 percent impacts, where the exact nature of impact is not appraisal was found to be of projects during preparation (see known at appraisal, projects can also prepare a 85 percent satisfactory. annex 1, table X1.2). Resettlement Policy Framework or an Indige- nous Peoples Framework that spells out the Identification and screening requirements and procedures for the client to IEG found identification and screening for follow during implementation, when subproj- environmental and social risks fully satisfac- ects are identified. Although not reflected in the tory in 87 percent of the projects reviewed. environmental policy, in practice projects under Twenty-four percent of the sample had been similar circumstances also prepare Environmen- classified as category A, 70 percent as category tal Management Frameworks (EMF) or Environ- B, and 6 percent as category FI. Screening mental and Social Management Frameworks determines the scope and depth of environ- (ESMF). 6 FI projects also normally require mental assessment and/or the social assess- preparation of similar policy frameworks, which Table 2.1: Safeguards Preparation and Appraisal in Bank Projects (percent satisfactory) Environmental Identification and and social impact Disclosure and Overall preparation and screening assessment at appraisal consultation appraisal # % # % # % # Region Rated Satisfactory Rated Satisfactory Rated Satisfactory Rated % Satisfactory Africa 60 85 58 86 57 91 60 78 East Asia & Pacific 49 92 48 96 46 100 49 94 Europe & Central Asia 49 82 47 87 45 93 49 84 Latin Amer. & 44 82 41 78 34 91 44 75 Caribbean Middle East & N. 19 89 19 100 15 87 19 89 Africa South Asia 31 100 31 100 30 93 31 94 TOTAL 252 87 244 90 227 93 252 85 Note: “Satisfactory” refers to portfolio review ratings of satisfactory (S) or excellent (E) on a four-point scale by IEG reviewers. 18 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P have to be endorsed by the client before project exclusively due to overcategorization. The portfolio review approval. In part, the increase in category B is confirmed a tendency of due to additional guidance issued by risk avoidance through The portfolio review confirmed a tendency QACU on classification of technical overcategorization. of risk avoidance through overcategorization, assistance and land administration which seemed to be emerging from the analysis projects, based on the realization that of categorization in the entire fiscal 1999–2008 some technical assistance projects had supported universe (table 2.2). The proportion of category- project preparation activities, and some land B projects increased steadily by a third, while administration projects were leading to changes category C dropped to less than half its fiscal in land use with potential environmental impacts. 1999 figure in the same interval. When assessed However, there is a fairly widespread perception against current norms, 15 projects were found by among task team leaders that the upward classi- IEG to be overcategorized, having overestimated fication is driven by risk aversion rather than the safeguards category—11 from C to B and 4 an empirical assessment of environmental and from B to A. One project in Africa had underesti- social risks. Among the B projects with category mated a category-A project as B. Projects miscat- deficiencies, task team leaders felt that 15 percent egorized as A had relatively limited, site-specific should have been category A, while 77 percent impacts that were not sensitive or irreversible, or should have been category C, because they had no impacts in this phase of the project. Overcat- no, or low, environmental and social impact. egorization results in additional preparation costs to the Bank and clients. Five other projects had The review also found some lack of clarity in use contradictions between policies triggered at of the FI category. Five projects were affected by appraisal and those reported on subsequently in this—4 out of 16 projects in category FI were not the Implementation Status and Results reports being administered by financial intermediaries and (ISRs), indicating improper, often overly cautious should have been classified as category-B projects, triggering of safeguard policies when impacts while 1 project that was reported as category B was were not known. in fact being administered by a financial intermedi- ary and should have been category FI. A tendency toward overcategorization in the Bank was also identified by IEG’s staff survey. Quality of environmental and social impact Eleven percent of task team leaders and 8 assessment at appraisal percent of safeguards specialists reported that The evaluation found the quality of EA/EMPs their projects were misclassified, mostly but not and RAP/Social Assessments satisfactory in over Table 2.2: Safeguards Category by Approval Year (percentage of projects) Category FY99 FY00 FY01 FY02 FY03 FY04 FY05 FY06 FY07 FY08 Average A 5 8 8 8 10 9 9 9 12 11 9 B 38 38 42 39 40 47 45 48 46 51 44 C 45 43 33 33 30 29 26 21 16 19 29 F 1 4 7 6 6 7 5 5 4 2 4 U 11 7 9 14 14 9 15 18 22 18 14 All projects 100 100 100 100 100 100 100 100 100 100 100 Note: Category U includes projects uncategorized for safeguards category. 19 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 90 percent of the projects reviewed. Of the 24 while the other 4 had weak RAPs that did not fully projects found deficient in environmental due identify the social impacts arising from involun- diligence, EA/EMPs could not be located for tary resettlement (box 2.1). 5; the EAs for 8 were generic, with insufficient assessment of the risks relevant to the project; 5 were health or education projects lacking Consultation and disclosure specific measures for medical waste manage- Overall, 93 percent of the projects had adequate ment or sanitation; 3 lacked clarity on mitigation consultation and disclosure. Deficiencies were actions leading to weak client implementation; largely related to timeliness of disclosure and and 2 had details not needed by the project. availability of the EAs or RAPs in the Bank’s public Of the 10 projects found deficient in document database. The major deficiency was the The quality of EA/EMPs RAP/SA preparation, 6 should have absence or inadequacy of any description of consul- and RAP/Social prepared detailed plans instead of tations conducted during project preparation, or Assessments was policy frameworks before appraisal missing EAs and Integrated Safeguards Data Sheet satisfactory in over since the alignments were known and in the the Bank’s internal document database. Task 90 percent of the microimpacts of these infrastructure teams are obliged to ensure that project documen- projects reviewed. projects could have been assessed, tation is in the public information domain and Box 2.1: Magnitude of Involuntary Resettlement in Bank Operations IEG was unable to obtain the magnitude of project-induced tion Reports showed 298,415 were actually affected (88 percent involuntary resettlement in the portfolio from Bank sources and of the original estimate), indicating that during implementation made a special effort to estimate this magnitude from the review it is possible to reduce impacts. Nonetheless, the resettlement sample. Half of the sample triggered involuntary resettlement. impact of Bank-financed activities is nontrivial and merits careful The total number of project-affected persons in the sample monitoring to ensure that it does not lead to impoverishment of was 418,049, of which 41 percent were physically displaced; affected persons. the rest faced impacts on livelihoods. Excluding fiscal 1999 as an outlier (which had 148,263 new project-affected persons), Planned and Actual Project-Affected Persons (PAPs) in Completed Projects, by Approval Year the magnitude of resettlement in each approval year averaged 180 29,976 new project-affected persons per year within the portfolio sample. The sample is 18 percent of the universe of projects 150 affected by safeguards. Extrapolating to the universe this gives PAPs (thousands) 120 an average of 166,535 new project-affected persons per year. With an average project life of 6–7 years, the total number of 90 persons subjected to involuntary resettlement in the Bank’s active portfolio falls within the range 999,207–1,165,742. IEG 60 estimates that at any given point in time over 1 million people, 30 two-fifths of them likely to be physically displaced, are affected by involuntary resettlement in active Bank-financed projects. 0 This is half of the 2 million estimated in 1994, when hydropower FY99 FY00 FY01 FY02 FY03 FY04 FY05 FY06 FY07 FY08 dams constituted a much larger share of the Bank’s portfolio. PAPs planned PAPs actual 3 year moving avg. Compared with the 339,519 project-affected persons identified (PAPs planned) in RAPs of 20 completed projects, the Implementation Comple- Note: Actual numbers of project-affected persons are not available for recent years. Source: IEG portfolio review. a. Michael Cernea undertook a review of resettlement in the 1990s and estimated that “Projects currently in the Bank’s active portfolio are expected to involve the resettlement of 2 million people over an eight-year period.” World Bank 1994. 20 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P accessible to all to ensure that stakeholder feedback relatively low-cost investments Of the 122 projects that is based on adequate access to project information. to ensure country ownership for triggered involuntary Documents that cannot even be retrieved inside Bank operations. During the period resettlement, 94 the Bank cannot be expected to be accessible to reviewed, the East Asia and Pacific percent had adequate in-country stakeholders. region was the most diligent in terms consultation and of disclosure and consultation regard- disclosure. Of the 122 projects that triggered involuntary ing safeguard impacts. resettlement, 94 percent were found to have adequate consultation and disclosure. Deficien- Evaluation of Process Implementation: IFC cies were noted in 7 projects, only 2 of which The staff and budget at IFC’s Environmental and involved insufficient disclosure of documents or Social Investment Support Group (CESI) have resettlement entitlements to people affected by increased as IFC commitments increased. Until the project. Two projects were due to resettlement recently there was only one full-time environ- needs identified during implementation whose mental and social specialist for FI projects, documents had not yet been placed in the public despite growing FI investments. In 2009 CESI had domain, and 3 were projects whose resettlement 50 environmental and social specialists, including documentation had not been filed in the Bank’s 13 specialists for social appraisal and supervision, internal or external document databases. and 5 specialists devoted to FI projects, out of a total staff of 72. Trends of IFC net commitments The lack of access to information can needlessly (excluding dropped projects) by categories (A, B, generate suspicion and criticism about Bank FI, C) and the environmental and social budget operations. Meaningful stakeholder consulta- and staff for real sector (non-FI) and FI sector tion, transparency, and timely disclosure of projects’ appraisal and supervision are shown in relevant project documentation in an easily figure 2.1. The global financial crisis and increase accessible manner by interested parties are of trade finance (mainly category-C projects) is Figure 2.1: IFC Net Commitments by Category and CESI Resources 20.0 100 18.0 90 16.0 80 Net committments ($bill) 14.0 70 12.0 60 CESI staff 10.0 50 8.0 40 6.0 30 4.0 20 2.0 10 0.0 0 FY00 FY01 FY02 FY03 FY04 FY05 FY06 FY07 FY08 FY09 Commitments: C & no category ($bill) Commitments: FI cat ($bill) Commitments: B cat ($bill) Commitments: A cat ($bill) CESI budget & consultants ($mill) CESI total staff CESI staff for FI sector Source: IFC’s MIS database and Environment and Social Department database. 21 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D clearly visible in fiscal 2008–09 figures in overall Identification and screening declining volume and increasing category-C Although most of the sampled projects (97 volume. percent) were correctly categorized based on the significance of environmental and social Quality of IFC appraisal impacts, IEG found that the guidance on project IEG found strong correlation between environ- categorization has been weak in both pre- mental and social appraisal and supervision and post-Performance Standard frameworks, quality and the Environmental and Social Effects leading to different approaches between the (ESE) indicator.7 The ESE rating is based on the Bank and IFC, and to incorrect categoriza- project’s environmental performance in meeting tion and reputational risks in some projects IFC’s requirements as well as the project’s environ- with supply chain risks. Project categorization mental impacts. Out of nearly 300 evaluated affects the depth of environmental studies, XPSRs, a high rating (excellent and satisfactory) public consultations, reporting, and frequency for environmental and social appraisal quality of supervision, and signals the urgency and resulted in high ESE ratings in 65 percent of the associated environmental and social risks to the projects. For environmental and social supervi- public. For example, EBRD9 has clearer industry sion, this resulted in high ESE ratings in 58 percent sector guidance for categorization. The share of of projects (table 2.3). category-A projects by number of projects has dropped by half, from 6.2 percent before the The overall quality of IFC’s work at prepara- Performance Standards to 3.6 percent after. tion and appraisal has been high for both real But by net commitment volume, the share has sector and FI projects on all measures except remained at 12 percent level after fiscal 2006, disclosure and consultation, where IFC lacked as the average size of category-A projects has information on a substantial proportion of increased more than that of B and FI projects projects, and shows no difference (table 2.5, also see figure 1.3). Interviews and between projects prepared before8 focus group discussions with IFC staff revealed Guidance on project and after the introduction of the selection bias and pressure from investment categorization has Performance Standards (pre- and departments to prefer category B instead of been weak in both pre- post-Performance Standard), despite category A in order to speed up appraisal and and post-Performance increasing environmental and social implementation. Standard frameworks. requirements (table 2.4). Several high-risk, category-B cases would have likely been categorized as category-A projects using the Bank’s screening system. In the evalua- tion’s judgment, this difference affects 27 percent (10/37) of the category-B projects in the sample. Table 2.3: Correlation of Environmental and Social In 5 cases that involved the construction of new Appraisal and Supervision Ratings with ESE Ratings infrastructure or greenfield facilities, the scale in IFC Projects of the impacts would have led IBRD to classify them as category A. In six additional cases, the sensitive nature of the impacts—associated as n = 298 p = 0.005% n = 293 p= 0.000% they were with hazardous waste, indigenous Environmental and Environmental and peoples, natural habitats, or cultural resources— social appraisal social supervision would have likely led IBRD to classify them as category A. Categorization, in principle, would ESE Low High ESE Low High be a major determinant of the eventual environ- High 4% 65% High 10% 58% mental and social outcomes. While the categori- Low 7% 24% Low 15% 17% zation of these projects appears to have been in Source: IEG’s Environmental and Social Reviews for XPSR projects, appraised fiscal 1999–2004. compliance with IFC’s procedures, IBRD would 22 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P Table 2.4: IFC’s Quality at Preparation and Appraisal Pre-Performance Post-Performance Standard Standard Number Satisfactory Number Satisfactory All projects 347 88% 35 89% Real sector (non-FI) overall 208 90% 23 87% Safeguards identification and screening 15 93% 24 92% Environmental and social impact assessment and appraisal 15 80% 22 95% Disclosure and consultation 14 57% 21 71% FI sector overall 139 86% 12 92% Source: Pre-Performance Standard; IEG’s Environmental and Social Reviews fiscal 2004–09 (209 non-FI and 139 FI projects) and study portfolio (15 real sector projects). Post-Performance Standard; study portfolio of 24 real sector and 17 FI projects. Note: The differences in pre- and post-Performance Standard ratings are not statistically significant. likely have classified them otherwise, pointing to issued a statement that the WBG Several high-risk, a lack of consistency of safeguards implementa- would not make public investments category-B cases (in tion across the WBG. in this specific commodity sector IFC) would have likely until a common approach in the been categorized as Due diligence for two trade finance projects industry sector was established. IFC is category A using the with supply chains to agribusiness was found currently developing a new approach Bank’s screening system. inadequate. Trade finance projects have grown to address supply chain and biodiver- rapidly and represented 23 percent of IFC sity issues in similar agribusiness commodity commitments in fiscal 2009.10 These projects are projects. categorized as category C under the Performance Due diligence for Standard framework11 and are only required to The appraisal process has been trade finance projects comply with the trade finance exclusion list.12 systematized in the post-Perfor- with supply chains A complaint was filed by NGOs, smallholders, mance Standard framework to agribusiness was and indigenous people’s organizations in 2007 with the improved ESRD and a found inadequate. against two IFC agribusiness commodity projects in Southeast Asia, which had been labeled category C because the projects were defined as trading facilities in spite of direct supply chains Table 2.5: Average Project Size at IFC by Category to client-owned agribusiness operations. After examining the complaint, the CAO concluded that commercial pressures dominated IFC’s $ million/project assessment process. Environmental and social due diligence reviews did not occur as required, Category 2000–05 2006–09 Increase % and IFC did not meet the intent or requirements A 34.7 105.4 204% of its policy for assessment of trade facility B 17.9 29.4 64% investments and each project’s supply chain. IFC FI 13.9 32.3 132% management accepted CAO findings on project categorization, strategic sector framework, and Total 17.1 32.2 89% supply chain due diligence, and the president Source: IFC database. 23 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D The quality of due structured approach to monitoring essential outcome indicators that both IFC diligence on social performance indicators, as earlier and the client can benchmark against IFC safeguards and recommended by IEG. Still, some requirements and industry best practices. Performance Standards important and sensitive environ- Global issues received better attention in the during appraisal mental and social aspects, such as 2006 Performance Standard framework, but in IFC-supported environmental legacies or social recycling and energy efficiency are not fully projects has generally liabilities, have been occasionally mainstreamed. been satisfactory. overlooked due to reliance on client reporting. Although overall appraisal The quality of due diligence on social safeguards quality is similar in pre-Performance Standard and Performance Standards during appraisal projects (table 2.4), the evaluation found that in IFC-supported projects has generally been IFC’s ESIA was satisfactory in the majority satisfactory. There has been a learning curve of real sector projects in both the pre- and in IFC’s appraisal of the new requirements on post-Performance Standard random sample Labor and Working Conditions (Performance for which an evaluative opinion was possible. Standard 2) and Community Health, Safety, and While in most cases, IFC found gaps between Security (Performance Standard 4), particularly the client’s initial environmental manage- for environmental specialists, which constitute ment provisions, national requirements, and four-fifths of CESI staff (the rest are social special- IFC’s Performance Standards, these had been ists). Social due diligence was especially strong in usually appropriately identified in the Correc- category-A projects that IEG visited for this evalua- tive Action Plans or Environmental and Social tion; IEG found that labor procedures conformed Action Plans that the client agreed to undertake with relevant requirements of Performance as a condition of IFC support. Standard 2, including establishing, maintain- ing, and improving employee-management EHS guidelines and indicators have not been relations; formation of workers’ organizations if adequately integrated with ESRD’s section on desired by employees; promoting fair treatment, Performance Standard 3 (Pollution Abatement nondiscrimination, and equal opportunities for and Control), and the annual monitoring workers; and ensuring compliance with national templates for the client lack production- labor and employment laws. However, there specific indicators. The introduction of Perfor- is a need to strengthen the templates given to mance Standard 3 and guidance on emission the clients at appraisal for annual monitoring by control and industry best practices in the including essential outcome indicators on health updates of EHS guidelines have the and safety.14 IFC due diligence was potential to improve environmental satisfactory in half of the appraisal in projects for processing Due diligence was also high on involuntary category-B projects that and manufacturing industries, which resettlement (Performance Standard 5), indige- the World Bank would form the majority of IFC’s real sector nous peoples (Performance Standard 7), and have categorized as A. investments. But the online ESRD cultural property (Performance Standard 8), system lacks performance indica- which occur less frequently in IFC’s portfo- tors for industrial pollution control, lio. Regarding land acquisition and involuntary and the AMR templates given to resettlement, 92 percent of the 12 projects in Clearer definition clients at appraisal for future annual the random sample for which an opinion was is needed on the reporting do not include important possible were rated satisfactory in terms of the environmental and production-specific indicators13 on identification of people to be displaced by the social review process air emissions, effluent discharges, project and those eligible for compensation and environmental water and energy conservation, and assistance through a baseline census with and social monitoring and waste management. The AMR appropriate socioeconomic data. The single system requirements for templates are not well tailored to unsatisfactory project was due to the absence of financial intermediaries. the project; they should focus on a Resettlement Action Plan. 24 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P The quality of due diligence was rated as satisfac- percent (24/39) of the real sector projects in the tory for the two category-A15 projects in the random sample. For the remaining 39 percent, sample, but only for half (3/6) of the category- either the only disclosure documented was in the B projects that the IBRD would have catego- WBG’s external document database (InfoShop) rized as A on the basis of the sensitive nature or no information on disclosure was found in the of their impacts. Where IFC’s due diligence was project documents. evaluated as partly unsatisfactory, the shortcom- ings relate to the inadequate coverage of risks Evaluation of Process Implementation: associated with the projects. MIGA Preparation and appraisal Clearer definition is needed on the environmen- Identification and screening. The Documentation on tal and social review process and ESMS require- evaluation of MIGA’s environmen- public disclosure and ments for FIs. IEG had earlier found performance tal and social screening and review consultation in IFC gaps, especially in FI projects appraised before process focused on the classification projects is relatively weak. the 2006 framework.16 Previously, for projects of projects, the quality of the ESIA, that received IFC corporate finance without and the extent of public consultation. The first direct IFC-financed subprojects, IFC focused on step is the classification of projects based on the the process of environmental and social manage- significance of their expected environmental ment in the institution rather than on specific and social impacts. In the evaluated sample, 11 subprojects. Under the post-Performance percent (4/35) of the projects had been classified Standard framework (ESRP 2009), where the as category A, 60 percent (21/35) as category B, portfolio review indicates that the FI’s invest- 14 percent (5/35) as category C, and 14 percent ments could have potentially significant environ- (5/35) as category FI. mental and social impact,17 the FI is obligated to ensure that its subprojects meet the relevant A major finding of the portfolio review is that elements of the Performance Standard in addition the MIGA/IFC and World Bank approaches to to applicable national environmental and social project classification differ from each other. In laws and regulations and exclusion lists. the evaluators’ judgment, this difference affects the classification of 17 percent (6/35) of the Thirty-three percent of 231 post-Performance projects in the MIGA sample and has various Standard FI projects were requested to apply origins. For two category-B projects that involved Performance Standard requirements for their the construction of major new facilities, the subprojects. In comparison, of 139 past XPSR magnitude of the impacts would have led IBRD projects validated by IEG, 46 percent were to classify them as category A. This was also the requested to apply Safeguard Policies and 29 judgment of the EBRD, which cofinanced one percent EHS guidelines. The level of these strict of these projects. In three additional category-B environmental and social requirements therefore projects, the Bank would have classified them remained about the same after introduction of as category A, based on the sensitive nature Performance Standards. Based on IEG evalua- of the impacts, that is, the fact that tion of 42 post-Performance Standard projects, they raised issues associated with The MIGA/IFC and World IFC’s decision to apply Performance Standards natural habitats, cultural resources, Bank approaches to for subprojects has broadly followed the rules set transboundary waters, retrenchment, project classification forth in the ERSP 2006–09, but these rules leave or tropical forests. In addition, the differ from each other. much room for interpretation. single pre-PPSSES financial interme- diary project in the sample had been classified Documentation on public disclosure and consul- as a C, in line with MIGA’s 1999 Environmental tation emerged as one of the weaker areas in Assessment Policy,18 but would have been put IFC’s due diligence. IEG found that informa- in the FI category by the Bank. Overall, since tion was available on local disclosure in only 61 only one of the six cases can be attributed to a 25 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D The preparation and specific difference in the language of documents, 38 percent (8/21) did not submit appraisal under the respective environmental assess- any ESIA, for which an adequate explanation was Performance Standards ment policies, these findings point provided in MIGA’s clearance memorandum, and has improved compared to a lack of consistency in safeguards 24 percent (5/21) provided unsatisfactory ESIAs with projects prepared implementation across the WBG. (all of which were from the pre-2007 sample). under safeguards policies. MIGA’s sustainability screening relies primarily Quality of environmental and on documents and information submitted by social impact assessment at appraisal. The the clients, which in turn reflect their corporate preparation and appraisal under Performance procedures and largely respond to the host Standards has improved compared with projects countries’ own requirements. The limited extent prepared under safeguards policies (table 2.6). to which full ESIAs were required is consistent Improvements were found in the appraisal of with the language of Performance Standard 1 projects’ SEMS, including for FI projects that that “depending on the type of project and the were previously categorized by MIGA as category nature and magnitude of its risks and impacts, C. MIGA also improved in the appraisal of labor the Assessment may comprise a… straight- and working conditions, following the introduc- forward application of environmental siting, tion of the PPSSES. Additionally, the Japan-MIGA pollution standards, design criteria, or construc- Trust Fund offered technical assistance to help tion standards.” clients in Africa enhance SEMS. The 24 percent of category-B projects that Across the entire sample, the evaluation found received environmental clearance with unsatis- that the quality of the ESIAs has been satisfactory factory EIAs are of concern. These shortcomings for category A, but only partially so for refer to sampled projects underwritten subject The 24 percent of category-B projects both in regard to to the pre-2007 safeguards system. Three of the category-B projects that environmental and social aspects. All cases involve agro-industrial projects for which received environmental four category-A projects in the sample the submitted EIAs only cover existing plants, clearance with submitted satisfactory ESIAs. Of the with no information on the new facilities as unsatisfactory EIAs category-B projects, 38 percent (8/21) well as the impacts of manifold expansion in are of concern. submitted satisfactory EIAs or similar crop production supported by the project. In Table 2.6: IEG Assessment of Process Implementation in MIGA Guarantees Environmental Identification and social Disclosure and Preparation and Client Quality of MIGA Overall results and screening appraisal consultation appraisal implementation monitoring Number of projects 35 25 22 35 10 8 Satisfactory (%) 88.6 60 81.8 71.4 40 62.5 Projects approved under safeguards Number of projects 21 19 17 21 9 7 Satisfactory (%) 81 47.4 76.5 52.4 33.3 57.1 Projects approved under performance standards Number of projects 14 6 5 14 1 1 Satisfactory (%) 100 100 100 100 N/A N/A Source: IEG. Note: Interpretation of results for indicators with small sample size should be treated with caution. 26 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P another case, a preliminary EIA was submit- by an environmental or social special- Public consultation ted, which a subsequent audit found to be ist. For the remainder, the entry-level emerged as one of the inadequate. In a further case, involving a solid review was limited to information weaker areas in MIGA’s waste treatment plant, while a due diligence available in documents provided by environmental and social mission had determined that the plant’s SEMS the client, plus their responses to assessment process. appeared to be satisfactory, no EIA had been follow-up questions. The evaluation submitted for verification and to serve as a found that 17 percent (6/35) of the projects basis for future monitoring and evaluation—an had been screened into a different safeguards important omission in light of the significance of category than the World Bank would have done, the potential impacts. using what are essentially the same criteria. While this is not a quality-at-entry issue per se, Consultation and disclosure. Public consul- it points to the need for improved safeguards tation emerged as one of the weaker areas in coordination across the WBG. Another issue MIGA’s environmental and social assessment arises from the exception on disclosure given process. This appears to be because MIGA’s 1999 to Small Investment Program projects (box 2.2), EA policy, which applied until 2007, only required resulting in an exemption from public scrutiny the sponsors of category-A projects to consult of ESRS and EIAs for a majority of category-B project-affected groups and local NGOs. There was projects. improvement among post-2007 projects. While all four of the category-A projects in the sample Quality of WBG Supervision had undertaken at least some minimal consulta- IEG assessed supervision quality through four tion with project-affected groups, only 24 percent indicators—the extent to which environmental (5/21) of category-B projects involved any form and social aspects of the project were addressed of public consultation. Two of these were projects by qualified staff or consultants, the accuracy of that the Bank would have classified as category A, for which consultations were required and/or sponsored by other financiers. The remaining 76 percent (16/21) of category B did not undertake Box 2.2: Spotlight on the Post-2007 Disclosure of a formal public consultation process and missed Environmental and Social Review Summaries (ESRS) out on the important opportunity to consult with affected communities on the projects’ environ- Before the 2007 policy reforms, MIGA’s disclosure of ESIAs had mental and social aspects that could potentially been limited to category-A projects. In its 2007 paper on the Draft affect them. PPSSES and Draft Policy on Disclosure of Information, MIGA com- mitted to disclose ESIAs, or at least the ESRS for both category-A Findings on MIGA preparation and and -B projects. However, an exception was made for Small Invest- appraisal. Overall, the implementation of ment Program projects, that is, projects with guarantee amounts MIGA’s environmental and social screening of under $10 million. This is a major loophole, since 67 percent (4/6) and appraisal has been only partially satisfac- of the category-B sample projects approved under the new policy tory. The portfolio review found that about 17 were Small Investment Program projects. As a result, out of six post- percent of the environmental assessments and Performance Standard category-B projects in the sample, only one most of the social assessments (relating mostly ESIA and not a single ESRS had been posted on MIGA’s website (as of to projects underwritten by MIGA before 2007) June 2009). Thus, the language of MIGA’s Disclosure Policy exempt- had been unsatisfactory, and only a third had ing the Small Investment Program projects from ESRS disclosure is some form of public consultation. MIGA’s very at cross purposes with the spirit of Performance Standard 1, which limited resources devoted to environmental and expects the client to provide communities that may be affected by social screening and appraisal appears to have risks or adverse impacts from projects with access to information been the major factor. Entry-level due diligence on the purpose, nature, and scale of the project, as well as any risks of only 27 percent (7/25) of category-A and -B and potential impacts. projects in the sample was based on a site visit 27 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D supervision ratings on safeguards/Performance cant differences between high and low perform- Standards, and the quality of monitoring and ers (figure 2.2).19 As in project preparation, East evaluation (M&E) of these issues. Candor in ratings Asia and Pacific has been the best performer and M&E were given more weight than supervi- on supervision quality, with most other regions sion by qualified specialists when determining the lagging significantly behind. Latin America and the overall rating for WBG supervision. Caribbean (LCR) and the Middle East and North Africa (MNA) were found to have overly optimis- Findings on quality of bank supervision and tic safeguard ratings compared with the evidence monitoring presented in the respective ISRs and aides Quality of supervision was assessed in terms of memoirs. Among the networks with substantial the supervision effort invested in following up safeguard issues identified within projects, the on the mitigation measures and action plans Human Development Network lagged signifi- prepared to address any safeguard policies cantly behind the others, both on supervision triggered. Policies could be triggered by individ- quality and on M&E (annex 1, table X1.12). ual projects, the composition of the supervision team, especially with reference to the deployment There is a substantial difference in supervision of staff or consultants with relevant skills, and the quality by project classification, category A receiv- appropriateness and supporting evidence for the ing much more attention on environmental and safeguard ratings in the available ISRs social safeguards than category B (figure 2.3). Consolidated results and related aides memoirs. In the final While this reflects better attention to high-risk on supervision instance, these results were compared projects, it does not follow that all category-A show considerable with the quality of M&E of safeguards projects have to follow the safeguards design variation by region. relevant to the project. approved at appraisal. Some projects have done an excellent job of adaptive learning to modify The consolidated results on supervision show the safeguards design when the project context considerable variation by region, with signifi- changed (box 2.3), lowering safeguards costs appropriately. While this reflects better attention to high-risk Figure 2.2: Supervision of Safeguards in projects, IEG findings support the concern Bank-Financed Projects, by Region expressed by environmental and social staff and management that category-B projects, most of Total which are delegated to respective sectors, are not being adequately supervised and monitored SAR in one-third to one-half of the projects where MNA safeguards are triggered. Delegation of projects to the sectors is thus having a perverse effect LCR of leaving safeguards aspects of a large number ECA of projects unsupervised, raising the likelihood that the next generation of reputational risks EAP could arise from low-risk projects financed by AFR the Bank. 0% 20% 40% 60% 80% 100% Percent of projects satisfactory The increasing reliance in World Bank projects on policy frameworks is a cause for concern Effectiveness of safeguards M&E Appropriateness of ISR safeguards ratings Quality of Bank supervision because these projects include multiple subproj- Source: IEG Portfolio Review, fiscal 1999–2008 approvals. ects and are less well supervised than projects Note: Satisfactory refers to ratings of satisfactory or excellent from the portfolio review by IEG reviewers. with proper risk assessments and mitigation plans. About one-third of the projects that trigger 28 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P environmental or social safeguards rely on policy Figure 2.3: Supervision of Safeguards in frameworks to address impacts that are not fully Bank-Financed Projects, by Safeguard Category known or assessed at appraisal. As mentioned previously regarding resettlement (box 2.1), IEG found projects with policy frameworks in the Total portfolio review to be less well supervised than those with RAPs, giving rise to potential reputa- FI tional risks if adverse impacts are not addressed adequately by clients during implementation. B IEG found similar differences in performance of projects with ESMFs. Supervision results perfor- A mance in terms of mitigation are both lower for Bank projects that rely on policy frameworks, 0% 20% 40% 60% 80% 100% whose environmental and social impacts cannot Percent of projects satisfactory be assessed at appraisal (figure 2.4). If the Bank Effectiveness of safeguards M&E Appropriateness of ISR safeguards ratings Quality of Bank supervision relies on Resettlement Policy Frameworks during preparation it needs to invest proportionately Source: IEG Portfolio Review, fiscal 1999–2008 approvals. greater resources in supervising these projects Note: Satisfactory refers to ratings of satisfactory or excellent from the portfolio review by IEG reviewers. to help the client implement them well. Policy frameworks can be powerful instruments to empower local communities and involve by design have stronger participatory The increasing reliance them in environmental management decisions. processes for project implementa- on policy frameworks Policy frameworks are frequently used by tion that can also be mobilized for is a cause for concern community-driven development projects, which safeguards implementation (box 2.4). because these projects IEG did not find any evidence of use are less well supervised of social audits or participatory M&E than projects with proper to monitor safeguards results. These risks assessments and Box 2.3: Adaptive Management on community monitoring mechanisms mitigation plans. Safeguards in Project Restructuring are employed to monitor project outputs and outcomes in an increasing Projects implemented by A Bank-financed power project in Asia initially ad- number of community-driven develop- financial intermediaries opted a sectorwide approach and triggered eight ment projects from agriculture, human are also a cause safeguard policies at appraisal. The project was re- development, and social development for concern. structured midway when it became apparent that the sectors, and as key instruments to largest component—attracting private investments monitor governance outcomes in a wider range of for subprojects in the sector—was no longer viable sectors. The use of these modalities of community due to conflict in the country. Project resources that monitoring could also strengthen safeguard had been intended for the component were reallo- implementation, particularly for projects whose cated to the remaining two components, which were dispersed nature precludes effective monitoring working successfully. When the largest component of all subprojects by the client. was dropped, the Bank and the client also agreed that only three safeguard policies (environment, re- Projects implemented by financial intermedi- settlement, and natural habitats) were applicable, aries are also a cause for concern, especially effectively restructuring the safeguards design. This because many of these rely on ESMFs during was a good example of adaptive management by appraisal on the assumption that the financial Bank staff working with the client in a fluid political intermediaries will be undertaking or commis- and security context. sioning environmental and social assessments during implementation.20 Without adequate 29 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 2.4: Projects with Policy Frameworks versus Projects with Mitigation Plans Supervision of Bank projects, FY1999–2008 Mitigation of adverse impacts in Bank projects FY1999–2008 100% 100% Projects with satisfactory supervision Projects with satisfactory mitigation 100% 91% 80% 80% 75% 71% 64% 58% 59% 59% 61% 60% 60% 56% 56% 51% 40% 40% 20% 20% 0% 0% Environment Resettlement Indigenous people Environment Resettlement Indigenous people Framework Plan Framework Plan Source: IEG Portfolio Review. Note: This chart compares the ratings of projects with an EA/EIA with those that relied on an Environmental Management Framework or Environmental and Social Management Framework. For resettlement it compares ratings of projects with a RAP with those that relied on a Resettlement Policy Framework or an ESMF, and for indigenous peoples it compares ratings of projects with an IPP with those that relied on an Indigenous Peoples Planning Framework or an ESMF. IEG found safeguards M&E supervision of these ESMFs, the and aides memoirs calls for urgent manage- to be the weakest aspect Bank cannot be confident about the ment attention. of Bank supervision. client’s due diligence and safeguard results. Even though the FI sample IEG found safeguards M&E the weakest aspect was small, the fact that only 50 percent of FI of Bank supervision. Except for resettlement operations had safeguard ratings that could be monitoring, which was of high quality in East supported by evidence presented in the ISRs Asia and Pacific and in South Asia but weaker Box 2.4: Local Implementation of an ESMF in a Community-Based Project in Africa A community-based rural development program in West Africa beneficiaries, who were included in the program’s M&E system. (category B), which financed over 14,000 microprojects, applied The guiding principle of the screening system was that effec- an Environmental and Social Management Framework (ESMF) tive mitigation of potential negative environmental effects would to subprojects at the village level. Fifty percent of project in- ensure the long-term sustainability of infrastructure investments vestment was directed toward small rural works projects. The since beneficiaries were expected to provide a significant level ESMF helped the project team classify all subprojects into two of in-kind contribution. This required a high level of village aware- categories: those with positive or minor negative environmental ness and capacity and increased safeguard technical assistance impacts (such as forest management), and those likely to have and supervision costs. However, with this assistance, despite a more negative environmental impacts (feeder roads through slow start the system was fully deployed and extended into the forests, small-scale dams, and rangeland management, for second phase of project implementation. Site visits confirmed example). Safeguards were identified and mitigation measures that the system had been used appropriately to mitigate negative designed in consultation with local village land committees and environmental effects for almost all subproject types. Source: IEG Field Study 30 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P in other regions, more than one-third of Bank supervision. But uneven quality of Overall, 73 percent of projects suffered from inadequate M&E. The staff, and inconsistencies in interpreta- NGOs responding to the weaknesses lie in lack of specificity of monitoring tion of policy was a more widespread IEG survey rated Bank indicators, underinvestment in client’s monitor- issue (appendix E). performance better ing capacity and poor follow-up during supervi- than in the 1990s, as sion. Safeguards monitoring will not be effective On the other hand, IEG’s NGO survey compared with 10 percent until safeguards indicators are integrated within and client survey revealed a positive that rated it lower. the overall results framework of the project. perception of Bank project perfor- Too often, safeguards activities are considered mance. Overall, 73 percent of NGOs responding an add-on, and left to environmental and social to the IEG survey rated Bank performance better specialists who are underresourced and not well than in the 1990s, compared with 10 percent who integrated into supervision teams. This is not rated it lower; and two-thirds of NGO respondents simply a resource constraint. Staff surveys reveal ranked safeguard performance of Bank-financed that task team leaders complain about unavail- projects better than projects financed by other ability of environmental and social specialists donors or by the countries themselves, compared when they are needed (box 2.5), and environ- with 15 percent who rated the Bank lower (annex mental and social specialists complain about the 2, table X2.2). An overwhelming majority of lack of predictability of demand for their services. clients rated Bank performance “better than in Matching skills to demand cannot be left to the 1990s.” Fifty-two percent rated Bank perfor- the labor market and will require management mance “much better,” while 21 percent rated it attention and up-front commitment of staff and “somewhat better” than projects financed by resources as an integral part of work program other institutions (annex 2, table X2.3). planning, if this constraint is to be overcome. Triangulation of bank findings on preparation and supervision Box 2.5: Staff Views on Resources for Safeguards The relatively better results for Bank-supported projects during preparation compared with Nearly a third of Bank task team leaders and environmental and social supervision were confirmed by data from other staff complain about inadequate resources to address safeguard issues. sources. IEG found that the current satisfaction level of Bank managers with the adequacy of • Twenty-two percent of task team leaders report inadequate support environmental safeguards in project prepara- from both environmental and social specialists, which increased to tion is 76 percent, compared with 52 percent 26 percent for social during supervision. during supervision. The same managers rated • Task team leader feedback on support from environmental spe- satisfaction with social safeguards about 10 cialists was slightly more positive than feedback on support from percent lower: 65 percent during project social specialists, and support was better during preparation than preparation and 41 percent during supervision. supervision (74 and 67 percent, respectively, for environment, and Environmental and social sector managers and 68 and 59 percent, respectively, for social). regional safeguard advisors were even more • While two-thirds of task team leaders found the task budget to critical of the quality of environmental and be adequate, 28 percent found the task budget to be low during social supervision than were country directors. supervision and 22 percent during preparation. Among the regions, Lack of incentives appears to be an even bigger task team leaders from the Africa Region were most negative (37 constraint than resources (box 2.6). percent “inadequate”) about the adequacy of effort and resources for supervision. The survey of task team leaders indicates that • A third of the environmental and social specialists felt the level of capacity issues within the Bank are more acute effort and resources during supervision was too low. in Africa than in other regions. Team leaders from that region were most negative (37 percent Source: IEG Staff Survey inadequate) about the effort and resources for 31 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Box 2.6: Incentives Are a Constraint on Safeguard Effectiveness in Bank Projects Country directors: • “[Supervision] could and should be done by a third party • “The Bank itself lacks capacity. We were always scrambling or the client, with independent monitoring to ensure for someone to come and complete the safeguards work objectivity.” on time.” • “Senior social safeguards staff are in short supply and can- Sector managers: not always handle the more complex issues. Some were • “Safeguards work is being delayed from preparation to im- former NGO liaison staff and haven’t really got the right plementation through increasing use of policy frameworks. background and training.” Project preparation is easier for programmatic lending [with] • “Delegation [of B and C projects] to sector managers means policy frameworks, but following through on the need for that no attention is being paid to their quality because they greater supervision is more difficult.” lack the skills to deal with safeguards. This is where there • “How do we transfer capacity? The quickest way is to do it are more problems.” yourself instead of relying on the client or focusing on the • “The incentives are not there. Nobody wants to work on long term. With the layering and fear in the World Bank, safeguards.” capacity is not a priority. Reputational risk is the primary • “Fiduciary people (financial management and procurement) concern.” are paid off the top and deploy staff as needed by task teams. • “Staff promotions are slanted toward own-managed projects The same should happen for environmental and social safe- more than toward providing safeguard services.” guards work.” Source: IEG Interviews with country directors and sector managers. Monitoring and supervision at IFC has improved. IEG has evaluated IFC’s environ- The portfolio review reveals that IFC’s supervi- mental and social quality since 2004 as a part of sion quality has been lower than appraisal the XPSR validation program. As shown in figure quality. IFC’s monitoring of project performance 2.5, IFC’s environmental and social supervision focuses on two aspects: (i) compliance with the quality for pre-Performance Standard projects safeguard policies and Performance Standards, has improved since 2007, with an increased project-specific environmental and social number21 of environmental specialists supervis- requirements, and applicable industry ing FI projects, but environmental and social The portfolio review guidelines, and (ii) client implementa- supervision quality is still below the real sector reveals that IFC’s tion of the Environmental and Social level, due to fewer staff resources devoted to the supervision quality has Action Plans agreed to as a condition FI sector (5 environmental and social special- been lower than appraisal of project approval. With increas- ists) compared with non-FI sector (57 environ- quality and lags behind ing resources, FI supervision quality mental and social specialists). However, IFC has real sector supervision. has improved but still lags behind developed rules for project supervision and site real sector supervision. Supervision visit efforts and its overall “knowledge gap”22 has of client implementation of actions plans and decreased from 12.5 percent in fiscal 2008 to 5.8 reporting needs to improve. percent in fiscal 2010. Based on IEG’s validation of XPSRs, Clients’ annual reporting has been a challenge Client’s annual reporting the quality of environmental and for IFC’s supervision. The staff survey reveals has been a challenge social supervision after fiscal 2007 for that about 30 percent of investment officers for IFC’s supervision. pre-Performance Standard projects and environmental and social specialists felt 32 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P the timeliness and quality of client monitor- Figure 2.5: IFC’s Environmental and Social Work ing was inadequate. Within IFC’s sustainabil- Quality at Supervision for Pre-Performance ity framework, the clients’ AMRs and IFC site Standard Projects visits are the main instruments for monitor- ing the projects’ performance. Since clients’ 100% first AMR is only due six months following the first year of project approval, the post-Perfor- 90% mance Standard portfolio review focused on Satisfactory rate 80% projects that had been approved at least two years earlier. Of the 28 random sample projects, 70% including all pre-Performance Standard and 60% post-Performance Standard real sector projects older than two years, only 50 percent (14/28) 50% provided IFC with satisfactory AMRs.23 In most 0% such cases, IFC identified the deficient informa- 2004–06 2005–07 2006–08 2007–09 tion in the AMR for correction in the following Fiscal Year year, but in many cases the deficiencies contin- All projects Non FI FI ued despite IFC corrective actions, reflecting Source: IEG’s Environmental and Social Reviews for fiscal 2004–09 XPSRs, 209 real sector (non-FI) insufficient communication and frequency of and 139 FI projects. IFC feedback, and poor client intake of correc- Note: Data are based on three-year rolling averages of IEG annual XPSR evaluations. By presenting tive requirements. the rolling average data, IEG; (i) meets the test of less than 5 percent sampling error, with 95 percent confidence level, when a 50+ percent stratified random sample from mature projects is sampled for annual XPSR evaluations, and (ii) sometimes large annual deviations in time series are eliminated. The portfolio review also found that IFC had not monitored the implementation of the Environ- mental and Social Action Plans in 21 percent (6/28) of the projects older than two years. Since the Environmental and Social Action Plans are designed to remedy gaps in the client’s social Box 2.7: IFC Staff Survey Feedback on Resources and environmental management system identi- for Environmental and Social Work fied during appraisal, they represent a major part of value IFC adds to the project. Without • Over half of the investment officers and a third of the environmental IFC monitoring of implementation, it cannot be and social staff felt that the share of resources used for interactions assumed that this value was added. with external community/NGOs was inadequate. • Investment officers were much more critical than environmen- Within the overall trends presented in figure 2.5, tal and social specialists of the resources devoted to work on IEG has not found significant differences in the safeguards/Performance Standards: 40 percent reported the task supervision quality of post-Performance Standard budget to be low during supervision, compared with 29 percent FI projects compared with the pre-Performance during preparation. Less than 10 percent of environmental and Standard ones. Supervision of environmen- social specialists complained about the task budget. tal safeguards and Performance Standards (3, • More than a quarter of the investment officers complained about 6, and 8) has been at the same level as social the timely availability of environmental and social specialists and safeguards and Performance Standards (2, 4, 5, the lack of integration of environmental and social specialists into 7, and 8); some deficiencies found in addressing regular supervision missions. the new social standards (2 and 4) are discussed • Two-thirds or more of IFC staff agree that implementation of Per- in chapter 3. formance Standards needs more resources compared with the pre-Performance Standard systems. IEG interviews with IFC environmental and social specialists confirmed that despite improve- Source: IFC staff survey. ments, resource constraints have not been 33 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D MIGA monitors overcome (appendix E). Environ- of issues requiring remedies suggests more category-A projects mental and social specialists also supervision of category-B and -FI projects is also more systematically, but reported difficulties in fully accurate warranted. Under its sustainability policy, MIGA the frequency of issues and reliable reporting on social is expected to monitor projects’ performance requiring remedies aspects of Performance Standards. against the applicable Performance Standards suggests more supervision IFC has a shortage of social special- after project approval by requesting periodic of category-B and -FI ists to fully assess and ensure client monitoring reports on the clients’ environ- projects is also warranted. performance on such sensitive issues mental and social performance, conducting as child labor, freedom of associa- site visits of selected projects, working with the tion, or discrimination. Generally, IFC clients to address adverse impacts if they occur, relies on the site visits and AMRs, prepared by and exercising remedies as appropriate. The the client, who are not candid on these issues portfolio review found that MIGA carried out 12 (also see box 2.8). monitoring missions covering all four category- A projects and 33 percent (4/12) of category-B Monitoring and supervision at MIGA projects in the sample that had been effective MIGA monitors performance of category-A for longer than two years. This is consistent projects more systematically, but the frequency with a risk-based approach that allocates greater Box 2.8: IFC’s New Social Performance Standards IEG found from the portfolio evaluation that most of the ap- management (Performance Standard 2), and cultural heritage praisal and supervision indicators regarding the five social (Performance Standard 8). Performance Standards (Labor and Working Conditions; Com- Based on IEG interviews, social Performance Standards are munity Health, Safety, and Security; Land Acquisition and perceived by both clients and IFC staff to support IFC’s devel- Involuntary Resettlement; Indigenous Peoples; and Cultural opment dialogue and mission. The two Performance Standards Heritage) achieved reasonably high scores. IFC’s performance unique to the IFC, numbers 2 and 4, are oriented toward enhancing in mitigating negative and, especially, enhancing positive so- human welfare, both inside and outside of the borrowing entity. cial impacts was better, or at least at the same level, com- However, there is still insufficient understanding and depth of pared with the earlier safeguard period. For example, IFC knowledge on how some details in social Performance Standards persuaded the client in a large category-A pipeline project in should be incorporated in the dialogue with the clients. While Latin America and the Caribbean to engage with the fisher- this may be true to some extent for all five social standards, it is men near the port facility. Compensation, labor, and health particularly relevant for Performance Standard 2, especially re- and safety issues, security management plan, and cultural garding freedom of association, discrimination in hiring practices, heritage plan were other major improvements that contributed and child and forced labor. These sensitive cultural issues do not to better project management procedures. Key observations lend themselves to transparency on the part of the borrower. Key of IEG’s site visit included excellent monitoring programs, IFC staff members have expressed awareness of this issue and including locally hired supervisors to monitor environmental are taking steps to remedy it through community surveys and and social performance, good health and safety statistics and discussions with local worker union organizations. IEG found practices, active community engagement, and careful archeo- that it is important to raise staff awareness on social aspects and logical rescue of identified artifacts. The client regarded the promote systematic training on sound monitoring techniques, social Performance Standards as adequate and fair, covering with illustrations from real-life cases, increased resources and environmental and social aspects that otherwise might have staff time for proper social supervision and site visits, and more been unnoticed, especially contractor management, resettle- in-depth, qualitative data collection and monitoring techniques to ment and compensation (Performance Standard 5), security better understand the social aspects of projects. Source: IEG portfolio review. 34 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P supervision resources to category-A projects and resources devoted to these functions. Monitoring and high-risk category-B projects. High-risk issues While the portfolio review concluded supervision of the (mostly related to resettlement) were found in that these functions are managed environmental and social 75 percent of the projects that MIGA visited, efficiently, in the sense of focusing on performance of MIGA including all category-A projects. While reassur- what are expected to be the highest- guarantees are seriously ing from a cost-effectiveness perspective, the risk projects, the high incidence of constrained by limited high incidence of issues found in the monitored problems identified (and addressed) capacity and resources. projects raises question about potential adverse by the very limited number of environmental and social impacts that may have monitoring missions points to the likelihood been missed. The concern arises because of the of a large number of environmental, health, large number of category-B and -FI projects that safety, and social risks that remain unidentified have not been monitored at all through either a and unaddressed in the large share of projects systematic review of and follow-up on periodic that have never been monitored. These hidden monitoring reports or field visits. risks point to a major gap in MIGA’s sustainability framework and represent a missed opportunity Active monitoring and follow-up can help address to help clients and enhance MIGA’s develop- social impact issues even in a high-risk environ- mental contribution. In IEG’s judgment, it is not ment. A mining project in a conflict-afflicted feasible for MIGA to fully meet the expectations country had attracted much controversy because of the Performance Standards under the PPSSES of the host-country armed forces’ use of the unless its environmental and social capacity is client’s plane, trucks, and drivers to suppress a substantially increased. local uprising. A CAO field audit found that, while MIGA’s initial environmental and social screening Accountability Mechanisms had been adequate, its follow-through on social The Inspection Panel (IPN) and the Compliance aspects had been weak, perhaps because of the Advisor and Ombudsman (CAO) are key account- absence of in-house social expertise. MIGA had ability mechanisms25 created for the World Bank expected the client to adhere to the Voluntary and IFC/MIGA, respectively, to assess complaints Principles on Security Forces and Human Rights from affected parties who feel that they are without assessing whether the client had the ability being harmed by the WBG’s failure to properly to do so.24 Following MIGA’s hiring of an in-house implement its policies. The roles of the IPN social scientist who visited the project, the client and CAO are to hold the WBG accountable for developed a protocol governing its interactions compliance with all WBG Operational Policies. with the host-country’s armed forces and obtained Their roles are not restricted to oversight of the the government’s signature. The acceptance and safeguards policies and Performance Standards dissemination of the protocol by local communi- only. The IPN reports to the Board, while the CAO ties and officials was assisted by training programs reports to the president of the WBG. In practice, held at three of the client’s mining sites, funded the CAO has maintained its independence from by the Japan-MIGA Technical Assistance Trust the line management of the IFC and MIGA. Fund. MIGA also assisted the project to design and implement a comprehensive community develop- Since 1995 when it was first established, the ment program. Follow-up monitoring suggests IPN has received 64 requests for inspection. IPN that this protocol has set a good example for other procedures require affected people to approach mining investors in the country. management first before complaints are considered for registration. Fifty- Environmental policies MIGA’s quality of monitoring and supervision seven of the complaints received by had more complaints Overall, monitoring and supervision of the the IPN were registered for review and violations but social environmental and social performance of and 31 (48 percent) recommended safeguards appear to the projects MIGA guarantees are seriously for investigation. Twenty-six of the be more challenging constrained by the limited capacity and requests were received between fiscal and contentious. 35 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Less than half of the complaints filed with the Table 2.7: Inspection Panel Complaints Filed and IPN were about the safeguard policies. Accord- Policy Violations, 1995–2009 ing to IPN annual reports these 64 requests contained 307 specific complaints, 138 (45 Number of Number of percent) of which related to the 10 safeguard complaints violations policies, while the rest were related to the Bank policy filed found Bank’s 25 other policies. Of those relating to Safeguard policies safeguards, 60 percent concerned environmen- tal safeguards and 40 percent were about social (Env) Environmental assessment 40 21 safeguards. The IPN found 110 policy violations, (Soc) Involuntary resettlement 32 17 63 of which were against safeguard policies: (Soc) Indigenous peoples 23 10 36 policy violations of environmental and 27 (Env) Natural habitats 17 7 of the social policies. Adjusting for the smaller proportion of projects that trigger the social (Env) Physical cultural property 11 6 safeguards, these policies received relatively (Env) Forestry 9 1 more complaints and policy violations were (Env) Pest management 2 1 more frequent, affirming the relatively greater (Env) Dam safety 3 0 challenges faced by projects in complying with the current social safeguards. (Legal) International waterways 1 0 (Legal) Disputed areas 0 0 The majority of safeguards violations in Bank Subtotal 138 63 projects were related to design issues stemming Other Bank policies from inadequate assessment of environmental Project supervision 49 15 or community impacts, and inadequate consulta- tion with affected people, which had still not been Information disclosure 24 10 resolved. One-fifth of deficiencies arose from Poverty reduction 23 8 inadequate attention to additional implementa- Economic evaluation of 12 8 tion issues. Supervision provides an opportunity investment ops to deal with unanticipated risks that arise during Project appraisal 6 6 implementation. Careful attention to complaints and resolution of grievances can decrease the Gender and development 1 0 demands on accountability mechanisms and also Other (various) 54 0 ensure that grievances are redressed. The IPN Subtotal 115 47 also found policy violations in all six complaints Total 307 110 filed against OMS 2.20, Project Appraisal, which Source: IEG compilation from Inspection Panel cases. provides guidance on assessing social issues during project appraisal (table 2.7).26 After the IPN findings are accepted by the 1995 and 2002 and 38 requests in the period Board, follow-on activities are between the fiscal 2003–10. Nearly 85 percent of the requests Board and Bank management. The action came from the Africa, Latin America and the plans and implementation progress reports Caribbean, and South Asia regions, with requests that IEG reviewed varied broadly in quality for inspection received from 34 countries. The and scope: from precise actions, with defini- highest number of requests among countries was tive schedules, and institutional responsibilities from India (8), and among the sectors from the to more qualitative statements of intent, with Energy and Mining sector (16). There has been a no clearly defined end and unclear linkages flurry of recent requests from Latin America and between action plans and implementation the Caribbean. progress reports. Progress reporting of IPN 36 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P interventions is more open-ended than for CAO interventions. In the latter case, probably Box 2.9: Manager Comments on the Inspection because of the CAO’s continued involvement in Panel monitoring implementation, there appears to be greater attention to the objective of reaching • “The Panel case on a pipeline project led to major benefits because and reporting closure. it substantially raised the level of government awareness and politi- cal will to resolve any future environmental and social safeguards IPN/CAO investigations do not come without issues. There is now substantive dialogue around how to build costs. The investigative stages of the IPN/CAO capacity in government.” work invariably place substantial demands on • “[The Bank should] be proud of its safeguard work in most cases WBG staff, and the implementation of any agreed because it is extremely rare to encounter self-correcting actions actions resulting from their investigation leads to in other organizations… But the Panel has resulted in a culture of remedial actions, largely paid for by the client. In fear and paralysis at the Bank.” some instances the IPN complaints also affected • “There is always a huge stress factor with an Inspection Panel relationships with the clients. case, because of the amount of information that has to be as- sembled and trawled through. There would be merit in some sort of However, independent investigations can arbitrage mechanism as many times the matter could be resolved identify systemic policy gaps. The CAO identified through negotiation and this would be much cheaper and more a systemic policy gap regarding the application effective.” of Performance Standards to supply chains of • “In some ways the Panel has provided some constructive feedback. primary clients; this issue is now being addressed But they are very legalistic and unrealistic… looking at process, by IFC management. IPN investigations in two not really at outcomes. [The Bank should] keep the Panel for seri- forestry projects surfaced systemic weaknesses ous cases, but if you have smaller complaints that need redress, a in dealing with livelihood interests of forest- conflict resolution mechanism would be better.” dependent communities, leading to lessons Source: IEG interviews with country directors and sector managers. learned for future natural resources management projects. Similarly, the IPN identified systemic gaps in appraisal of social issues due to inatten- tion of relevant provisions of OMS 2.20. the IPN has received a total of 64 complaints in Managers acknowledge the benefits of having 15 years. Distributing estimated costs over the an accountability mechanism, but believe that entire portfolio yields an incremental expendi- the creation of a more transparent grievance ture in the 15-year portfolio of about $9,000 per redress mechanism or ombudsman through project, an expense that serves to demonstrate which complainants can seek management accountability within a public institution. The intervention is overdue (box 2.9). The biggest costs of the IPN or those incurred by clients in concern staff and managers expressed was that responding to these complaints are additional. the intensive scrutiny involved with IPN cases Since a significant proportion of the complaints had huge opportunity costs since prolonged received do not go to full investigation, staff and investigations take a toll on project implementa- managers feel that a lower-cost grievance redress tion. IEG interviews with managers revealed that mechanism could help resolve some of the additional costs in responding to IPN complaints complaints of affected people. Although the IPN ranged from $120,000–$1,000,000 in manage- process requires complainants to first approach ment and staff time. Staff costs on projects that management, the Bank does not yet have a go to full investigation are estimated to be over system for receiving or resolving such $500,000, but even the preliminary manage- complaints and continues to deal with The Bank does not have ment response for cases that do not go to a full such issues on an ad hoc basis, or a system for receiving investigation tend to cost over $100,000. While when complaints have already been or resolving complaints these costs are not trivial, it should be noted that filed with the IPN. from affected people. 37 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Accountability mechanisms have yielded tion, and monitoring of follow-up actions ensures important benefits, some in the form of redress achievement of results. Although the CAO for aggrieved parties, others less quantifiable appears to have maintained its independence but no less significant. As always, costs need with respect to IFC and MIGA management, it to be compared with benefits. These include reports to the president of the WBG. Institutional the management of reputational risk—since credibility and its external perception would be the WBG is now able to demonstrate that it is enhanced through an independent compliance publicly accountable—the ability to take correc- review process, which ensures that the CAO tive action against legitimate complaints, and the submits its audits directly to the Board. ability to identify systemic policy issues that may need additional guidance or clarification. In the The Bank does not have a formal grievance opinion of some experts from academia and the redress system or a conflict resolution NGO sector27 these accountability mechanisms mechanism, although more recently the IPN has have helped people to recognize that the costs of been providing management with opportuni- noncompliance are actually significantly higher ties to solve problems brought to their attention than the costs of compliance. And the existence by affected persons. In recent IPN cases (for of a mediation mechanism at CAO enables half example, on projects in Democratic Republic of of the complaints against IFC to be resolved and Congo, Panama, and Yemen) the IPN deferred benefits transferred to aggrieved parties without investigation on a case-by-case basis to provide the need for eligibility studies or investigations. management with an opportunity to resolve complaints received before deciding on whether The existence of grievance redress mechanisms a full investigation was justified. The formalization increases the efficiency of accountability of a transparent grievance redress and conflict mechanisms. As noted in a recent IPN publica- resolution mechanism for affected people to tion,28 some of the relatively new or restructured approach management, or the extension of accountability mechanisms established by other the terms of reference of the Bank’s Ombuds- IFIs and development agencies following the man to review complaints from people affected Bank’s example contain features that the IPN by projects, may increase the efficiency of the lacks, such as mediation or grievance redress accountability process. mechanisms and the authority to monitor actions following an investigation. Both of these features Summary are present in the CAO structure as well as among The focus of this chapter has been on prepara- the accountability mechanisms in some other tion and supervision quality to assess the extent multilateral development banks (such as EBRD of compliance with safeguard policies. IEG and the Asian Development Bank). Since its found inconsistency between the categoriza- inception in 1999, the CAO has been responsible tion of projects at IFC/MIGA and those at the for 126 complaints and requests for audits. While Bank. The quality of preparation in the WBG is 40 percent were found ineligible for investiga- much better than the quality of supervision, this tion, 32 percent were settled after the Ombuds- difference being more pronounced at the Bank man assessment, and another 10 percent were where policies and procedures are heavily front- closed after the Ombudsman assessment and loaded. M&E of safeguards is the weakest aspect compliance appraisal.29 Since 11 percent are still of Bank supervision, followed by lack of candor in the process of an Ombudsman assessment, in supervision reporting. Supervision quality at less than 10 percent of the complaints received the IFC has improved since the introduction of went on to a compliance audit. an online ESRD system with key environmental and social performance indicators, but client The CAO experience suggests that the use of an reporting through the AMRs needs to improve. Ombudsman function considerably reduces the The Bank is constrained by lack of effective proportion of projects that need a full investiga- instruments and clearly specified indicators, 38 P R O C E S S I M P L E M E N TAT I O N B Y T H E W O R L D B A N K G R O U P uneven supervision capacity across the regions, policy frameworks and to category- The existence of grievance and underinvestment in client systems to B projects, which have mostly been redress mechanisms collect monitoring data. MIGA’s supervision is delegated to sector management units increases the efficiency constrained by limited capacity and resources. which lack the capacity and incentives of accountability to supervise them. IFC does not use mechanisms. Despite some recent improvements in IFC policy frameworks but agribusiness supervision, IEG found WBG procedures and supply chains have similar vulner- practice to address environmental and social abilities. Given the dispersed nature of most risks of projects with multiple subprojects to be of these projects, the WBG needs to develop inadequate. This applies to financial interme- alternate mechanisms for third-party monitoring diaries across the WBG and, in the Bank, and increased transparency to ensure effective to projects with multiple subprojects using supervision and implementation. 39 Chapter 3 Evaluation Essentials • The WBG does not have a clear framework to assess the impacts of its safeguards and sustainability policies. • Environmental and Social Effects are one of four dimensions used to evaluate project development out- comes of IFC and MIGA projects but not of Bank projects. • Implementation of the IFC/MIGA Performance Standards is too re- cent to evaluate outcomes. • Strengthening the commitment and capacity of the FI sector clients re- mains a challenge. A women’s group participating in an environmental project in Kenya. Photo by Curt Carnemark, courtesy of the World Bank Photo Library. Environmental and Social Performance Analytical Framework corporate standards that the client would be The main focus of this chapter is the question: normally expected to follow in the absence of “To what extent have safeguards and sustainabil- WBG involvement. Safeguards and Performance ity policies led to improved environmental and Standards, and also EHS guidelines in the case of social performance and impacts at the project IFC, are triggered when projects are expected to and sector level?” The chapter goes beyond result in adverse environmental or social effects. compliance to look at implementation perfor- Thus all projects are likely to involve some mance. This task is particularly challenging in the measures to mitigate these negative effects. World Bank because the frameworks themselves In addition, the project design may include do not provide clear guidance on how to identify measures to promote environmental and social and monitor performance. And in the absence sustainability. Typically, these could consist of of clear indicators and baseline data, measure- efforts to strengthen the client’s systems and ment of performance in most projects is limited institutional capacity in a manner that outlasts or nonexistent. the project, or environmental and social actions that go beyond the minimum requirements for In order to compare implementation perfor- mitigation specified in the policy. mance of safeguards and sustainability policies across the variety of projects found in the IEG rated a project’s MNI satisfactory when the WBG’s portfolio, IEG assessed each project in environmental and social design was appropri- the portfolio sample against three indicators— ate and when supervision and Implementation mitigating negative impacts (MNI), enhancing Completion Report (ICR)/Expanded Project positive impacts (EPI), and strengthening client Supervision Report (XPSR) documentation capacity (SCC). 1 The WBG’s safeguards and provided evidence that agreed mitigation actions Performance Standards aim to benefit its public had been successfully completed and safeguards and private sector clients through better manage- objectives achieved or, for projects under ment of environmental, social, health, and safety implementation, were well advanced at the time risks; improved community and government of the assessment. For the most part, satisfac- relations; and access to concessional funding. tory ratings were based on IEG’s assessment These benefits arise from the clients’ adoption of effective implementation by the client and of the environmental and social standards, the satisfactory supervision. In a few cases, implemen- scope and reporting requirements of which are tation weaknesses identified early were overcome often broader and more stringent than those during supervision, resulting in satisfactory MNI of national laws and regulations, as well as the ratings. Projects that did not contain supporting 43 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D evidence in supervision or completion reports 2009 and therefore had not used these indica- were downgraded even if the supervision reports tors, but IEG evaluates the same four dimensions rated safeguards as satisfactory. for MIGA guarantees, each indicator measur- ing a distinct aspect of the guarantee’s perfor- IEG rated a project’s EPI successful when there mance, with the project development outcome was evidence that environmental and social rating being a synthesis of the four.2 The lack of actions had exceeded the minimum require- specification and measurement of safeguards ments of MNI, that is, (a) the environmental performance indicators as an integral part of and social measures introduced for the project the results frameworks of investment projects were being applied across the sector more limits the Bank’s ability to track and evaluate broadly beyond the physical footprint of the social and environmental outcomes at exit. The project; or (b) institutional arrangements had lack of instruments and performance indicators been made to continue environmental and for systematic client monitoring and supervision social measures beyond the life of the project; reporting (identified in the previous chapter), or (c) the socioeconomic measures introduced often compounded by the absence of baseline included gender programs, or went beyond data on social and environmental conditions, compensation for adverse impacts and resulted make it impossible to draw conclusions regard- in a sustainable stream of benefits or livelihood ing the outcomes and benefits of most of the standards for local communities that exceeded safeguards policies. On this matter, the resettle- preproject levels; or (d) the environmental and ment policy is a notable exception. The prescrip- social measures introduced for the project went tive nature of the policy and the requirement beyond compliance with standard requirements, to conduct baseline surveys and for third-party for example, with energy efficiency, greenhouse monitoring of resettlement outcomes has made gas mitigation, or biodiversity conservation. it easier to document resettlement results. IEG rated a project’s SCC successful when there However, the absence of provisions for reporting was evidence that the client’s capacity had been on social and environmental performance in ICRs assessed, gaps identified, and actions included means that despite the exhortations of some of the to build and sustain client capacity. The purpose safeguards policies, completion reporting remains of this indicator was to separate projects that weak and outcomes are either unreported or hard invested in strengthening client systems and to verify. If the Bank were to adopt environmental institutions from those that relied heavily on and social outcomes as a dimension of the overall technical assistance or external consultants to rating for the project’s development objective ensure satisfactory completion of the project. in the ICR and in IEG’s ICR Review, that would significantly change the incentive structure and go Factoring Environmental and Social a long way toward mainstreaming environmental Performance in the Assessment of and social outcomes. Project Development Outcomes The Bank does not consider environmental and Self-Assessment of Safeguard and social performance a significant dimension of Sustainability Results3 a project’s development outcome. However, Self-assessment of performance: World Bank environmental and social effects is one of the About one-fifth of category-A and half of four dimensions used by IFC and IEG category-B projects do not track or report on The World Bank to assess the development outcomes safeguard performance. Three of the environ- does not consider of IFC projects. The other three mental safeguard policies4 include provisions for environmental and social dimensions are: business success, evaluating environmental impacts and reporting performance a significant economic sustainability, and private on the achievement of environmental objectives dimension of a project’s sector development. MIGA had not and the effectiveness of mitigatory measures development outcome. undertaken self-evaluation until fiscal in completion reporting. However, there is 44 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E relatively little guidance beyond this statement. points to shortcomings in this area, About one-fifth of The two social safeguard policies5 contain more indicating that IEG needs to revisit its category-A and half of explicit guidance on supervision, monitoring, own methodology for assessment of category-B projects do and evaluation of impacts in the ICRs. This may completed projects to ensure more not track or report on be one reason why the quality of ICR report- robust assessment in accordance safeguard performance. ing on resettlement and indigenous peoples, with the provisions of the safeguard in projects where these policies are applicable, policies. is more rigorous (78 percent satisfactory) than on environment (56 percent satisfactory; figure Self-assessment of results: IFC 3.1). None of the five completed FI projects (all From its evaluation of the XPSRs IEG found triggering OP 4.01) reported on safeguards. some inconsistencies in project teams’ assess- ments of environmental and social outcomes. Some regional variations are significant. East The XPSRs are self-evaluations by operations Asia and Pacific did relatively better (76 percent) staff. IEG validates these ratings to arrive at an in terms of overall performance on environ- overall assessment of Environmental and Social mental and social reporting in ICRs, while Latin Effects. The ESE rating is based on the project’s America and the Caribbean (38 percent) and environmental performance in meeting IFC’s Africa (50 percent) lagged significantly behind. requirements and the project’s actual environ- Among the networks, Human Development mental impacts. Self-evaluation of outcomes Network performance was weak (21 percent) and performance in regard to requirements at in terms of ICR reporting on environmental appraisal has been partly insufficient in XPSRs. performance but strong on ICR reporting of In fiscal 2009, IEG changed the ESE relevant social results.6 ratings in XPSR evaluations in 23 There are inconsistencies percent of the sample (17/88) and in the way project teams An examination of of IEG ICR Reviews and downgraded net ratings for some assess environmental and Project Performance Assessment Reports also projects as well. social outcomes in XPSRs. Figure 3.1: Adequacy of ICR Reporting on Safeguards in Completed Bank Projects, by Safeguard Category 100% 81% 78% 78% 78% 77% 80% Percentage of projects 60% 54% 56% 47% 48% 40% 20% 0% n=98 n=31 n=62 n=97 n=31 n=61 n=55 n=27 n=27 Overall E&S Environment Social All projects Category A Category B Source: IEG portfolio review. Note: Excludes four projects for which ICRs were not available. 45 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D IFC has introduced an internal online Environmen- ing agency staff interviewed by IEG felt that the tal and Social Review Document (ESRD) as a more Bank’s resettlement policy had led to a greater structured platform for self-evaluation. ESRDs focus on the poor, especially those lacking legal include ratings for project-specific performance title to land. Without the Bank’s policies the risk indicators. Their effectiveness will be assessed assessment and corresponding mitigation plans after post-Performance Standard projects become would have been less comprehensive, would only mature for XPSR evaluations in 2011. have benefited titled landowners, would usually be limited to monetary compensation, and would Stakeholder Perceptions of Safeguards not have aimed to restore livelihoods. As a result, Performance there would have been weaker environmental Client perceptions, ownership, and implementa- and social protection. At the same time, the higher tion have bearing on the ability of the WBG to standards of the Bank’s social safeguards (annex promote environmental and social sustainability. 2, table X2.3) and rigidities in policy application IEG solicited feedback from clients on the value inhibit client ownership. Feedback was also more and effectiveness of WBG safeguards and Perfor- muted in terms of client capacity building: high mance Standards for all projects that were visited or substantial capacity building was reported by in the field. 55 percent for environmental and 58 percent for social aspects. IEG also assessed the quality of client implemen- tation as part of the portfolio review through an Clients also prominently identified more public examination of client capacity and actions taken participation and disclosure of information as to implement the environmental and social important actions that took place on projects as management plans agreed to during appraisal. a result of Bank involvement and the safeguard Client implementation is affected by ownership policies. Implementing agencies felt that the and commitment, client capacity, and institu- safeguards resulted in better selection of sites tional efficiency. IEG’s assessment took into and greater participation and consultation with account variations in implementation experience project-affected persons. over the life of the project to arrive at a rating for quality of client implementation for each Nonetheless, client feedback about the deterrent project in the portfolio review sample. When the effect of safeguards was confirmed by the staff client demonstrated commitment early during survey. Bank-wide, 38 percent of task team the project life, this resulted in higher ratings on leaders, 72 percent of social specialists, and 55 the indicator. When implementation was uneven percent of environmental specialists had encoun- and results were achieved only through intensive tered clients who wanted to avoid all or part supervision or through external consultants, this of a project because of safeguard policies. The was reflected in a lower rating. impact of this chilling effect was reported by a majority of team leaders from Latin America and Stakeholder perceptions of safeguards the Caribbean and over 40 percent from East Asia performance in Bank projects and Pacific and South Asia, which have the most Feedback from country clients7 revealed general active safeguards portfolios. Almost a fifth of agreement that safeguard policies result in team leaders had encountered a situation where more robust appraisal of project risks the team revised the scope or design of a project Country clients are and better environmental and social to avoid being classified as category A because in general agreement results; however, clients also expressed this high-risk category leads to higher levels of that safeguard policies concern about the rigidity and cost of scrutiny and higher costs. resulted in more robust the Bank’s requirements. Resettle- appraisal of project risks ment and compensation for project- On the other hand, project beneficiaries, local and better environmental affected persons dominated comments NGOs, and cofinanciers viewed the safeguard and social results. received from clients. Implement- policies much more positively. Over half of the 46 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E task team leaders surveyed reported that the Figure 3.2: Quality of Client Implementation in Bank’s safeguards increased acceptability of the Bank Projects, FY 1999–2008 Approvals project among beneficiaries, and the safeguard policies also increased acceptability among 100% nearly 30 percent of cofinanciers. Percentage of projects satisfactory 80% Local NGOs felt that the safeguards facilitated greater awareness of social and environmental 60% protection, and public participation, knowledge of, and acceptance of projects. 40% Client Implementation 20% Client implementation in Bank-financed projects 0% AFR EAP ECA LCR MNA SAR Total IEG assessment of client implementation of the mitigation plans for projects in the portfolio Source: IEG Portfolio Review. sample varied considerably across the regions from a high of 78 percent in East Asia and Pacific and 77 percent in Europe and Central Asia, to 37 percent and 41 percent in the South Asia and Africa, respectively,8 with a Bank-wide average of or better than implementation of Local NGOs felt that the 58 percent (figure 3.2). Higher scores indicate Safeguard Policies in all evaluated safeguards facilitated both ownership and implementation capacity. aspects. 9 Implementation quality greater awareness of South Asia scores were affected more by client has been good for management of social and environmental resistance to safeguards than by capacity hazardous materials10 and appears protection, and constraints, while Africa projects revealed more to have improved for provision of public participation, capacity gaps. Within East Asia and Pacific, results replacement and compensation knowledge of, and for China were even higher (88 percent), indicat- to project-affected persons, but acceptance of projects. ing high commitment and ability to implement implementation of EHS guidelines mitigation measures agreed with the Bank. and Corrective Action Plans has been success- ful in only 62 percent of post-Performance Client implementation was somewhat more Standard projects.11 Since many of the clients of effective in compensating project-affected evaluated post-Performance Standard projects persons for resettlement impacts or in mitigat- have not yet submitted their AMRs and some ing adverse impacts on indigenous peoples, IFC requirements have been applied only for both cases where explicit mitigation plans were very few projects, it is not possible to compare easier to identify and monitor, compared with implementation of pre- and post-Performance implementation of EHS guidelines, a third of Standard projects with regard to grievance which demonstrated lack of clarity on specific mechanisms and avoidance of adverse impacts measures to be implemented and monitored to indigenous people and cultural heritage sites. by the client. However, client implementation Likewise, because of insufficient data, it is not of resettlement was noticeably weaker in Africa possible to compare implementation of Perfor- compared with East Asia and Pacific. mance Standards with Safeguard Policies in IFC regions and industry sectors. Findings on pre- Client implementation in IFC-financed and post-Performance Standard projects are operations presented in annex 1, table X1.6. Overall quality of client implementation of Performance Standards in real sector projects As discussed previously, IEG found that about has been reasonably good and about the same half of the FIs implemented the social and 47 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D About half of the environmental management system Bank use of standardized performance metrics financial intermediaries in a satisfactory manner.12 This perfor- Overall, MNI received much more attention in implemented the ESMS in mance reflects the fact that the FIs the portfolio than enhancing positive impacts a satisfactory manner. do not have any financial incentives or strengthening client capacity (figure 3.3). or regulatory pressure to make sure These findings were corroborated by feedback that subprojects meet local and IFC environ- from staff, management, and NGOs. MNI could mental and social requirements, contrary to real be substantiated for 66 percent of the projects sector projects, where environmental and social in IEG’s portfolio sample. In addition, a third risk management is a part of the corporate risk (34 percent) of these projects went beyond management process to avoid costs of correc- minimum mitigation measures to enhance tive actions and reputational damage in case of positive impacts, while half (49 percent) had serious environmental and social incidents. credible measures to strengthen client capacity. Promotion of environmental and social sustain- Environmental and Social Performance in ability within client countries lags behind Bank Projects mitigation of immediate environmental and Supervision and monitoring reports suffer from social risks in Bank-financed projects. While the absence of systematic, structured informa- mitigation of adverse impacts has to remain tion on safeguard implementation and outcomes. a priority, over time the Bank needs to move Documentation for each project in the portfolio beyond protecting Bank-financed projects review was carefully examined to obtain evidence toward measures that strengthen environmen- of performance, as opposed to compliance with tal and social sustainability within borrower safeguard procedures. As highlighted in chapter countries. 2, information on safeguards in ISRs and monitor- ing reports was found to be miniscule and generic, Performance depends largely on the quality or absent, compared with the more detailed of client implementation and Bank supervi- information found in appraisal documents. For sion. While both were strong in East Asia and comparability, IEG assessed all projects on their Pacific, in Europe and Central Asia, which had achievement of the three composite performance considerably lower safeguards challenges, strong indicators—MNI, enhancing positive impacts, client implementation sufficed to overcome the and strengthening client capacity. limitations of somewhat weaker supervision. In Figure 3.3: Safeguards Performance in Bank Projects 100% 80% Percentage of projects 60% 40% 20% 0% AFR EAP ECA LCR MNA SAR Total Mitigating negative impacts Enhancing positive impacts Strengthening client capacity 48 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E contrast, greater supervision effort and invest- Figure 3.4: Bank Safeguards Performance by EA ment in client capacity building in South Asia Category overcame some of the constraints of weaker client implementation resulting in somewhat 80% better ratings for MNI than would have been otherwise (figure 3.3). Although safeguards Percentage of projects performance can also be affected by poor design, 60% this was rarely the case in the portfolio review sample. 40% In Bank-financed projects attention to sustain- 20% ability is much greater in category-A projects than in category-B and -FI projects. Data on implementation performance (figure 3.4) 0% A B FI Total indicate that category-A projects, to which the Mitigating negative Enhancing positive Strengthening client client tends to pay more attention, do a much impacts impacts capacity better job of strengthening client capacity. There are substantial missed opportunities in promot- ing sustainability in category-B and -FI projects, the last group being of greatest concern because of the heavy dependence on client capacity for (figure 3.5). Weaker preparation Attention to sustainability implementation. and much weaker supervision is much greater in in IDA countries compound the category-A projects The quality of preparation and Bank supervi- inherent weaknesses in IDA client than in category-B sion can profoundly affect client implementa- capacity and lead to a huge differ- and -FI projects. tion and project performance. Segmentation ence in the proportion of projects of performance for IDA and IBRD projects demonstrating satisfactory mitigation of provides the clearest example of this effect negative impacts. Figure 3.5: Process Implementation and Performance for IDA/IBRD Projects Preparation and appraisal Bank supervision Client implementation MNI results 0% 20% 40% 60% 80% 100% Percentage of projects satisfactory IDA IBRD Blend Source: IEG portfolio review. 49 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Investing in client capacity can lead to high implementation performance for Environmen- positive impacts over the long term. Recogni- tal Health & Safety, Involuntary Resettlement, tion of the Bank’s knowledge and expertise and Control of Hazardous Materials was less on infrastructure, including on managing its evident than in other areas. Among the projects environmental and social risks, induced China with involuntary resettlement, 40 percent of to maintain its relationship with the Bank even affected persons were physically displaced, though their need for financial support had and 60 percent had impacts on their liveli- attenuated (box 3.1). The transport sector in hoods. Among the projects in the sample with China adopted a structured learning process to satisfactory resettlement outcomes, livelihoods extend the knowledge gained from the Bank’s were found to be 15–85 percent better than capacity-building efforts in a few projects to the their preproject status. Projects that did not entire sector. succeed in restoring livelihoods, or did not provide any evidence of having done so, were Performance for different safeguard impacts rated wholly or partially unsatisfactory. IEG’s Evidence of mitigation of environmental and desk-based portfolio review was constrained social impacts was available in about two-thirds by the quality of reporting in supervision and of the sampled projects. In addition to the completion reports, but the findings are consis- composite performance indicators, the portfolio tent with those from the field case studies. This review also enabled IEG to assess the environ- performance is further compounded by the mental and social effects of different safeguard incomplete thematic coverage of Bank policies, policies. Performance for the 102 completed as illustrated by a transport project in Bangla- projects was combined with those for another desh (box 3.2). 66 projects with substantial implementation at the time of the assessment. Evidence of satisfac- Triangulation of findings on bank safeguards tory implementation, summarized in figure performance 3.6, ranged from 63–86 percent. Satisfactory IEG portfolio review findings are consistent with the information obtained from other sources of data. Of the Bank managers interviewed,13 the vast majority said that safeguards were Box 3.1: Structured Learning Process in the Chinese effective in mitigating adverse impacts in Transport Sector their project portfolio but were less effective in building client capacity than they would have liked (figure 3.7). Based on their project One interesting example consisted of 10 transport sector projects in portfolio, two-thirds of managers reported that China. China had approached the WBG not for its financial support, the Bank’s approach to safeguards had been but for its “world-class knowledge and expertise on infrastructure effective in building environmental capacity, and urban development projects.” What is most remarkable about but effectiveness in building client capacity for these projects is the extent to which the Chinese consciously and the current social safeguards was reported by systematically applied lessons learned about social and environmental less than half. safeguard policies, extending the knowledge gained on a few projects to the entire sector. Throughout this process, they strengthened their Staff feedback on safeguards performance has own internal capacities and institutions, thus significantly contributing generally been positive, with environmental to better preparation and implementation. The assessment by IEG was and social specialists rating performance higher reinforced by feedback from staff who reported that “local practices than task team leaders: 57 percent of task team are moving ever closer to the Bank’s policies… being constantly leaders and 69 percent of environmental and strengthened by the government. As a result, in-country capacity is social specialists surveyed said the application high.” of Bank’s safeguards policies resulted in mitiga- Source: IEG interview with regional manager. tion of adverse impacts, while 10 percent of task team leaders said there was no mitigation at all. 50 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E Figure 3.6: IEG Ratings for Implementation Performance in Bank Projects Indigenous peoples 75% 17% Involuntary resettlement 68% 42% Physical cultural resources 86% 12% Natural habitats 72% 13% Control of hazardous materials 68% 28% 63% 100% EHS/EMP outcomes 0% 20% 40% 60% 80% 100% Percent of projects satisfactory Satisfactory outcomes Percentage of projects Source: IEG portfolio review. Box 3.2: Protecting Vulnerable Populations—Do Safeguards Ensure Adequate Coverage of Social Risks? Dhaka’s streets, like those of many cities in developing coun- Unanticipated project effects: A year before project close it tries, are shared by cars, buses, nonmotorized vehicles, and became evident that a major social risk had been overlooked pedestrians. They are plagued with gridlock and accidents, during appraisal, despite close scrutiny by environmental and caused in part by poorly developed road infrastructure in Dhaka social safeguards specialists. The ban on bicycle rickshaws relative to the population and urban area served. on major arterial roads was causing significant loss of income due to labor impacts on 54,000 affected rickshaw pullers, whose The Dhaka Urban Transport Project (DUTP) aimed to improve remittances to impoverished rural districts helped keep families transport infrastructure and services to improve traffic flow and out of poverty. overall mobility for the traveling public. The project financed con- version of highly polluting two-stroke auto-rickshaws to cleaner What the safeguards contributed and what they missed: Project four-stroke engines, 175 kilometers of pedestrian footways, 63 preparation included substantial up-front work for the environ- kilometers of arterial roads, traffic signals at 69 junctions, and mental and resettlement safeguards (OP 4.01 and OD 4.12), the rehabilitation of three major bus terminals. After a midterm only safeguards policies triggered. Under the current safeguard review, several major bus arteries were closed to nonmotor- policies, no social impact assessment was required, and no ized vehicles. comprehensive assessment was made of how the project might affect local communities or vulnerable groups. A comprehensive Positive project outcomes: Traffic surveys indicated improved social assessment of community impacts would have allowed traffic flow, reduced traffic accidents (fatal accidents down by early identification of vulnerable populations and timely intro- 33 percent, injuries down by 85 percent), and average travel duction of an appropriate mitigation plan as part of the DUTP. time reduced by about 30 percent. The majority of transport As the project was about to close, a mitigation package con- users surveyed were positive. The project also brought about sisting of technical training and microcredit had to be designed significant improvements in urban air quality (lead pollutants somewhat belatedly, and financed from another IDA credit down by 30–60 percent). (Microfinance II) to address these adverse impacts. Source: Project files. 51 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 3.7: Manager Feedback on Bank Safeguards Effectiveness Social capacity building Environmental capacity building Social mitigation Environmental mitigation 0% 20% 40% 60% 80% 100% Percentage of responses Ineffective Mostly ineffective Mostly effective Fully effective Source: IEG interviews with country directors and sector managers. Results varied by region, environmental mitiga- ing the project’s development objectives, tion being better in Africa, East Asia and Pacific, thereby contributing to development effective- Europe and Central Asia, and Middle East and ness. Environmental and social specialists North Africa, while social mitigation was rated viewed the impact of safeguards even more better in Latin America and the Caribbean and positively than task team leaders: about three- South Asia (see figure 3.8). Half of the respon- quarters of environmental and social special- dents to the task team leader survey felt that the ists felt safeguards impacts in their own area of application of the Bank’s safeguard policies to specialization enhanced the project’s develop- the project enhanced the likelihood of achiev- ment effectiveness. But country directors are Figure 3.8: Task Team Leader Assessment of Mitigation Outcomes of Bank Safeguards, by Region mitigationto a large or moderate extent 80% 68% Percentage of TTLs reporting 65% 63% 61% 61% 60% 59% 60% 52% 53% 53% 54% 45% 42% 40% 36% 20% 0% Africa East Asia Europe & Latin America Middle East & South Asia Total & Pacific Central Asia & Caribbean North Africa Mitigation of environmental impacts Mitigation of social impacts Source: IEG staff survey. 52 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E less sanguine and would like the rigidities in implementation to give way to a risk manage- Box 3.3: Concerns of Country Directors about ment approach with greater focus on develop- Safeguards ment effectiveness (box 3.3). Country directors are most concerned about rigidities in safeguard poli- Nonetheless, IEG surveys indicate that staff cies and implementation, as well as weak links with development results. are acutely aware of the discouraging effect of safeguards and Performance Standards on • “There is no problem with the safeguard standards… It is the clients. Feedback from Bank staff described performance metrics and procedures that are the constraint. We above is mirrored by findings from the IFC must judge task teams by their contribution to safeguards, not by staff survey. Among investment officers, 40 identifying what the government has not complied with.” percent reported that Performance Standards • “According to Bank policies, safeguards compliance is the re- had a negative impact on IFC-client relation- sponsibility of the client and it is the responsibility of the team to ships. Forty-seven percent of investment supervise compliance. Currently in our region the pendulum has officers, and 56 percent of environmental and swung the other way with the region expecting the teams to be social specialists, had experienced a situation responsible for compliance… We need to rebalance the distribution where the client wanted to avoid dealing with of responsibility between the client and the Bank.” the IFC because they thought it would be too • “We need to make sure that a risk management approach is properly expensive or time-consuming. These numbers designed and implemented in our portfolio in order to manage both are of the same order of magnitude as among reputational risk and our own resources more efficiently. Currently Bank staff. the Bank is risk averse, and this has practical implications for our staffing and resources and for those of our client.” While the majority of NGOs reported that WBG • “[Safeguard policies] are certainly useful for development effective- performance on safeguards had improved, less ness. If you are doing huge infrastructure projects and not taking than half were satisfied with the WBG’s impact safeguards seriously you will not have the strong effect you would on building client capacity to manage environ- like. But we should look closely to see to what extent safeguards mental or social impacts. The majority of NGOs have contributed to development results.” agreed that environmental and social perfor- • “We are trying to do the right thing in paying attention to safe- mance of WBG-financed projects was better guards, but we can bring a lot of improvement to the policies and than that of other IFIs and client governments. how we implement them… The policies have evolved over time in However, only half of the respondents assessed a reactionary manner. This may be the time to look at the policies the effectiveness of the safeguards and sustain- and their impact on the client.” ability policies as being high or substantial,14 local NGOs being more positive than international Source: IEG interviews with country directors. NGOs about WBG effectiveness. NGO percep- tions on Bank capacity building are similar to IEG’s assessment. Recent organizational restructuring appears to have contributed to a growing divide between increasing void between safeguards and the safeguards and nonsafeguard environmen- sustainability agenda. There is evidence of tal and social work programs. Feedback from more careful screening of projects at entry, focus group discussions with staff revealed greater attention to category-A projects, and that recent institutional restructuring, includ- more risk aversion reflected in an inflation of ing the relocation of the Quality Assurance projects being classified as category B. Staff and Compliance Unit (QACU) from ESSD have pointed to more centralized control and, to Operational Policy and Country Services in some instances, divergence in interpretation (OPCS)15 has led to greater attention to the of policies and standards between the regions safeguards but has come at the cost of an and centrally based staffs. 53 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D The artificial separation of environmental increased attention is being paid to enhanc- and social staff between those who work on ing positive impacts and strengthening client safeguards and those who work on social or capacity without significant differences in pre- environmental sustainability is a cause for and post-Performance Standard project per- concern. The merger of infrastructure sectors formance. The level of performance regarding with the environmental and social development mitigation of negative impacts and enhancing sectors under one vice presidency has given rise positive impacts of Performance Standard proj- to a surge in demand for safeguards services, ects is similar to that of pre-Performance Stan- but the demand-driven nature of the relation- dard projects. While increased attention is being ship between infrastructure task team leaders paid to developing clients’ SEMS under Per- Increased attention and environmental and social staff is formance Standard 1, satisfactory project per- is being paid by IFC forcing a division of labor among the formance remains a challenge, particularly for to enhancing positive social and environmental staff that is financial intermediaries. impacts and strengthening unnecessary. A formal examination client capacity without of the organizational incentives and Real sector projects significant differences effectiveness is beyond the scope of Among real sector projects the share of satisfac- in pre- and post- this evaluation. However, the separa- tory performance of post-Performance Standard Performance Standard tion of QACU from the oversight of projects in applying the first three Performance project performance. the environmental and social Sector Standards and EHS guidelines has been at about Boards appears to have exacer- the same level as for pre-Performance Standard bated this divide. The effect of these projects covering similar environmental and tendencies is that across the Bank, most of the social aspects (figure 3.9). However, the portfolio Bank staff who work on safeguards do not work review does not permit an assessment of Perfor- on environmental and social sustainability, while mance Standards 5–8, which were triggered those who work on sustainability no longer work in too few projects in the portfolio sample to on safeguards. This is not an optimal use of Bank evaluate their performance. resources, and is in contrast to IFC and MIGA where the mitigation agenda is an integral part IEG has reviewed clients’ annual reports and of social and environmental sustainability. other environmental and social documents for validation of the ESE section in the XPSRs and Sustainability Performance in IFC collected information on main performance Projects indicators since 2004, covering project appraisal IFC use of standardized performance metrics years 1999–2004. The share of satisfactory Since the sample was quite small and several implementation performance along key environ- post-Performance Standard projects had not yet mental and health and safety indicators in 209 submitted Annual Monitoring Reports or Annual pre-Performance Standard projects is presented Environmental Performance Reports, it was in figure 3.10. The upper dark bar represents possible to obtain information only from 18–28 the percentage of satisfactory performance and projects for each indicator. The performance the lower light bar the percentage of projects of post-Performance Standard projects will not showing the respective Performance Standard in be available for full evaluation before 2011; the total XPSR sample. therefore, the results presented here are indica- tive only. A review of Performance Standards IEG found 71 percent of the real sector, implementation based on the portfolio review pre-Performance Standard projects have and projects visited in the field is presented in satisfactory environmental and social perfor- annex 4. mance overall. Most projects demonstrated a satisfactory EMS.16 Project performance has Preliminary findings indicate that while main- also been good in emergency preparedness taining the focus on mitigating negative impacts, and management of solid wastes, hazardous 54 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E Figure 3.9: Implementation Performance of Safeguard Policies and Performance Standards (real sector) PS1 S&E assessment and management system 86% OP 4.01 Env. Assessment 82% PS2 labor & working conditions 78% Guideline on occupational H&S 75% PS3 pollution prevention & abatement 67% WBG env. guidelines 73% PS4 community health & safety 100% No pre-PS requirements PS5 land acquisition and inv. resettlement (NOP) OD 4.30 Inv. Resettlement 92% PS6 biodiversity conservation (NOP) OP 4.36 Forestry & OP 4.04 Natural Habitats 97% PS7 indigenous people (NOP) OD 4.20 Indigenous People 100% PS8 cultural heritage (NOP) OP 4.11 Cultural Property 96% WBG guidelines overall 62% 67% WBG guidelines – overall 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Post-PS performance Pre-PS performance Source: Pre-Performance Standard real sector results are based on IEG’s fiscal 2004–09 Environmental and Social Reviews for 209 XPSR projects appraised in fiscal 1999–2004, and post-Performance Standard real sector results are based on the portfolio review of 24 projects. Note: PS = Performance Standard, NOP = No opinion possible. The differences between pre- and post-Performance Standard ratings are not statistically significant. materials, and fuels, and satisfactory in occupa- mance Standard projects in the FI sector, the tional health and safety, control of air emissions, sample for this evaluation was expanded to 181 liquid effluents, and energy efficiency. However, investments—139 pre-Performance Standard emergency preparedness and energy efficiency projects from the XPSR database, which contains aspects were present in only half of the projects. compliance information regarding IFC’s standard The 139 evaluated pre-Performance Standard FI requirements for FIs, and 42 randomly sampled projects (not depicted in the chart) achieved 65 post-Performance Standard projects with Annual percent satisfactory rate for their Environmental Environmental Performance Reports submitted. and Social Effects (the difference was within the Summary performance on quality of the social sampling error of 3.3 percent). and environmental management systems at FIs, and their implementation in appraising and Financial intermediaries supervising subprojects are shown in table 3.1, Assessment of FI project performance in which reveals no significant difference between post-Per for mance Standard projects is pre- and post-Performance Standard projects. constrained by scarcity of information on subprojects. Since Annual Environmental Perfor- SEMS implementation was satisfactory only in mance Reports were available for only 10 projects about half of the evaluated pre-Performance in the original random sample of 16 post-Perfor- Standard and post-Performance Standard FI 55 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 3.10: Satisfactory Environmental and Social Performance in IFC’s Real Sector, Pre-Performance Standard Projects Appraised, FY 1999–2004 Environmental and social effects (ESE) 71% Real sector projects (Non-FI) 100% 87% Environmental management system (EMS) 80% 77% Occupational health & safety 85% Non-FI projects. n=209 92% Emergency preparedeness & fire safety 51% 93% Management of solid wastes 75% Management of hazardous materials & fuels 87% 71% Control of liquid effluents 73% 71% Control of air emissions 80% 66% Energy efficiency 75% 53% 0% 20% 40% 60% 80% 100% Satisfactory rate Percentage of sample Source: IEG’s fiscal 2004–09 Environmental and Social Reviews for 209 XPSR projects appraised fiscal 1999–2004. Table 3.1: Performance of FI Projects in IFC’s Portfolio Post-Performance Standard, Pre-Performance Standard, n=42 n=139 (performance (safeguards) standards) Satisfactory Satisfactory Indicator Number results Number results Social and environmental management: system documentation 112 71% 41 78% at entry (policy, procedures, and organization) Social and environmental management: implementation 71 56% 38 50% (appraisal and supervision and ensuring environmental and social compliance of subprojects) Source: Pre-Performance Standard real sector results are based on IEG’s fiscal 2004–09 Environmental and Social Reviews for 139 XPSR projects appraised fiscal 1999–2004, and post- Performance Standard real sector results are based on the expanded study portfolio of 42 projects. Note: The differences between pre- and post-Performance Standard ratings are not statistically significant. 56 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E projects. About three-fourths of the FIs had for professional and independent environmental developed an SEMS document describing the and social audits because of the associated costs company policy, procedures for screening, in a competitive market situation. As intermediary appraising, and monitoring subprojects, as well financiers they do not pose reputational risks to as nominated people to oversee and implement IFC comparable to that of IFC direct investments environmental and social appraisal and supervision in category-A or some category-B real sector of subprojects. However, SEMS implementation— projects. Therefore, until recently IFC has not paid proper screening, appraisal, and supervision of sufficient attention to poor environmental and subprojects, and ensuring compliance with EHS social performance of FI projects. CESI staffing, guidelines, safeguard policies, and Performance although increased from one to five in the past five Standards as appropriate—was satisfactory only in years, is still small compared with 57 environmen- about half of the evaluated pre- and post-Perfor- tal and social specialists dedicated for real sector mance Standard projects. The other half of the projects. Contrary to some real sector projects, FIs seem to have taken only the first easy step to IFC has never disinvested a financial intermedi- develop a social and environmental management ary project because of environmental risks— document, but were not willing to implement it. although its finance to one high-risk subproject or numerous SME industry projects may have signifi- The FIs often lack guidance on proper catego- cant cumulative environmental and social impacts rization, resulting in systematic downgrading of and improvement opportunities. category-B subprojects to category C and insuffi- cient information to IFC on subprojects that The financial sector needs more capacity building should have been in category A. However, a great and incentives to improve subproject perfor- majority of the FIs reported that they followed mance; IFC’s supervision alone cannot improve the exclusion list and host-country environ- performance. The IFC’s Environment and Social mental laws and regulations. Legal compliance Department discontinued the Competitive provides some comfort on mitigating negative Business Advantage sustainability training program impacts at subproject level in countries with for financial intermediaries in 2006 because it was strong environmental law enforcement, but as regarded as a one-off event and insufficient to observed during IEG site visits, law enforcement build sustainable client capacity in environmental in many countries is weak and subprojects in management. Instead, the Environment and Social industries with small and medium enterprises Department has started e-learning and partnership are outside the purview of regulatory authorities. programs with local entities in emerging markets to deliver training for financial intermediaries. Although IFC’s environmental and social work quality in FI projects has improved substantially Although IEG welcomes such long-term capacity in the past three years with increased supervi- building programs, to date the program has a sion resources, IEG’s ratings for ESE in FI projects very narrow regional scope—in Brazil, China, and is still low (63 percent, fiscal 2007–09 average)17 India only. Many financial intermediary clients compared with real sector projects (71 percent), have changed their environmental coordinators and SEMS is implemented effectively only in half and managers in the past three years and new of the FI projects. There are three reasons for low staffs in the financial intermediaries do not have performance. First, financial intermediaries do sufficient capacity for environmental appraisal not have a legal obligation to the host country to and monitoring of subprojects. The discontinua- ensure sustainability of their subprojects, which in tion of IFC’s global training program has created turn do not have direct contractual obligations with a major gap in IFC’s ability to build the capacity IFC to meet its environmental and social require- of financial intermediaries, which urgently needs ments. Second, many financial intermediaries to be addressed. IFC should make better use of lack environmental management capacity. Third, opportunities to build capacity by visiting their many are unwilling to hire external consultants subprojects and demonstrating good appraisal and 57 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D The financial sector needs supervision practices. Visits to subproj- a project, expands upon the scope of MIGA’s more capacity building ects would also give IFC an opportu- previous (1998) Environmental Assessment and incentives to improve nity to verify SEMS implementation by Policy by including social impacts and clarifying subproject performance. the financial intermediaries in subproj- the clients’ responsibilities. Thus, MIGA’s review- ect appraisal and supervision, as well at-entry of a project that is expected to have as mitigation of adverse environmental significant environmental and social impacts and social impacts by their clients. should include an assessment of the capacity of the client’s SEMS and the identification of correc- This review of implementation performance tive actions to bring it up to standards. identifies the benefits of IFC’s new policy framework in terms of the scope of the policies— IEG’s portfolio review found that the new such as the increased coverage of social PPSSES is leading to greater attention to the aspects—and the focus on the clients’ social clients’ SEMS, but there is still a gap in reaching and environmental management the goals of Performance Standard 1. While The PPSSES is leading systems. This determines the nature every category-A, -B, and -FI project is subject to to greater attention to of environmental and social impacts review-at-entry, MIGA’s environmental and social MIGA clients’ SEMS, but and the client’s capacity to undertake clearance memorandum discusses the adequacy there is still a gap in mitigation measures being examined. of the clients’ SEMS for only 39 percent (7/16) of reaching the goals of The preceding assessment has looked the pre-PPSSES projects and 70 percent (7/10) of Performance Standard 1. at the stringency of the targets, the the post-PPSSES projects in the sample for which effect of categorization (A versus B), it would have been expected. and follow-up by clients in the new system, which together provide proxies for the environmental The risks associated with inadequate SEMSs and social impacts, although direct measurement are illustrated by a category-B project for of the impacts would have been preferable. which the client requested an EHS audit. The audit, funded by the Japan-MIGA Trust Fund to Sustainability Performance in MIGA Address Environmental and Social Challenges Projects in Africa, found that the client’s social and As stated in the PPSSES, MIGA expects to achieve environmental management system did not its environmental and social objectives through meet MIGA’s requirements, even though the the application of a comprehensive set of Perfor- client had obtained an environmental license mance Standards to the projects it guarantees. from the host country. The audit also found While the Performance Standards have only been that a fire had recently occurred and not been in place since October 2007, they are largely car- reported to MIGA. For each of the EHS gaps, ried over from the previous safeguards policies the audit suggested corrective actions. A recent and represent MIGA’s most recent and authori- IEG mission found that the audit had been well tative statement of what its sustainability frame- received and the client was actively remedying work is intended to achieve. However, as a result the identified shortcomings. of MIGA’s limited monitoring of project perfor- mance and implementation, only sparse infor- To address client capacity challenges, MIGA has mation is available on actual performance on established the Trust Fund to Address Environ- minimizing negative impacts, enhancing positive mental and Social Challenges in MIGA-guaran- impacts, and strengthening client capacity. teed projects in Africa. Since its launch in 2007, this program has supported nine environmen- Strengthening client capacity tal and social capacity strengthening activities, The application of Performance Standard 1 is of which three have been completed.18 One MIGA’s main instrument for strengthening client of these activities paid for the aforementioned capacity. Its objective, to promote the clients’ EHS audit, which helped a client address a effective use of a SEMS throughout the life of number of SEMS gaps. MIGA follow-up client 58 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E surveys indicate that client satisfaction with since IEG does not have information from the these activities has ranged from very good to review sample on performance for the other two excellent. (PS7,PS8). MIGA’s 2007 sustainability framework requires The broadened scope of the Performance an assessment of the social and environmental Standard on Labor and Working Conditions management system of financial intermediaries, (PS2) has led MIGA to pay much more attention which MIGA had not previously examined. Even to labor issues than before. IEG found that so, out of the six post-Performance Standards the client’s human resource policies had been financial sector projects in the sample, IEG reviewed and found satisfactory for 100 percent found that the review-at-entry commented on (7/7) of the category-B projects approved after only 50 percent (3/6) of the client’s SEMS. This is Performance Standard 2 became effective, an important gap, since financial sector projects whereas they had been discussed for only one constitute a large and growing share of MIGA’s of the pre-PPSSES sample. This is fully in line portfolio, and most of these projects support a with MIGA’s PPSSES, but raises a question as to diversified portfolio, including projects in sectors why Performance Standard 2 is not extended that typically face significant environmental and to category-C and -FI projects, since there is no social risks. prima facie indication that the same issues may not also be relevant to them. Furthermore, MIGA’s due diligence is limited to the corporate policies of the parent bank. Under the Performance Standard on Commu- How well the corporate parent is able to ensure nity Health, Safety, and Security (PS4), MIGA that its subsidiaries—which are supported has devoted greater attention to the clients’ by MIGA—abide fully by these policies is not management of potential community risks and always clear. Given the likelihood of significant impacts. Two recent projects illustrate what the differences in implementation capacity between process involves. For a medium-sized indus- a parent bank’s headquarters and its subsidiar- trial project in Asia, the client committed to ies, with less sophisticated developing country disclose planned activities and other environ- partners and country systems, this approach mental and social effects to the communities’ may not provide an accurate assessment of the leadership, as well as to undertake regular con- subsidiary financial intermediary’s SEMS com- sultations and procedures for resolving commu- pliance with national laws and MIGA’s Perfor- nity grievances. For a power transmission line mance Standards, as required by Performance in Latin America, the host country’s legislation Standard 1. To date, MIGA has not received any requires, and the client committed to a social monitoring reports on the environmental and communication program that provides land- social performance of its financial sector proj- owners and local residents the opportunity to ects that would support the adequacy of its report environmental and social concerns that approach, which relies solely on a review of the arise throughout the construction period. In corporate policies of the parent banks. the absence of MIGA monitoring, however, it is not possible to ascertain if these commitments Minimizing negative and enhancing positive have been met. social impacts Aside from Performance Standard 1, which is the As a supporting initiative under the 2007 policy overarching framework for enhancing environ- reforms, MIGA proposed that it would regularly mental and social management, five Perfor- report on the local community impacts for a mance Standards19 constitute MIGA’s tools for small number of projects where such impacts minimizing negative and enhancing positive may be significant. On this basis, two action plans social impacts. The discussion that follows were under preparation, but the guarantees did focuses on the three standards (PS2, PS4, PS5), not go forward.20 59 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Given limited MIGA monitoring, implementation minimizing the negative and enhancing the of the Performance Standard on Land Acquisi- positive environmental impacts of the projects tion and Involuntary Resettlement (PS5) is it guarantees. difficult to assess, and some risks are evident. In a power transmission project in Africa, approved MIGA’s limited monitoring of the Performance under the safeguards framework, no RAP had Standard on Pollution Prevention and Abatement been prepared, even though the due diligence (PS3) makes it difficult to assess clients’ perfor- mission had found that over 200 homesteads mance. While IEG found that environmental were going to be affected. Four years after, MIGA’s problems were identified in 25 percent (2/8) first monitoring mission found that the majority of the mature sample projects (older than two of resettlement and compensation cases had not years) that MIGA had monitored, the absence been satisfactorily handled, leading to numerous of monitoring cannot be assumed as represent- grievances, and committed the client to correct ing satisfactory performance for the 75 percent outstanding problems. A follow-up monitoring of projects that have not been monitored. The mission two years later found that, while a few single case (mentioned earlier) for which the important claims had been settled, most of the comprehensive audit requested by the client claimants had moved and could no longer be found numerous environmental and social traced and adequately compensated. management inadequacies illustrates the problems that may be hidden when monitoring A second example shows how active monitor- is not done. ing can help address unexpected problems and improve performance. Land compensation for MIGA has been fairly consistent in its imple- a highway project in Latin America had been mentation of its Performance Standard on Bio- delayed because of conflicting ownership claims diversity Conservation and Sustainable Natural on the same land arising out of faulty histori- Resource Management (PS6), although some cal transfers. Two years after project approval, a avoidable losses occurred. IEG found that sev- monitoring mission staffed by a qualified social eral projects support the sustainable use of natu- scientist scoped out the problems and requested ral resources, by incorporating environmentally the government to give priority to resolving such friendly, energy efficient, and resource saving cases through the courts system. A year later, technologies, to ensure project impact will be a follow-up visit verified that the resolution of positive relative to the facilities they replace and these cases had accelerated, but the project is local comparators. A few projects incorporate still much delayed from the original schedule. intensive water treatment and recycling technol- ogies to ensure their viability and minimize their MIGA’s limited monitoring capacity is constrain- impact in water-scarce locations. ing its ability to ensure that its projects meet the applicable Performance Standards. While it can be Two of the projects in the MIGA sample supported argued that the resettlement process and perfor- the restoration of previously degraded habitats: mance would have been worse if MIGA had not a power plant in Europe included the creation been involved, they also point to areas needing of small ponds and wetlands, which will not more attention and resources. These cases only compensate for the losses from its own ash highlight the importance of preparing a full RAP disposal site, but also for some of the habitat before starting construction and the importance converted by an older plant it replaces. The of timely supervision. other case is a wastewater treatment plant in the Middle East that will significantly improve Minimizing negative and enhancing positive the currently anoxic water flow into a major environmental impacts reservoir, enabling its restoration as a viable Along with Performance Standard 1, Perfor- aquatic ecosystem and wintering ground for mance Standards 3 and 6 are MIGA’s tools for migratory birds. 60 E N V I R O N M E N TA L A N D S O C I A L P E R F O R M A N C E Implementation performance from the at the Bank, IEG was able to find evidence of perspective of MIGA’s new sustainability satisfactory mitigation of negative impacts in only framework two-thirds of Bank projects. Comparing the implementation performance from projects processed under MIGA’s 2007 Per- The adoption of the PPSSES provided IFC and formance Standards with projects subject to the MIGA with an integrated framework to assess earlier safeguards, IEG found some improvement environmental and social performance. This in the following areas: has been strengthened by the introduction of additional monitoring instruments by IFC, such as • Appraisal of the projects’ SEMS, including the online ESRD document with specific perfor- those of Financial Sector projects (the latter mance indicators and a rating system; however, were previously classified and screened as its efficacy depends on the quality of reporting by category-C projects) clients, and independent assessment of impacts • Appraisal of labor and working conditions is not the norm. The limited experience to date • Appraisal’s attention to community consulta- does not permit IEG to assess rigorously the tions performance of IFC projects prepared under the • Provision of technical assistance to SEMS-chal- PPSSES, yet there appears to be greater emphasis lenged clients in Africa. on sustainability outcomes in the Performance Standards approach. On the other hand, IEG found a substantial gap in MIGA’s ability to monitor implementa- The Bank’s safeguards work appears Although there have been tion performance and provide assurance that to be more focused on the immedi- some improvements, the objectives of the Performance Standards ate effects of the projects it finances there is a substantial are being met. Most of this gap is due to a (and related reputational risks) gap in MIGA’s ability to shortage of capacity and resources devoted to than on longer-term environmental monitor implementation MIGA’s environmental and social unit. In IEG’s and social sustainability. The Bank's performance and judgment, it is not feasible for MIGA to fully policies and systems emphasize provide assurance that comply with the requirements of the PPSSES appraisal of environmental and social the objectives of the and meet the expectations of the Performance risks and proposed mitigation plans Performance Standards Standards unless its environmental and social to ensure compliance with safeguard are being met. capacity is substantially increased. policy requirements. Strengthen- ing client capacity and enhancing Summary of Main Findings positive impacts tend to take a back The main purpose of this chapter has been to seat to mitigating the adverse impacts induced assess implementation performance of the WBG’s by Bank-financed investments. IFC's and MIGA's safeguards and sustainability policies. The WBG new policy emphasizes environmental and social lacks a clear framework to assess the impacts sustainability and achievement of the Perfor- of their safeguards and sustainability policies. mance Standards. IEG found increased emphasis Environmental and social outcomes of Bank on strengthening the client's social and environ- projects are not clearly articulated, performance mental management system and enhancing indicators are rarely specified and integrated in the client responsibility for implementation and project’s results framework, and data to monitor achievement of the Performance Standards since and evaluate are not routinely collected and used. introduction of the PPSSES framework in IFC and The Bank relies primarily on compliance with the MIGA. IFC's approach encourages its clients to environmental and social management plans as a adopt additional measures to improve community proxy for assessing mitigation of adverse impacts. impacts and EHS outcomes. Chapter 5 will return Despite the emphasis on mandatory procedures to these alternative paradigms for doing business. 61 Chapter 4 Evaluation Essentials • The WBG’s safeguards framework generates significant benefits, but these are not systematically mea- sured or quantified. • IFC’s budgetary resources devoted to its sustainability framework are being efficiently allocated toward projects with higher risks and benefits. • Bank costs are well targeted to- ward high-risk projects but findings on category-B projects indicate suboptimal allocation of resources. • Bank clients tend to allocate re- sources efficiently in meeting safe- guards requirements, but allocative efficiency cannot be assessed for IFC clients because data are not available. • Better monitoring, documentation, and reporting of environmental and social impacts are needed to improve the quality of benefit-cost analysis. • The assessments of benefits and costs show that the WBG's safe- guards framework generates sig- nificant benefits for the mitigation of environmental and social risks of projects. Crews working on a road in Peru. Photo courtesy of Jouni Martti Eerikainen. Risks, Benefits, and Costs of the WBG’s Safeguards and Sustainability Policies Introduction client capacity). The use of those standardized This chapter explores the efficiency of safeguards indicators was aimed at comparing performance work across the WBG by juxtaposing the distri- across the portfolio. The Risks and Benefits bution of benefits against costs. The WBG Model developed by IEG (box 4.1), and applied does not collect data on environmental and to the portfolio review sample, takes this analysis social benefits and costs. As an alternative, IEG further by weighting the performance by the risk developed proxy measures to estimate benefits rating. The model used a log scale developed that were compared with project costs obtained by IEG—building on criteria described in the by IEG from the WBG’s resource management Environmental Assessment policy (OP 4.01)—to databases and project data. This chapter extends rank projects by their significance and severity of the analysis of performance by ranking projects risks, with higher risk ratings reflecting multiples based on the significance of their likely environ- of lower risks. mental and social impacts, and uses this ranking to analyze the efficiency of the WBG’s safeguards IEG applied the model to develop a risk profile and sustainability frameworks. The likelihood of of the Bank’s portfolio, based on environmen- their occurrence and the extent to which impacts tal and social data for the 102 projects in the will be mitigated is uncertain; hence risk is a more review sample that had been completed by the accurate descriptor than impacts. The relative time of this assessment and for which ICRs were risk scale provides IEG with alternate metrics available. The risk profile was compared with to analyze the distribution of risks embedded the safeguard categories for these projects and in the project categorization system. It is also econometric analysis was undertaken to assess a practical technique for estimating benefits of the robustness of the safeguards categorization the safeguards and sustainability frameworks. system. The number of IFC and MIGA projects Even though IEG was only able to obtain limited from the portfolio review, for which comparable data on costs, comparing those findings with data were available, was too small to replicate the estimated benefits allows partial evaluation of econometric analysis of risks. the efficiency and cost effectiveness of safeguards and Performance Standard implementation. IEG estimated risks and benefits for the projects whose costs were available to compare the Chapter 3 reported on IEG ratings for projects in benefits and costs of the safeguards and sustain- the portfolio review sample against three perfor- ability frameworks at the Bank and IFC. The risk mance indicators (mitigating negative impacts, model provided a modality to impute the value enhancing positive impacts, and strengthening of benefits by multiplying the MNI performance 65 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D of WBG projects by the aggregate environmental triggered Involuntary Resettlement, and 22 and social risk ratings. By adjusting MNI accord- percent triggered the Indigenous Peoples policy, ing to the significance of risk, this measure which is about the same proportion as in the provides a proxy of the actual benefit (B) from Bank’s portfolio. Data for estimating risks were the implementation of the WBG’s safeguards obtained from IEG’s review of appraisal and for a specific project. IEG was able to collect supervision documentation. Safeguard perfor- data on both Bank costs for safeguards appraisal mance for these completed projects was assessed and supervision and client costs for safeguards by IEG using the Risks and Benefits model implementation for 35 Bank projects that described in box 4.1. Results were tabulated had triggered both social and environmental separately for the social risks (RS) and environ- safeguards. This provided the empirical basis for mental risks (RE), estimated using the indicators the analysis of efficiency of resource use. described in table 4.1 and compared with the results from project categorization. IEG applied the same model to estimate the distri- bution of risks and benefits for projects in the IFC The risk assessment provided IEG with the means portfolio sample where these data were available. to explore three issues: (a) consistency between MNI ratings were only available for 37 projects in objective environmental and social risk criteria IFC’s sample. For proprietary reasons, IFC does and safeguard categorization in Bank projects, not have access to data on client costs incurred (b) the extent to which safeguard categories are on safeguards or Performance Standards. IFC determined by social versus environmental risks, costs therefore include costs incurred by IFC and (c) which variables are key determinants of only, which too are likely to be a fraction of costs category-A projects. incurred by IFC clients. Comparative MIGA data are not available at this time. Regression analysis A probit regression was carried out on the What Determines Safeguard completed projects in the Bank portfolio review Categorization? to test the effect of environmental and social Categorization is a screening mechanism risk ratings, regional effect, and network effect described in the Environmental Assessment on project categorization (see regression results policy (OP 4.01) to classify projects based on the in annex 6). The sample contained 32 category- significance of anticipated environmental risks. A projects, 65 category-B, and 5 FI projects. Category-A projects require a more exhaustive The regression analysis included 97 completed Environmental Impact Assessment (EIA) and projects (after excluding the 5 FI projects), 53 have more stringent disclosure requirements.1 of which had both social and environmental risk Categorization has had significant resource ratings. The dummy for category A was the regres- implications. The degree of environmental sand while the regressors included environmen- risk signified by the assigned category served tal and social risks (8 ordinal gradations of risk as the primary mechanism to determine the level), sector (dummy for infrastructure), and intensity of effort expected of the client and the region (dummy for East Asia and Pacific region). A task team. Although, in the past decade, social zero was assigned to social risks in the remaining risks have also been considered when catego- 44 projects, which only had environmental risks. rizing projects, the extent to which social risks Each project was rated on the significance of influenced the ratings was unknown prior to this social and environmental risks using a four-point evaluation. scale for the four criteria described in table 4.1 All of the 102 completed Bank projects in the Misalignment between categorization portfolio review sample triggered environmental and risk ratings safeguards, while 53 triggered social safeguards: IEG findings from the risk analysis indicate that 49 percent of the subset of completed projects categorization is not always determined by the 66 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S Box 4.1: Risks and Benefits Model IEG developed an analytical model to rank the environmental (MNI), the best documented performance indicator. This rating and social risks of each project along four parameters, using from the portfolio review reflects IEG’s assessment of the extent transparent criteria to rank each project on the basis of data to which the risks identified at appraisal have been mitigated. and documentation obtained for the portfolio review. The model Each project’s success in mitigating negative impacts was rated postulates risks (R) to be a function of Magnitude (M), Intensity as excellent (E=1.00), satisfactory (S=0.75), partially unsatisfac- (I), Duration (D), and Sensitivity (S), with separate indicators tory (PU=0.5), or unsatisfactory (U=0.25). On this basis, a measure for rating social risks (RS) and environmental risks (RE) along of the actual benefit (B) from the implementation of the WBG’s these four criteria. Data for estimating risks were obtained from safeguards for a specific project is estimated as: IEG’s review of appraisal and supervision documentation. The aggregate risk (R) is the sum of RS and RE, where B= MNI (RS +RE) RS = log (MS + IS + DS + SS) and RE = log (ME + IE +DE +SE) While B is only an ordinal indicator of the benefits of safe- guards implementation, it can be appropriately used to compare The risk model provides a modality to impute value to benefits benefits against costs, to analyze allocative efficiency and cost by weighting the environmental and social outcomes of WBG effectiveness of WBG, and client resources expended on meeting projects by the significance of environmental and social risks. For safeguards and sustainability objectives. this purpose we rely on the rating for Mitigating Negative Impacts Table 4.1: Indicators for Estimating Social and Environmental Risks Risks: High Substantial Moderate Low SOCIAL Magnitude (No. of project- >10,000 project- 1001 <=10,000 project- 101 <=1000 project- <=100 affected persons—displaced affected persons or affected persons or up to affected persons or up persons get 10 times the weight of >1,000 displaced 1,000 displaced to 100 displaced other project-affected persons) Intensity Physical displacement Economic displacement Workplace safety Community impacts Duration Permanent (beyond the (Late project life) (Early project life) <1 year or by project’s closing date) Mid-term review—closing >1 year mid-term review effectiveness date Sensitivity—Indigenous peoples Substantial risks to be Potential risks identified in Projects mainstreaming Targeted IP projects (IP) mitigated as per IP Plan IP Framework benefits to IP ENVIRONMENT Magnitude—Area affected Global, regional, or National or multiprovincial State or provincial Localized transnational Intensity Irreversible Severe Moderate Mild Duration >100 years >10–100 years >1–10 years <1 year (seasonal or intermittent) Sensitivity—Natural habitats Significant impact on Significant degradation Degradation other NH, Conservation and (NH) critical NH of NH parks or reserves rehabilitation of NH 67 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D riskiness of a project; neither is it based on use • In all, 12–16 percent of the sample projects of objective criteria to assess environmental and were overclassified, while 8 percent were un- social risks. The application of clearly specified derclassified, leading to a total classification risk indicators to the subset of completed error of 20–24 percent among category-A and projects, whose adverse impacts were fully -B projects alone2 when the assigned catego- known provided even stronger evidence of the ries were compared with the results from the weaknesses in the current practice of safeguards application of the risk indicators. categorization. Almost a third of projects with high-risk levels were incorrectly classi- IEG also found substantial regional variation in fied as category B, while one sixth of use of safeguard categories; however, regional Safeguards categorization category-A projects had much lower variations in distribution of environmental is not always risks. Based on the environmental and categories cannot be attributed to differences determined by the social risk ratings derived from IEG’s in categorization. East Asia and Pacific has a riskiness of a project. model: high proportion of category-A projects, while Latin America and the Caribbean seems to avoid • Of the 32 category-A projects in the sample, them, a practice confirmed by IEG staff survey one had moderately low environmental risks results. Econometric analysis shows that regional and no social risks, and was overclassified as composition has no statistically significant effect category A. on a project being category A when controlling • If the threshold were adjusted to include only for social and environmental risks, and network projects with high or moderately high risks, effects. Looking at all possible risk profiles, the 5 of the 32 category-A projects would appear likelihood of an East Asia and Pacific project to be overclassified, resulting in an inclusion being category A is only 2–4 percent higher than error of 5 percent of the sample universe. other regions. The substantial regional variation • Twenty-eight of the 97 projects had social in the distribution of environmental catego- and/or environmental risks that were high or ries appears to be driven more by the choice highly substantial, and on the basis of the risk of lending operations than by differences in ratings could have been classified as category categorization. A; in fact, 8 of these were not, which suggests an exclusion error of 8 percent of the sample Social risks drive project identification as universe. category A • The results do not change significantly if the IEG analysis confirms that social risks are a risk threshold is adjusted. The econometric significant determinant of classification as analysis also corroborates this result: for in- category A. Table 4.2 indicates that social and stance, the likelihood of a project being cat- environmental risks both influence project egory A is 72 percent if both environmental categorization although not necessarily equally. and social risks are highly substantial. The Among the 32 category-A projects, 22 had high likelihood increases to about 90 percent if or substantial social risks, and 19 had high or both environmental and social risks are very substantial environmental risks, 10 of these high. having both substantial social and environ- • On the other hand, 11 of the category-B proj- mental risks. Econometric analysis suggests ects had social and/or environmental risks that that a project with any social risk is 32 percent were in the three lowest ordinal lev- more likely to be category A than a project with Regional variations els on both environmental and social no social risks when controlling for environ- in distribution of risks, which would indicate overclas- mental risks and region and network effects. environmental sification of 11 percent of the sample, Taking into account gradations in environmen- categories cannot be which resonates with the observation tal and social risks3 for all possible risk profiles, attributed to differences made in chapter 1, based on the trends social risk is a stronger determinant of project in categorization. in figure 1.2. identification as category A than environmen- 68 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S tal risk (table 4.3). For instance, projects with high social risks and low environmental risks Table 4.2: Distribution of Projects by Significance of are 26 percentage points more likely to be Risks in Completed Bank Projects category A than projects with high environ- mental risk and low social risks.4 By contrast, Social risks Environmental risks environmental risk does not have a statistically Risks (n=53) (n=102) significant effect on projects being classified as High 4 0 category A. Category A Substantial 18 19 The Bank has moved from an environmentally Moderate 4 13 driven classification system toward one based Low 2 0 on social and environmental risks. When only Total 28 32 environmental risks or social risks are used to High 0 2 predict the likelihood of a project being category Category B A, while controlling for network and region, they Substantial 10 30 are both statistically significant on their own. Moderate 12 35 But the likelihood of a project being category A Low 3 3 is much higher for social risks than for environ- Total 25 70 mental risks in the same category (figure 4.1). Although social risk better predicts category A, High 4 2 differences in the predicted probabilities are Substantial 28 49 smallest when risks are low or very high and are All Moderate 16 48 greatest when risks are moderate or substantial, Low 5 3 giving an inverted U-shaped curve. Total 53 102 Infrastructure sector projects are more likely to Source: IEG portfolio review. Note: Category B includes five FI projects with moderate environmental risks. This table presents the be category A than noninfrastructure projects distribution of projects by safeguard category and significance of social risks based on the results from An infrastructure project is 37 percent more the application of the risk model to all 102 completed projects in the Bank portfolio review. likely to be category A than a noninfrastruc- ture project, when controlling for social and The Bank has moved environmental risk ratings and regional effects. possible risk profiles. For instance, from an environmentally The higher probability of classifying infrastruc- when both environmental and social driven classification ture projects as category A is evidenced for all risks are highly moderate (that is, in system toward one based on social and environmental risks. Table 4.3: Likelihood That a Project Is Category A (min-max) Social risks Low Moderate Substantial High Low 3%–8% 11%–23% 27%–45% 50%–69% Environmental Moderate 6%–15% 19%–35% 40%–59% 64%–81% risks Substantial 12%–25% 30%–49% 54%–72% 77%–89% High 21%–38% 43%–63% 68%–83% 86%–94% Source: IEG portfolio review. Note: Min-max is based on eight ordinal gradations of risk levels. 69 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 4.1: Predictive Probabilities for Category-A Projects Predictive probability for category A Gap between predicted probabilities for category A using environmental and social risks using social versus environmental risks 100% 30% 90% 86% 80% 77% 25% 73% 70% 66% 20% 61% 19% 19% 60% 17% 16% 52% 50% 47% 15% 13% 13% 40% 39% 34% 9% 10% 30% 27% 22% 20% 17% 6% 13% 5% 10% 7% 10% 4% 0% 0% Low-Low Low-High Mod-Low Mod-High Substantial-Low Substantial-High High-Low High-High Low-Low Low-High Mod-Low Mod-High Substantial-Low Substantial-High High-Low High-High Environment Social Infrastructure projects the upper three ordinal risk grades), not necessarily riskier. A risk-based assessment are more likely to be the likelihood of an infrastructure would also need to consider the context, includ- classified category A project being category A is 56 percent ing the institutional capacity of the implement- than other projects. versus 15 percent for a noninfrastruc- ing agencies, sensitivity of the location, and ture project. political economy effects on the project. The WBG, as a whole, would benefit from the From the empirical evidence we conclude that development and introduction of transparent (a) the current categorization system in use at criteria for assessing social and environmental the Bank is based more on magnitude of physical risks (as in IEG’s model) to ensure more consis- impact (infrastructure projects) than significance tent, risk-based categorization of the projects it of risks, and (b) social risks now have a substan- supports. tial effect on categorization, justifying renaming the labels as “environmental and social catego- Estimating Benefits from Risks ries” or simply “safeguard categories.” The risk ratings in IEG’s model provide estimates of the significance of environmental and social However, the Bank does not use transpar- risks and potential benefits in individual projects. ent, objective criteria to determine safeguard The higher the risk, the higher the potential categories, leading to inclusion and benefit from mitigating the risks. However, identi- The WBG would benefit exclusion errors, with an overall fying high risks at appraisal is only the first step from development tendency toward overcategoriza- in a long process toward realization of benefits. and introduction of tion, particularly among category-B These benefits only materialize when the risks are transparent criteria for projects. Use of transparent criteria fully mitigated. Careful supervision and monitor- assessing social and does not eliminate the room for ing helps to achieve intended results. The MNI environmental risks to judgment, but the judgment needs rating captures the effectiveness of this process ensure consistent, risk- to be risk-based, not driven by scale and provides a good measure of the extent to based categorization. alone. Large-scale physical works are which potential risks have translated into actual 70 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S benefits. For example, the riskiness of Africa Investing in client this implies better integration of the projects is at the Bank-wide average, but weaker capacity and mitigation objectives of safeguards MNI ratings compared with other regions (see enhancing positive with the objectives of enhancing social chapter 3) leads to significantly lower benefits. and environmental sustainability.environmental and MNI ratings are significantly higher in East Asia social effects can enhance and Pacific and, to a lesser extent, in South Asia, IEG’s evaluation identified some safeguards benefits. so their ability to realize safeguard benefits is constraints emanating from Bank greater. safeguard policies that could be addressed to improve the cost There is broad Safeguard benefits can be enhanced when effectiveness of environmental and support among staff performance in terms of mitigating negative social policies. The majority of Bank and management impacts is augmented by strengthening client managers interviewed reported that for the objectives of capacity and enhancing positive impacts. The clients find safeguards cumbersome, the safeguards and Risks and Benefits Model has been generated some policies being more contentious sustainability policies. to compare relative benefits against costs (see and needing further refinement. The next section). However, benefits accrue not just main issues identified revolved around client from mitigation of adverse impacts. Benefits can ownership, policy rigidity, and capacity (box 4.2). be augmented by investing in client capacity and enhancing positive environmental and social The current social safeguard policies appear effects. to create relatively greater discomfort among clients than environmental policies because Benefits and Costs of the WBG’s of the prescriptive and restrictive nature of the Safeguards and Sustainability Policies social safeguard policies and perceived rigidities Qualitative feedback on benefits and costs in policy application. Unlike the social safeguards, IEG found broad support among staff and the requirements of environmental assessment management for the objectives of the safeguards and sustainability policies. This was reiterated constantly during the interviews with WBG managers, and focus group discussions with WBG staff and NGOs. This finding is reinforced Figure 4.2: Cascade of Benefits from Safeguards by data from the Staff Survey 2009, which and Sustainability Policies covers all Bank staff. Overall, 61 percent of WBG staff rated safeguards policies and procedures Reduced vulnerability Local Improved liveihoods favorably in terms of its effect in delivering Enhanced citizen’s voice high-quality results to clients, a rating that is higher than for fiduciary (53 percent), resource Civil Sustainable resource management society Greater citizen ownership management (47 percent), and OPs and BPs (55 percent). The safeguards (which in the Staff Laws and regulations Client Strengthened institutional capacity Survey subsumed the Performance Standards) Triple bottom line: social, economic and environmental sustainability were rated higher by IFC staff (66 percent) and MIGA staff (61 percent) than by Bank staff (59 WBG Greater development effectiveness Reputational risks managed percent). Equitable resource use From its manager interviews and focus group Global Enhanced global public goods discussions, IEG also found wide recogni- tion that benefits from these policies accrue at multiple levels, including local beneficiaries, civil society, clients, the WBG, and in the form of global public goods (figure 4.2). However, 71 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D ties for the very people who need them most. Box 4.2: Feedback from Bank Clients and Managers Managers see this risk as most acute due to the lack of sensitivity to different contexts in Bank clients: the treatment of squatters by the Involuntary • The Bank always considers safeguard policies superior to the cli- Resettlement policy, and in application of ent country’s own laws and systems, reflecting lack of trust and the Indigenous Peoples policy, both of which undermining client ownership. have, in some instances, resulted in errors of • Social safeguards are of greatest concern, 82 percent reporting that exclusion. In one African country, the borrower Bank standards are significantly higher than country requirements. decided to drop one component of a loan that • Some also see the notification requirement of the International was to improve water supply and sanitation Waterways policy as an unnecessary imposition, especially for services in low-income areas because of the projects that will also benefit neighboring countries. inability to reach agreement on how to deal with illegal squatters. In this instance, dropping Bank managers: that component led to the denial of develop- • The primary concerns on Involuntary Resettlement arise from its ment benefits from a Bank-financed project to treatment of illegal squatters (which contradicts national laws), those urban squatters. The inability to distin- and from imposition of high compensation rates, which in some guish between projects designed to benefit instances have inflationary effects. squatters from those that are simply displac- • The policy on Indigenous Peoples works better for groups whose ing them can thus undermine development rights are recognized and enshrined in national laws; in other re- effectiveness. gions the term is alien, and rights and identity are often contested and politicized, and can give rise to ethnic tensions. In another example cited, the insistence on • Capacity constraints among clients need greater attention to ensure imposing the IP policy on an ethnic group, which sustainability. was not legally recognized as indigenous and • Staff capacity is weaker in some regions; there is greater need of was engaged in conflict with the government, technical oversight to ensure consistency in policy interpretation. derailed the process of project preparation. In one region, task teams complain about the lack of differentiation between projects that are designed to extend social services to indigenous peoples from projects that might adversely affect The current social are more open-ended, allowing greater them. Staff expressed concern that insistence safeguard policies appear room for dialogue and prioritization of on preparing additional mitigation plans in a to create relatively risks. However, one recurring theme project that is designed to benefit, not harm greater discomfort in the staff interviews was that affected them leads to duplication of effort and risks among clients than people have the ability to change their marginalizing indigenous peoples rather than environmental policies. minds about adequacy of compensa- bringing them into the mainstream of develop- tion and entitlements. In effect, social ment benefits. risks include risks from the project and political economy risks to the project from IEG’s survey of task team leaders and interviews affected people who are dissatisfied with its with Bank managers provided some evidence immediate effects, and who may then challenge of a deterrent effect of safeguard policies on the project, leading to costly delays. In contrast, lending. Client avoidance of Bank lending was environmental risks are risks from the most prominently reported by team leaders Lack of sensitivity to project but are less likely to be a threat from Latin America and the Caribbean (figure different contexts can lead to the project’s development outcome. 4.3) but was also found in other categories to missed development and regions: 38 percent of survey team leaders opportunities for In some instances, the lack of sensitiv- said that clients avoided a project or dropped the very people who ity to different contexts can lead to a component because of safeguards, while 18 need them most. missed development opportuni- percent reported that the project team revised 72 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S Figure 4.3: Impact of Safeguard Policies on Lending, by Region 70% Percentage of task team leaders 60% 50% 40% 30% 20% 10% 0% AFR EAP ECA LCR MNA SAR Total Client avoided WB project or component because of safeguards Project team revised project scope to avoid category-A classification Source: IEG Staff Survey. the scope of a project to avoid category-A classi- safeguards, is essential if the WBG is to create fication. Two-thirds of managers interviewed broader ownership among its clients for environ- reported that some clients had avoided or were mental and social sustainability. dropping a Bank project because of safeguard policies. At the same time, 15 percent had clients Assessing Costs and Benefits who actively sought out Bank lending because Assessment of the benefits and costs of safe- of the stamp of approval of the Bank’s safeguard guards and Performance Standards of the policies, which makes it easier for them to portfolio sample proved challenging because mobilize additional funding. IEG’s survey of IFC benefits of safeguards had not been systemati- staff also identified similar chilling effects on their cally monitored and documented, and environ- clients: 47 percent of investment officers and 56 mental and social costs were rarely available. In percent of environmental and social staff had the rare cases where benefits were recorded, encountered a situation where the client wanted there was no attempt to put a monetary value to avoid dealing with the IFC on a project or a on nonmarketed environmental goods and ser- part of a project because they considered that the vices. It is also very difficult to establish a coun- Performance Standards might be too expensive terfactual to determine the incremental value or time-consuming. While avoiding environmen- of WBG safeguard policies in relation to what tal and social risks through project redesign is would have happened without WBG involve- clearly desirable, the extent to which this leads to ment because very few projects contain base- missed opportunities that weaken development line information. Given these data constraints, effectiveness is also of concern. the assessment relied on qualitative description of benefits and costs, and benefit-cost The benefits of safeguards are undeniable, but reasoning to assess efficiency of re- Benefits of safeguards they do come at a cost. Evaluating these benefits source allocation from the perspec- had not been and costs makes it easier to demonstrate their tive of the WBG and clients. A stylized systematically monitored value and help identify ways of increasing benefit-cost model presented in the and documented, the cost effectiveness of WBG support. This end illustrates the kind of assessment and environmental does imply, however, that measurement of the and insights that could be drawn with and social costs were benefits accruing from, and costs incurred on adequate data. rarely available. 73 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Qualitative description of benefits safeguards and sustainability frameworks. For The review of project documents and field studies IFC and MIGA projects, these costs arise from the provided rich qualitative descriptions of benefits environmental and social corrective action plans and costs associated with safeguards from the developed in the course of appraisal, which the perspective of society, clients, and the WBG. client agrees to implement as a condition of IFC Benefits to society include: and MIGA support. In the absence of real cost data, corrective action plans provide a proxy of • Reduction in accidents or health damages, and client costs. Costs of safeguards are more difficult improved safety standards for the population to identify for Bank-financed projects since • Enhanced developmental opportunities they are subsumed under total client costs on through compensatory mechanisms, protec- safeguards, differing from country projects largely tion of use rights from common property re- in the extent to which additional analysis, consul- sources, and improved livelihood opportunities tation, and mitigation is required, and in terms of • Avoidance of harmful project and impacts of the rigor with which they are implemented. negative externalities. Bank costs for preparation and supervision of Client adoption of the WBG’s safeguard and safeguard elements in the sample projects are Performance Standards generates incremental a small fraction of the cost incurred by clients. benefits including: Bank costs include direct staff costs of environ- mental and social specialists5 and travel costs • Better management of environmental, social, for identification, appraisal, and supervision health, and safety risks of safeguard aspects of projects. Safeguard- • Improved community and government rela- related Bank costs are not directly identifiable tions and access to funding in the Bank’s cost accounting database because • Enhanced reputation, brand value, and mar- these costs are not recorded separately. The ket potential that is associated with improved costs reflected in table 4.4 are based on the sustainability performance, particularly for the costs attributable to environmental and social private sector. specialists from the project’s Bank budget only, which therefore do not capture the full costs For the WBG, the main benefit of the sustain- incurred on safeguards. Feedback to IEG from ability framework is recognition of its leader- the Regional Safeguards Advisors indicates ship role in setting and promoting benchmarks that full Bank costs incurred on safeguards for environmentally and socially sustainable being supervised by task teams ranged from projects, and management of reputational risks. $116,000 to $250,000 for category-A, and up to Historically, the WBG has been widely acknowl- $130,000 for category-B projects. 6 Additional edged as being in the vanguard of promoting costs include those incurred by other staff sound environmental and social policies; some and consultants on safeguards, as well as costs clients see the WBG as representing the “gold charged to other project codes. However, some standard” on these policies. Similar policies have category-B projects were also found to have been adopted by many IFIs (and more negligible safeguards expenditures, lowering Bank costs for preparation recently by private banks) to provide the median cost for category B. The median and supervision of assurance that the development cost incurred by the Bank on safeguards in table safeguard elements in activities they support are consistent 4.4, as a proportion of total Bank cost identifi- the sample projects are with globally accepted standards for able by IEG, was about 8 percent for category-A a small fraction of the environmental and social perfor- projects and 4 percent for category-B projects cost incurred by clients. mance. in the sample projects. The median Bank cost on safeguards was 3.3 percent of the prepara- Clients incur substantial costs on mitigation tion budget and 7.6 percent of the supervision measures to address risks highlighted by the budget. The median for safeguard costs incurred 74 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S Table 4.4: Average and Median Costs for Safeguards (US$) Bank costs (n=60) Bank client costs (n=53) IFC costs (n=37) Average Median Average Median Average Median A 72,412 51,061 19,230,200 8,357,000 254,450 60,264 B 45,675 22,876 5,168,489 4,031,200 24,654 12,195 Sample total 59,766 38,700 13,544,300 5,920,000 62,953 19,062 Source: World Bank and IFC data. Note: These cost tables include data from completed and active projects in the portfolio sample and provide an incomplete picture of full costs on safeguards/Performance Standards at closure. World Bank are based on 60 projects: 28 completed projects (22 category A, and 16 category B), and 32 active projects (15 category A, and 17 category B). IFC data are based on 37 projects: 6 completed projects (all category B); and 31 active projects (6 category A, and 25 category B). While they are instructive in providing the relative proportion of safeguard costs, and in comparing costs of individual projects with risk-adjusted benefits, they are not appropriate for drawing inferences on resource allocation for safeguards. by Bank clients was almost $6 million, which is Efficiency of safeguards resource allocation 5 percent of the loan amount, and 3 percent from the Bank’s perspective of total project cost from the sample of Bank From a resource management perspective, a projects on which client data could be obtained. simple test of the efficiency of the WBG’s sustain- Nonetheless, client costs on safeguards consti- ability framework is whether the costs incurred tute the bulk of total expenditure on safeguards. are allocated in proportion to the environmen- tal and social risks of projects and achieve the For IFC, the median cost reflected in table 4.4, desired outcomes. Efficiency was assessed along was 13 percent of total IFC cost for category-A a quadrant of the risk-adjusted benefit (B) and projects and 4 percent for category-B projects. costs, with the separation between high and Client costs were not available for IFC. The low based on median values of B and costs for majority of projects in the IFC sample are still safeguards and Performance Standards to the active. IFC data for the benefit-cost analysis WBG and clients. Table 4.5 shows the distribu- are based on 36 projects from the portfolio tion of WBG performance along the benefit-cost review sample for which cost and MNI data quadrant. were available, and may not be representative. Additional data obtained from IFC for a set of 30 Analysis of the risk-adjusted benefits and Bank projects in its current portfolio gave an average costs on safeguards in the sampled projects for cost of $163,410 and median of $129,583 for which cost data are available do not provide category-A, and average of $51,814 and median clear evidence of allocative efficiency, particu- of $36,450 for category-B projects. A compari- larly for category-B projects. The analysis in son of these two sets of safeguards-related costs figure 4.4 shows that category-A projects tend to suggests that actual costs may be much greater incur higher costs and yield higher benefits, 15 of than that available for the cost-benefit analysis 22 being at or above the median level of benefits. from the historical portfolio. Category-B projects are more dispersed. Most category-B projects had lower benefits, but 6 of The benefits from the safeguards and sustain- 10 projects in this sample incurred costs at or ability frameworks are computed from estimated above the median level. All three projects in the environmental and social risks associated with extreme upper-right corner are “high risk–high each project, as described in box 4.1, and the MNI reward” infrastructure projects7 that incurred rating, which reflects IEG’s assessment of the high costs and yielded high benefits. The project extent to which the risks identified at appraisal in the lower-right corner had high costs but have been successfully mitigated. yielded low benefits because of unsatisfactory 75 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D tation and improved supervision by project staff Table 4.5: Distribution of Projects on Benefit-Cost are key factors driving cost-effective mitigation Quadrant of environmental and social risks. Most of the projects fell into the “low cost – low benefit” Bank (percentage) IFC (percentage) or “high cost – high benefit” quadrants. With Benefit–cost quadrant (n=35) (n=36) two exceptions, all the category-B projects were in the “low cost – low benefit” quadrant, High benefit—low cost 26 19 while four category-A projects were in the High benefit—high cost 29 33 “low cost – high benefit” quadrant. Only two Low benefit—low cost 23 28 projects achieved low benefits despite high Low benefit—high cost 23 19 client expenditures. Source: IEG risk analysis. Note: These figures are based on costs incurred by the WBG and exclude client costs. Efficiency of resource allocation at IFC Analysis of costs incurred by IFC to implement the Performance Standards in relation to the IFC’s allocation of ESHS MNI. Greater attention to outcomes distribution of environmental and social risks resources has been would have improved its benefit-cost shows that IFC’s allocation of ESHS resources broadly aligned with ratio. has been broadly aligned with risks, and that risks, and the alignment the alignment has improved since introduction has improved since The analysis of Bank client expendi- of the Performance Standards. The greatest introduction of the tures shows much greater allocative costs have been incurred on projects facing Performance Standards. efficiency (figure 4.5), higher costs relatively higher risks and higher benefits (see being incurred on projects with figure 4.6).8 Of the six category-A projects in relatively high benefits. An in-depth the chart, five lie in the upper-right quadrant of analysis of specific projects in the different higher benefits and costs, while the sixth falls quadrants suggests that better client implemen- just below the median of risk adjusted benefits. Both projects in the upper-right quadrant with the highest benefits and costs are from extractive industries. Projects implemented with Performance Standards also show more Figure 4.4: Bank Benefits and Costs for Environmen- efficient allocation of resources compared with tal and Social Safeguards projects under the safeguards policy. These 20 data are time sensitive, however, subsequent 18 analyses over the course of this evaluation indicated that, even with updated costs data, 16 the distribution of projects in the chart did not Risk adjusted benefit 14 change significantly. 12 10 Client costs for IFC projects could not be analyzed 8 by IEG because relevant information was unavail- 6 able. However, client feedback reveals that 4 incremental benefits were perceived as commen- 2 surate with the costs. Interviews with a small 0 sample of clients indicate that incremental costs 0 50 100 150 200 250 associated with IFC’s sustainability framework WB safeguard costs (thousand US$) were perceived as necessary and appropri- A-Category B-Category ate. IEG’s staff survey and interviews with IFC Source: IEG. managers also confirm the perception that incremental costs associated with the implemen- 76 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S tation of the Performance Standards appear to be Figure 4.5: Bank Client Benefits and Costs for proportionate to the benefits received. Environmental and Social Safeguards Current arrangements for safeguards supervi- 20 sion have created inefficiencies in resource 18 allocation. The contrast between the ability 16 of the Bank and IFC to target institutional Risk-adjusted benefit 14 resources efficiently toward higher-risk projects appears to be related to the differing administra- 12 tive arrangements for environmental and social 10 services. In IFC (as in MIGA), environmental 8 and social supervision is funded off the top and 6 not left to the discretion of investment officers. 4 Supervision efficiency and quality has improved 2 markedly. By contrast, safeguard oversight for 0 most category-B and -FI projects at the Bank 0 20 40 60 80 has been delegated to the sector manage- Client safeguard costs (million US$) ment units, and supervision services are paid A-Category B-Category through cross-support in all regions. Delega- Source: IEG. tion of authority also means that the respon- sibility for determining loan covenants rests with the delegated sector managers. While this permits a demand-driven approach, delegation reduces the ability of social and environmental Figure 4.6: IFC Safeguard Benefits and Costs sector managers to influence safeguards design, ensure that critical actions are covenanted, and 20 deploy their staff resources strategically where 18 they are most needed. 16 Risk-adjusted benefit 14 IEG’s risk analysis provides evidence of misallo- cation of resources and IEG’s findings on 12 supervision illustrate the neglect of safeguard 10 supervision in one-third of Bank projects that 8 trigger safeguards. In recent years several regions 6 (Africa, Middle East and North Africa, and South 4 Asia) have moved to more structured arrange- 2 ments with off-the-top payment for operational 0 support in fiduciary areas, which provides the 25 50 75 100 145 310 990 institutional space for regional managers to plan IFC budgetary cost for ESHS (US$ thousands) and allocate staff and budget resources strategi- Safeguards PS cally. The arrangements of environmental and social operational support for safeguards have Source: IEG; cost data from IFC. not been changed, and due to delegation of responsibility for category-B projects to sector management units, oversight does not rest with the social and environmental sector managers poor and uneven supervision quality, Current arrangements for in the regions. Organizational responsibility and particularly for category-B and -FI safeguards supervision incentives are not well aligned with environ- projects. This deficiency requires have created inefficiencies mental and social supervision needs, leading to urgent attention. in resource allocation. 77 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Delegation reduces the Stylized Model of Benefits and provided in greater detail in annex 7, show that ability of environmental Costs the estimated benefits from the environmental and social sector To illustrate the insights that could and social policies are greater than the incremen- managers to influence be drawn from a benefit-cost analysis, tal costs in every case. safeguards design, ensure a stylized benefit-cost model was that critical actions estimated for a scenario with IFC A similar benefit-cost model was estimated for are covenanted, and intervention and another without IFC two stylized Bank projects. The results from the deploy staff resources intervention. Model costs were based Bank model show that the estimated benefits strategically. on expert opinion and information from the environmental safeguards far outweigh from the literature and validated with the incremental costs. In the case of social examples from IFC project portfolio and discus- safeguards the benefits do not exceed the costs, sion with IFC staff. A benefit-transfer method was but a number of benefits cannot be quantified used to estimate benefits from safeguards. The (box 4.4). results from the IFC stylized model (box 4.3), Box 4.3: Quantitative Estimates of Costs and Benefits from Stylized IFC Projects The stylized model estimates the net benefit-to-cost ratio for have been estimated applying a discount rate of 10 percent and two hypothetical IFC interventions: a gold mine in West Af- a lifetime of 25 years for each project. The results show that the rica, employing 500 people; and a General Manufacturing and safeguards had positive payoffs in every case when compared Services (GMS) project in the Middle East and North Africa, with a situation without IFC interventions. employing 1,500 people on a greenfield site. Benefit-cost ratios IFC Interventions for Which a Benefit-Cost Ratio Has Been Estimated Gold mine project category A GMS project category B Labor and working conditions Reduced fatalities and accidents at the workplace Pollution prevention and abatement Reductions in spills of toxic materials such Reductions in biological oxygen demand as cyanide and in emissions of volatile organic compounds Community health and safety Reduced fatalities and accidents from the increased flow of traffic generated by the project Land acquisition and resettlement Time savings due to less disputes on — resettlement Benefit–Cost Ratios for the Selected IFC Interventions Gold mine project category A GMS project category B Labor and working conditions 4.8 7.7 Pollution prevention and abatement 3.6 5.1 Community health and safety 1.5 8.2 Land acquisition and resettlement 4.4–6.5 — Source: IFC. 78 R I S K S , B E N E F I T S , A N D C O S T S O F T H E W B G ’ S S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S Box 4.4: Quantitative Estimates of Costs and Benefits from Stylized Bank Projects The stylized model estimates the net benefit-cost ratio for two and 5 percent for the transport project, given the long-term archetypal Bank interventions: a transport project in Sub-Sa- nature of benefits from the forest ecosystem being valued. The haran Africa, through virgin rainforest and degraded savan- results show that the safeguards had positive payoffs when nah, adversely affecting 2,000 people; and an urban water and compared with a situation without Bank interventions. How- sanitation improvement program in Asia involving small-scale ever, the absence of policies on labor and community impacts resettlement. Benefit-cost ratios have been estimated applying reduces the net benefits that can be attributed to social safe- a discount rate of 12 percent for the water and sanitation project guards policies. Bank Interventions for Which a Benefit-Cost Ratio Has Been Estimated Policies triggered African road project Asian water sanitation project OP 4.01: Environmental Assessment Strategic Environmental Assessment EA with individual subproject EIAs and conducted with supplemental EMS/EMAPs EMPs. Support to 3 project management for main road sections units to improve supervision. OP 4.04: Natural Habitats Road sections pass through virgin Not applicable rainforests and protected areas OP 4.36: Forests Increases in legal and illegal logging Not applicable operations anticipated OP 4.10: Indigenous Peoples (IP) 1,600 IPs negatively affected by project Not applicable and compensated OP 4.11: Physical Cultural Resources “Chance Find” procedures and manual “Chance Find” procedures and manual prepared prepared OP 4.12: Involuntary Resettlement 400 people lacking clear property title are Land acquisition, restoration of livelihoods, relocated and livelihoods restored and resettlement for 1,000 project-affected households and 100 businesses Benefit-Cost Ratios for the Selected Safeguard Policy Interventions Policies triggered African road project Asian water sanitation project Environmental safeguards (OPs 4.01, 4.04, $330 million / $9 million = B:C ratio of 37 Costs = $0.75 million. Benefits in total are 4.36, and 4.11) $20 million. If at least 13% of health and tourism benefits are due to safeguards then B:C>1. Social safeguards Costs = $3.2 million Costs = $13.3 million. Benefits range (OPs 4.10 and 4.12) Benefits = $2.6 to $4.3 million; from $14.9 to $47.2 million, depending on Assuming successful livelihood starting income level of project-affected restoration, B:C ratio ranges from 0.8–1.3 persons and benefits from restoration of livelihoods. B:C>1if income of those persons affected is less than 75% of the national average, and is even higher for lower income levels. Total combined B:C ratio $333.5 million / $12.2 million = B:C ratio of 27.3 79 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Such a systematic framework for assessment of significant benefits for the mitigation of environ- benefit-cost analyses can be used to evaluate: mental and social risks of projects even though these are not systematically measured or quanti- • The costs associated with each safeguard or fied. Budgetary resources devoted by the IFC to impact area its sustainability framework are being allocated • The incremental costs and benefits of WBG relatively efficiently. However, the evidence on requirements versus those of country systems Bank resource allocation for safeguards work • Opportunities for cost savings from, for ex- is more mixed. Bank clients tend to allocate ample, alternative project design parameters, resources efficiently in meeting safeguards technology choices, or siting options. requirements. The same could not be assessed for IFC clients because data were not available. Summary of Main Findings Benefit-cost analysis can provide useful insights IEG developed and applied a model to test the into environmental and social performance. effectiveness of the project categorization system Better monitoring, documentation, and reporting as a proxy for safeguard risks. There was consid- of environmental and social impacts are needed erable variability in the risk profiles leading to to improve the quality of benefit-cost analysis. inclusion and exclusion errors in categorization. The WBG should systematically integrate indica- The risk model enabled scrutiny of the relative tors of environmental and social performance effects of social and environmental risks on within project results frameworks and collect project categorization and provided evidence data to monitor and evaluate safeguards and that social risks are strong, indeed marginally Performance Standards. stronger, determinants of projects classified as high risk. The current practice of signaling high IFC has taken effective steps to strengthen risk if either social or environmental risks are high supervision to ensure that environmental and is appropriate. The assessment of risks, which is social benefits are achieved. However, the integral to the safeguards model, is a proxy for Bank lags significantly behind. Resolving this potential benefits that would materialize if the deficiency at the Bank will require (1) revamping mitigation measures were fully implemented. the policy framework to correct the overempha- The focus of safeguards thus needs to evolve sis on compliance with frontloaded procedural beyond perfecting the design of action plans requirements and inadequate attention to toward ensuring effective implementation with supervision, monitoring, and completion report- clear allocation of resources and accountabilities ing; (2) development and application of transpar- for safeguards supervision. ent criteria for risk assessment and safeguards categorization; and (3) revamping the incentives, The assessment of benefits and costs shows resource management, and accountability that the WBG’s safeguards framework generates arrangements for safeguards supervision. 80 Chapter 5 Evaluation Essentials • The thematic coverage of the Per- formance Standards is more rel- evant than the safeguards suite to the WBG’s investment project portfolio. • Current World Bank social safe- guards do not provide adequate coverage of community impacts, labor and working conditions, and health, safety, and security issues at the project level. • The Bank’s approach to country systems for safeguards needs to be substantially revised. • The safeguards approach is heavily frontloaded with manda- tory requirements at entry but has weaker guidance and systems for effective implementation. • The instruments and practices in- troduced by IFC in parallel with the Performance Standards have im- proved implementation. • Lack of disclosure and independent verification of monitoring by IFC's clients are serious concerns. Road running through a deforested area in Bhutan. Photo by Curt Carnemark, courtesy of the World Bank Photo Library. Safeguards and Sustainability Frameworks: Policy Issues T his chapter relies on comparative analysis to examine the effects of the policy frameworks on implementation results, and explores ways of ad- dressing policy constraints. This is not designed to be a comprehensive analysis of the WBG’s safeguards and sustainability policies. Rather, it limits the examination to the systemic constraints giving rise to the shortcomings and constraints previously identified and the solutions being tried to address them within the WBG, or in other multilateral development banks (MDBs). Five issues are covered in this chapter: lending portfolio gives some indica- The thematic coverage (1) relevance of the policies and Performance tion of the relevance of these policies of the Performance Standards to the current portfolio; (2) lessons to the portfolio. Safeguards data from Standards is more from the use of country systems; (3) compara- the 10-year portfolio for the Bank, and relevant to the WBG’s tive analysis of MDB safeguards and sustain- from the 3 years since the introduc- investment project ability policies; (4) lessons from other MDBs tion of Performance Standards at portfolio than the and IEG evaluation on social safeguards; and IFC are shown in the two charts in policies in the current (5) alternative paradigms for environmental figure 5.1.1 The data indicate that safeguards suite. and social risk management within the WBG. the thematic coverage of the Perfor- This chapter is diagnostic and is not intended mance Standards is more relevant to to prescribe solutions. It will, however, identify the WBG’s investment project portfolio than policy and institutional constraints that need the policies in the current safeguards suite, due to be addressed and, where possible, identify to the addition of explicit provisions on labor solutions derived from the comparative impacts, community impacts, and pollution analysis. prevention and abatement. Relevance of the Safeguards and The Environmental Assessment policy is triggered Performance Standards to the Current by 72 percent of the investment lending portfolio Portfolio and subsumes pollution prevention issues; the The frequency with which the safeguards and Involuntary Resettlement policy is triggered by 30 Performance Standards are triggered by the percent. However, 6 of the 10 safeguard policies 83 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure 5.1: Safeguards and Performance Standards in WBG Portfolio Safeguards triggered in World Bank projects, FY1999–2008a Performance standards triggered in IFC investments, FY2007–2009b 1% Disputed areas 2% Indigenous peoples (PS7) 6% Dam safety 8% Cultural heritage (PS8) 6% Forests 14% Biodiversity (PS6) 6% International waterways 8% Pest management 23% Land acquisition/Involuntary resettlement (PS5) 11% Natural habitats 44% Community health and safety (PS4) 12% Physical cultural resources 49% Pollution prevention and abatement (PS3) 17% Indigenous peoples 52% Labor and working conditions (PS2) 30% Involuntary resettlement Environmental and social assessment 72% Environmental assessment 53% and management (PS1) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of projects n=2,459 Percentage of investment projects n=506 Sources: World Bank and IFC databases. a. Bank data are based on results for all 2,056 investment projects in the portfolio, approved in fiscal 1999–2008. This excludes 439 development policy loans and structural adjustment loans from the same portfolio; 6 percent of the structural adjustment loans, approved prior to Sept 2004 and that triggered OP 4.01 (Environmental Assessment), are reflected in this chart. IFC figures are based on the entire IFC portfolio, including FI- and C-category projects. b. IFC data depicted are an underestimate as they do not portray the Performance Standards triggered by FI subprojects. were triggered by less than 12 percent of the private sector clients, many investment projects Bank’s investment lending portfolio (figure 5.1).2 in the Bank resemble IFC’s real sector projects. There is no obvious reason to presume that In contrast, the first four Performance Standards community and labor impacts are not relevant are triggered by about half of IFC’s total portfo- to the Bank’s portfolio. lio. The relevance of these four standards increases to over 90 percent among category-A Comparing portfolios helps to identify opportu- and category-B operations. nities to increase policy relevance by harmoniz- ing thematic coverage, and to improve efficiency The policies that are more frequently triggered by learning from each other’s practices. Labor at the Bank, and are common to both Bank issues, such as occupational safety and working and IFC, are triggered in roughly similar conditions, are just as relevant to the public proportions. The Environmental Assessment sector as to the private sector. Bank assistance policy affects more than half, while Involuntary to help its borrowers deal with retrenchment or Resettlement is triggered by 23 percent and employment effects is already being addressed Biodiversity by 14 percent of IFC’s portfolio, in the Bank’s portfolio but is not covered by which is similar to that in the Bank’s portfolio. the safeguards policies, leading to some missed But the Performance Standards on Labor and opportunities (see the Dhaka transport example Working Conditions; and Community Health, in box 3.2). Attention to community impacts Safety, and Security apply to about twice as ensures that the people directly affected by a many projects, and to over 90 percent of real project are better off as a result of the project. sector projects. While there are some differ- A focus on adverse community impacts also ences between the priorities of public and provides an entry point to engage borrowers 84 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S on social impacts in country contexts where an the country level, which is considered appropri- exclusive focus on indigenous peoples or on ate for small countries with projects identified gender issues may not be culturally or politically that would be piloted based on the outcomes feasible. An integrated assessment of community of the SDR process.” But because of more impacts, as already encouraged under IFC’s significant differences with Bank policies and Performance Standards also increases the procedures, the piloting of Involuntary Resettle- efficiency of social assessments. On the other ment and Indigenous Peoples was avoided hand, the Bank’s policy suite has stronger entirely in the first phase, either through project provisions on dam safety, which is also relevant design, or by simply applying normal Bank to IFC. safeguard procedures. In the second phase, involuntary resettlement is being piloted in Lessons from the Use of Country Systems the parastatal corporations. However, even in In 2005 the Paris Declaration on Aid Effective- Brazil, the Indigenous Peoples Policy has not ness established global commitments from been triggered. IEG visited three of the six first- donors and partner countries to improve the phase countries involved—Egypt, Romania, and management and effectiveness of aid in reducing Tunisia—to see how well country systems were poverty and inequality. Subsequently, the World being implemented. Clients were contacted Bank launched a pilot program3 to promote the a second time after eight months to assess use of country systems (UCS) for environmental progress on the country systems since the field and social safeguards. Managers interviewed by visit. In addition, IEG undertook a desk review IEG, including some involved in developing the of other UCS pilots and included a question country systems approach, said that the rationale concerning UCS in its manager interviews. This was to scale up development impact by encourag- information was used to prepare a background ing the use of improved systems for government paper for the evaluation. expenditures to increase country ownership, build institutional capacity, promote donor Initial borrower ownership of the UCS pilot harmonization, and increase cost effectiveness scheme was mostly positive, but interest has for both the Bank and the borrower. Although dissipated. Participating governments wanted to these objectives are still relevant, the country get away from the use of dual systems and hoped systems approach adopted for the safeguard that the UCS approach could be extended to policies has proved to be too self-limiting and additional sectors and projects. However, recent not sufficiently robust and flexible for scaling up, experience suggests that the anticipated time and has lost ownership among Bank staff and and cost savings in the processing of subsequent their clients.4 operations have not materialized because new SDRs5 have been required for subsequent Seven projects in six countries were included projects in the same country. Client feedback in the initial phase of the UCS pilot program. regarding the UCS pilots indicates that there is All piloted Environmental Assessment and an inconsistency between client expectations of two triggered Physical Cultural Resources. the purpose of UCS and that presumed by the By December 2009, another eight pilots had Bank. For example, client expectations that Bank been initiated in seven more countries. These safeguard responsibilities would be transferred included three corporate systems (in Brazil, to the borrower did not occur. Management has India, and South Africa) and two proposed clarified that this was never the intention of the country-level pilots (in Mauritius and Croatia). A pilots. Anticipation that UCS could automatically state-level pilot planned in Brazil appears to have be applied to subsequent Bank-financed projects been dropped. A progress report to the Board has likewise been frustrated. The benefits of the in 2009 describes the country pilots as follows: UCS pilots to clients thus remain unclear. Even “Two SDRs [safeguards diagnostic reviews], in countries like China, which IEG’s evaluation Croatia and Mauritius, are being conducted at confirms as having one of the best records on 85 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Although initial safeguard policies, have declined to were found to have irreconcilable differences borrower ownership of participate in the UCS experiment. with national laws. the UCS pilot scheme • Harmonization. Stakeholder concern that was mostly positive, Within the Bank, enthusiasm for the UCS approach would water down the safe- interest has dissipated. participating in the UCS pilot has also guard policies turned out to be incorrect. On waned. Some country directors and the contrary, the Bank has insisted that coun- sector managers are concerned that try safeguard policies be brought up to the UCS will increase the cost and time of project Bank’s standards, rather than starting from the preparation and supervision, while increasing country’s own systems and accepting compro- reputational risk. Of the regional safeguards mises in equivalence. Because of the rigidities advisors and the environment and social in the policy (OP 4.00) governing the pilots, development sector managers interviewed which leaves no room for experimentation, its by IEG, three-quarters maintain that the right uptake has been very slow. approach is not being followed, and not one • Social safeguards. The social safeguard believes that the UCS approach in its present policies were found to have irreconcilable form can be scaled up. They perceive that the differences with national laws in UCS pilot current costs of UCS for safeguards outweigh countries. Not a single pilot country was found the potential benefits, despite some positive equivalent to the World Bank on social safe- aspects of the initial pilots. guards in the first phase, and in the second phase there has been more headway in cor- Although the pilots are ongoing and additional porations than at the country level. projects have been added, the UCS program • Downstream issues. Many of the challenges objectives are only partly being achieved, and in associated with the UCS are the same as those several areas performance has fallen short: associated with the application of regular Bank safeguards. Often these have more to do with • Strategic scope. The countries and projects “downstream” issues such as implementation selected were not a representative sample monitoring, enforcement, and compliance. of Bank operations, and UCS was limited to • Rigidity. More generally, some staff members safeguard areas where country policies were interviewed by IEG affirm that an underlying already close to the Bank’s and institutional ca- problem with the efforts to apply country sys- pacity and government commitment were rela- tems, to date, is the highly process-oriented tively strong. The second-stage pilot projects nature and rigidity of the Bank’s own safe- have been extended to parastatals using cor- guard policies. porate systems6 rather than country systems.7 • Decentralization and devolution. Scaling Continuing to pursue individual projects and up of UCS is most challenging. In the near policies rather than adopting an integrated term, unless there is a change in the UCS pol- countrywide or sectorwide approach to ad- icy, there will be a continued need for intensive dress capacity, however, is proving to be of Bank supervision to ensure compliance with little value. the UCS policy. This may continue to frustrate • Thematic scope. The UCS pilots were unable some borrowers, who expected UCS would to apply the entire safeguards policy suite in lead to greater safeguards-related responsibili- any country. A progress report to the Board ties vested in the clients, with the Bank’s role states that the UCS approach has not worked being more supervisory in nature. for social safeguards in any country. The only • Cost implications and ownership. The attempt to pilot use of the resettlement policy incremental preparation cost to the Bank was in Jamaica, but this was for a project that for UCS projects was expected to decline as did not have any resettlement impacts, so the the fixed cost of diagnostic work was shared approach could not be tested. In all other UCS among more projects and sectors. However, pilot countries, the social safeguard policies the Bank’s experience indicates that this will 86 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S not be the case, especially when the invest- A quick comparison of the environ- The social safeguard ments fall in other sectors and/or involve mental and social policy frameworks policies were found additional or different safeguard policies. Bor- depicted in table 5.1 reveals three to have irreconcilable rowers do not perceive significant benefits overarching differences. Other MDBs differences with from UCS, so ownership is low. (except the African Development national laws in UCS Bank) tend to have an integrated pilot countries. It does not appear likely that the UCS program can policy framework, an umbrella policy be mainstreamed in its present form. The method on environmental and social sustain- prescribed in OP 4.00 has not worked satisfactorily ability, and a relatively balanced treatment of and has led to a lack of ownership by Bank staff social and environmental issues. and borrowers. While there is a general consensus that the concepts underlying the use of country From policy innovation to bureaucratic inertia systems are sound, the piecemeal approach, at the Bank which focused on individual projects and policies In the 1980s and 1990s, the Bank was the in the UCS pilots, appears unworkable and needs trailblazer in policy innovation. Many of the a major redesign for it to be successfully scaled up. individual safeguard policies were originally National systems can and should be used, where developed at the Bank and then emulated by possible, in some countries, in some sectors, and other MDBs. The Bank was the first to acknowl- for some safeguards, particularly environmental edge the need for enhanced accountability and assessment. Nonetheless, significant revisions will established the Inspection Panel in 1995. IFC and need to be made to the policy framework and MIGA followed suit, as did the Asian Develop- approach before country systems can be used for ment Bank (ADB) and EBRD, but safeguard policies. they all refined the accountability The piecemeal approach mechanism, introducing a conflict to UCS, which focused At the same time, the UCS pilots’ inability to resolution mechanism alongside the on individual projects address social safeguards brings home the accountability function. The Bank and policies, appears inherent weaknesses in the safeguard policies has retained the original institutional unworkable and needs a themselves and in the lack of a comprehensive model and lacks a grievance redress major redesign for it to framework that addresses environmental and mechanism. be successfully scaled up. social issues equally. Gaps in both equivalence and acceptability will likely remain too wide to The procedure for safeguard policy revision permit UCS to be scaled up unless the Bank at the Bank has proven so cumbersome and modernizes its own policies. time-consuming that there is great reluctance to revise and improve the policies even when the MDB Safeguard and Sustainability lessons of experience indicate that this would be Policies beneficial in achieving the broader objectives of Overarching differences with policy environmental and social sustainability. During frameworks of other MDBs the review period, the safeguard policies were Feedback from clients and NGOs reveals that the converted individually from operational directives WBG has significantly improved environmental to operational policies (see table 5.1). Among the and social results, as compared with the 1990s. changes adopted during that process were the This is most visible among the high-risk category- increased attention to economic displacement A projects, where the Bank’s experience provides and livelihood restoration under the policy on important lessons for other MDBs. But challenges Involuntary Resettlement, and the recognition of remain among category-B and financial interme- free, prior, informed consultation as a requirement diary projects and, within the World Bank, on the under the Indigenous Peoples policy. However, growing number of projects that rely on environ- staff who participated in these revisions informed mental or social policy frameworks, leaving risk IEG that they found the process extremely assessment to the implementation phase. cumbersome and inefficient. Consequently, they 87 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table 5.1: Comparison of WBG Safeguards and Performance Standards with Other MDBs IFC/ MIGA Policy and Performance Asian Standards on African Development Inter-American Social and EBRD Environmental Development Bank Safeguard Development World Bank Environmental and Social Policy Bank Group Policy Bank Safeguard Stability and Performance Safeguard Statement Sustainability Policiesa (2006/2007) Requirements (2008a) Policies (July 2009) Standards (2006) PS1: Social and PR1: Environmental and SR4: Special Env & soc Environmental Social Appraisal and Requirements Assessment and Management for Different Management Financing System PR 9: Financial Modalities Intermediaries 4.01Environmental PS6: Biodiversity PR6: Biodiversity Policy on the SR1: Environment Environment Assessment (1999) Conservation Conservation Environment and Safeguard and Sustainable and Sustainable (2004) Compliance Policy 4.04 Natural Natural Resource Management of Living (2006) Habitats (2001) Management Natural Resources Policy on Integrated Water Disaster Risk Environmental 4.36 Forests (2002) PS3: Pollution PR3: Pollution Prevention Management Policy Resources Prevention and and Abatement Management 4.09 Pest Disclosure of Abatement (2000) Management (1998) PR8: Cultural Heritage Information Policy PS8: Cultural 4.11 Physical Agriculture Heritage Cultural Resources and Rural (2006) Development Sector (2000) 4..37 Safety of Dams (2001) 4.12 Involuntary PS5: Land PR5: Land Acquisition, Involuntary SR2: Involuntary Involuntary Resettlement (2001) Acquisition Involuntary Resettlement Resettlement resettlement Resettlement Policy and Involuntary and Economic Policy (November 4.10 Indigenous Resettlement Displacement 2003) SR3: Indigenous Operational Policy Peoples (2005) Peoples on Indigenous PS7: Indigenous PR7: Indigenous Peoples safeguards Peoples (IPP) Peoples – Operating Social PR2: Labor and Working Guidelines (2006) PS2: Labor and Conditions Working Conditions PR4: Community Health, PS4: Community Safety and Security Health, Safety and Security PR 10: Information Disclosure and Stakeholder Engagement 7.50 International Waterways (2001) Legal 7.60 Disputed Areas (2001) Source: IEG. Notes: PS = Performance Standard, PR = Performance Requirement, SR = Safeguard Requirement. Four multilateral development banks have policies on gender, although these are not considered safeguards. IFC integrates gender issues under the guidance for PS1. a. Except for pest management, all Bank Operational Policies (OP) have accompanying Bank Procedures (BP). 88 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S have little incentive to attempt further revisions which has been reviewing compliance with all even when practical experience demonstrates the Bank policies, not just those labeled safeguard need for further refinement. policies. The goal of 100 percent compliance with the safeguard policies has proved elusive, When the Bank created the suite of safeguard despite the additional resources allocated for policies, it made public commitments of 100 it. Yet the priority given to mitigation effectively percent compliance with these “do no harm” crowded out attention to other social impacts on policies, a goal that ultimately proved unrealistic. local communities,9 including gender impacts in At that time, the Bank’s safeguard policies suite Bank-supported projects, as shown by a recent excluded some existing policies that addressed IEG evaluation. The label “safeguards” has also social impact—including Sociological Appraisal, creates an artificial barrier precluding attention to which is a part of the Bank’s policy on Project emerging environmental themes such as climate Appraisal (OMS 2.20), and the policy then on change and occupational health and safety under Gender and Development (OP 4.20). In addition the safeguards framework. to those two, the Bank subsequently developed guidance on financing severance payments in Most of the safeguard policies provide clearer Bank operations that involve labor retrench- guidance on procedures and mandatory require- ment in the public sector.8 The WBG has jointly ments for risk assessment and development prepared and issued EHS guidelines. However, of risk mitigation plans prior to appraisal but Bank provisions on these two issues do not receive relatively little guidance on the implementa- adequate attention by safeguards practitioners tion phase.10 Since 1999, the Bank has dedicated even in projects where these risks are relevant. Of additional resources managed by QACU, to the 23 projects in the portfolio sample approved ensure high-quality safeguards performance. since the EHS guidelines were prepared, IEG Those resources are targeted largely toward found references to the guidelines in 5 projects clearance of all projects during preparation (22 percent). In contrast, IFC draws on the EHS and tracking of high-risk projects, with some guidelines when preparing loan covenants. resources invested in staff training. However, Neither has the Bank integrated the Performance feedback from staff indicates that the incentive Standard on Labor and Working Conditions, or the structure flowing from the policy and resource one on Community Health, Safety, and Security allocation prioritizes safeguard design rather adopted by IFC in 2006 and MIGA in 2007. The than supervision and monitoring. This has had safeguards suite has functioned as a prescrip- detrimental effects on safeguards performance. tive framework for existing social policies and a restrictive framework excluding consideration of More recently, Bank policy initiatives have been other social risks that are routinely integrated by overtaken by other MDBs. Over the past decade, other members of the WBG. several development institutions—including IFC, MIGA, EBRD, and ADB—have consolidated The reluctance to add more social issues to the their environmental and social policies into suite of safeguard policies was based on the an integrated policy framework. But safeguard priority given in the policies to mitigation of policies at the Bank have evolved piecemeal. The adverse impacts. The other social issues were current set of OPs and BPs has continued to be assumed to not involve social risks (although revised individually since 1998; the latest example the experience with downsizing of state-owned is the ongoing revision of the policy on Interna- enterprises proved this assumption incorrect) tional Waterways (OP/BP 7.50). The robustness and they lacked the standards and procedures and conceptual integrity of the Bank’s safeguards for compliance found in the safeguard policies. as an integrated framework has never been The exclusion of these additional social policies subjected to scrutiny. The UCS approach was (such as OMS 2.20) from the list of safeguards an opportunity to test the integrity of the policy does not diminish their importance for the IPN, framework. However, the safeguards policies 89 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D could not be applied at the country level as an climate change. (IFC-specific gaps related to integrated policy framework. lending instruments have already been discussed in chapter 3.) IFC has put in place a process to The crucial question associated with these policy review and update the PPSSES. IEG endorses the changes will be their effectiveness in achieving practice of periodically reviewing environmental environmental and social results. The reforms and social policies to clarify areas that need more introduced by other MDBs are too recent to guidance and consider mechanisms to improve compare results but the comparison helps to implementation performance and results. highlight the need for a more systemic approach at the World Bank, which complements risk Some NGOs expressed concern about the assessment requirements with implementation weakening of safeguard standards under the and careful supervision to get better results. PPSSES, but IEG did not find any evidence of a It will be important to develop indicators and dilution of standards. From meetings and survey benchmarks to measure and monitor environ- responses, IEG found a perception among some mental and social outcomes as an integral part of NGOs that IFC and MIGA would no longer be further reform. directly accountable, under the PPSSES, if their clients do not comply fully with the Performance Policy innovation at IFC Standards. Some NGOs expressed concern that The IFC’s Policy and Performance Standards the PPSSES would result in a watering down of for Social and Environmental Sustainability accountability standards set by the safeguard (PPSSES), approved in 2006, has since been policies since they depend heavily on the private emulated by others. The PPSSES is an integrated sector client or partner for implementation and policy framework with an umbrella policy on monitoring. 11 The previous chapters of this environmental and social sustainability that evaluation compared IFC performance under applies to IFC and MIGA, and relatively well the two different sets of policy frameworks. balanced Performance Standards on environ- IEG’s evaluation did not find evidence of deteri- mental and social impacts for their clients. IFC oration of policy content among IFC’s projects added two Performance Standards—Labor and in the sample of projects reviewed. However, Working Conditions (Performance Standard civil society concerns about the relative lack 2) and Community Health, Safety, and Security of disclosure and the absence of independent (Performance Standard 4)—to the two Perfor- verification of implementation results in IFC mance Standards derived from the Bank’s social projects are valid. safeguards. IFC does not have Performance Standards on dam safety or on gender. However, Private sector ownership of social and some IFC clients integrate gender impacts within environmental sustainability their community impacts study for Performance The consistency of the Performance Standards Standard 4. On the whole, environmental and with the voluntary standards of the Equator social staffs feel the Performance Standards have Principles appears to have increased acceptance had a positive impact on clients (see appendix E). of the latter among the private sector clients of IFC and MIGA. The Equator Principles are IFC and MIGA have framed their Policy and Perfor- voluntary standards for determining, assessing, mance Standards as principles to be followed by and managing social and environmental risk their largely private sector clients. IFC and MIGA in project financing. They were developed by continue to be responsible for supervision of client private sector financial institutions and launched implementation of the Performance Standards. in 2003. Those institutions chose to model the IFC’s policy challenges lie in financial interme- Equator Principles on the safeguard policies of diary projects and frontier areas—corporate the IFC. By October 2009, 67 financial institutions finance, listed equity projects, trade finance had adopted the principles, which have become in instruments, and emerging themes such as the de facto standard for banks and investors on 90 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S how to assess major investment projects around ing modalities. ADB also has a policy on gender the world. The Equator Principles represent a and development going back to 1998. It added significant industrywide initiative. In July 2006 guidance on integrating gender and development the principles were revised after IFC’s new into ADB operations in 2006, and on integrating sustainability policy was approved, increasing social dimensions into ADB operations in 2007, but their scope and strengthening their processes to it has followed the Bank’s lead in excluding these match those of the Performance Standards. from its safeguards suite. Recent policy changes in other MDBs By using the terms “Performance Require- The EBRD, whose architecture resembles that ments,” and “Safeguard Requirements,” at least of IFC and MIGA, adopted the PPSSES in 2008 in their choice of language, both agencies have with some improvements. While retaining the signaled a commitment to greater accountabil- structure of social and environmental sustainabil- ity. However, it will be a few years before the ity, EBRD used the term “Performance Require- environmental and social results of these policy ments” instead of Performance Standards. revisions at EBRD and ADB are known. Significant responsibility for ensuring implemen- tation lay more clearly with EBRD. It has also Safeguards architecture added a performance requirement on financial Most MDBs have umbrella policies on environ- intermediaries and one on Information Disclo- mental and social sustainability. In terms of sure and Stakeholder Engagement. The growing the internal architecture (see table 5.1), each share of financial sector projects at IFC and MIGA MDB dealing largely with the private sector— suggests that similar, more explicit guidance EBRD, IFC, and MIGA—has an explicit policy for financial intermediaries would be useful for on environmental and social sustainability with IFC too. IFC is also in the process of updating a set of Performance Standards (IFC/MIGA) or its disclosure policy. Disclosing supervision and Performance Requirements (EBRD) under the monitoring results to key stakeholders would umbrella. Among the regional development augment IFC’s ability to improve results. banks, the Inter-American Development Bank (IDB) does not have an umbrella but maintains In the treatment of indigenous peoples, EBRD a balance among environmental and social has adopted the language of “free, prior, and aspects with three separate policies on environ- informed consent” rather than “free, prior, and ment, involuntary resettlement, and indigenous informed consultation” provided for in WBG peoples. IDB also situates its safeguards work in policies, thereby meeting a longstanding demand the context of environmental and social sustain- of advocacy groups for indigenous peoples. ability. ADB has an umbrella Safeguard Policy IFC is currently reviewing the corresponding Statement with a similar balance on environmen- language in the PPSSES as well. tal and social aspects under the policy. The Bank and the African Development Bank have neither In July 2009, the Board of Directors of the ADB an umbrella safeguard policy nor a policy on approved a new Safeguard Policy Statement, 12 environmental and social sustainability. Given the which brings its previous safeguard policies on the renewed concern for environmental and social environment, involuntary resettlement, and indige- sustainability and climate change, a review of the nous peoples into one single policy that enhances Bank’s safeguards as an integrated framework consistency and coherence, and more comprehen- that complements up-front risk assessment with sively addresses environmental and social impacts implementation and sustainability is needed. and risks. ADB refers to the safeguard areas under the policy as “Safeguard Requirements.” In addition An unintended consequence of the lack of a to the three previous safeguard areas, ADB has policy on environmental sustainability at the added a fourth dealing with special environmen- Bank is that it has resulted in a divide between tal and social requirements for different financ- safeguards work and the considerable body of 91 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D work being done by the Bank’s environmental policies on gender, integrated water resources units to strengthen national environmental laws management, and agriculture and rural develop- and systems. Safeguards work is aimed at mitigat- ment, but it does not have a policy on indigenous ing the adverse impacts of investment projects peoples. while environmental policies and institutions are being supported by analytical work, techni- The absence of a comprehensive policy at the cal assistance, and lending operations, includ- World Bank limits the scope of social policies to ing investment projects and environmental a far greater extent than environmental policies. DPLs. However, that work is kept operationally The Bank safeguards consist of six environmen- distinct from the work on safeguard compli- tal, two social, and two legal policies (issues ance, leading to a growing disconnect between addressed in those legal safeguards are not the two and running the risk of undermining covered to the same length or complexity in client ownership for environmental and social other MDBs). OP 4.01 (Environmental Assess- safeguards. This separation has been reinforced ment) provides well-rounded coverage of by the transfer of staff providing quality environmental issues. However, as discussed assurance and oversight for safeguards from previously, the social safeguards are limited to the former ESSD network to OPCS and their the two (Involuntary Resettlement and Indige- counterparts in the regions, on the grounds that nous Peoples) that are labeled safeguards. The housing the safeguards compliance staff within set of safeguard policies is skewed toward the the same Sustainable Development Network environment both through the exclusion of whose infrastructure and agricul- social policies existing within the Bank, and The lack of a policy tural projects generate safeguard the Performance Standards on labor issues and on environmental impacts could lead to a conflict of community impacts adopted by IFC, MIGA, sustainability at the Bank interest. The effects of Sustainable and EBRD. The lack of a comprehensive policy has resulted in a divide Development Network integration for coverage of social risks thus appears to be between safeguards and on environmental and social perfor- resulting in the inability to capture social risks the work to strengthen mance have not been assessed in adequately. However, the addition of other national environmental this evaluation because it will likely themes would have to be accompanied by laws and systems. be examined under IEG’s planned policy consolidation under one social umbrella matrix evaluation. policy to ensure synergies and efficiency gains. Policy balance Two considerations militate against consolida- Overall, the environmental and social policies in tion of all the safeguards under one umbrella. IFC, MIGA, ADB, EBRD, and IDB provide more Internally, the current safeguards framework comprehensive coverage of environmental has resulted in a narrow approach to social and social risks than those of the Bank. Among safeguards, restricting them to two specialized the MDBs, the World Bank has the most visible issues. Expansion of thematic coverage will be imbalance among its environmental and social more practicable if the assessment of social risks policies and also lacks an umbrella policy on is consolidated under one assessment that priori- environmental and social sustainability. However, tizes these risks according to their relevance to unlike the other three, ADB has confined each operation. Externally, the country dialogue itself to three safeguard areas—environment, on safeguards has hitherto been dominated by resettlement, and indigenous peoples—while environmental concerns and complemented it maintains a separate focus on gender in its by a large volume of environmental analytical operational manual. Unlike IFC, MIGA, EBRD, work, institutional strengthening, and lending and ADB, the African Development Bank does operations. Expanding the scope of country not comingle its environmental and social legislation and environmental institutions to policies, and has separate policies on environ- ensure comprehensive coverage of social risks is ment and involuntary resettlement as well as not realistic since most countries have separate 92 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S agencies to deal with environmental and social risks. Rewriting environmental legislation, and Box 5.1: Country Director Views on Social Safeguard revamping institutional responsibilities and Policies regulatory structures for environmental agencies to deal with social risks would require significant Country directors find Bank social safeguard policies narrow, rigid, and institutional change and impose substantial costs difficult to implement. on Bank clients without the assurance that those institutions would have the capacity to address • “There is a serious disconnect between what countries are doing and social impacts adequately. our social safeguards. The resettlement policy is way out of line with what our clients have. It is amazing how little clarity there is on how However, there is merit in consolidating the this policy applies to legal or nonlegal, poor or nonpoor claimants, and environmental safeguards within one umbrella how to deal with people who refuse to accept the compensation rate.” policy. The environmental assessment policy • “By mandating some things that are onerous we encourage task is always triggered when other environmental teams to take a very narrow look at those social issues… They focus risks occur, and the EA/EIA encompasses the on resettlement but not on other social issues.” risks covered in the remaining environmental • “We are creating distortions in countries, for example, on restoration safeguards. In terms of managing safeguard risks, of livelihoods. Our recommendations are not consistent with reality treating them in isolation can lead to duplica- and distort the market. The compensation is far greater than what tion or redundancy. While the separate themes the market is offering. This is more apparent in Africa but is also true remain relevant, they could be amalgamated as in other regions.” requirements under a consolidated environmen- • “The policies are open to a lot of interpretation between the com- tal policy. pliance unit in OPCS, the regional safeguards unit, and the sector units. It depends a lot on personalities, and the transaction costs Fitting Social Safeguard Policies to are huge for us.” Context: Lessons from MDBs The two most contentious issues regard- Source: IEG interviews with country directors. ing social safeguards have significant policy implications: the treatment of squatters in the Involuntary Resettlement Policy, and the lack social sector projects with no immediate physical of differentiation in the treatment of people impacts. Classifying as category B projects that affected by resettlement or indigenous peoples require little or no mitigation measures widens in projects designed to benefit them compared the category band and undermines the purpose with projects where they are not the intended of categorization. IEG found that clients incurred beneficiaries. expenditures of less than $1 million in a third of category-B projects in its portfolio sample. IEG interviews with country directors (box 5.1) Except for one project, which was rated unsatis- and clients revealed that the social safeguard factory, these expenditures seemed adequate, policies were difficult to implement. Country indicating that these projects need not have directors were most concerned about the been placed in category B. prescriptive and restrictive nature of the policies, which they felt often differed from national laws, In the focus group discussion with Bank staff an observation that resonates with the findings (appendix E), the absence of an umbrella policy for of the UCS review. Sector managers’ views were an integrated social assessment was highlighted more variable (box 5.2), some sharing similar as a key constraint faced by social development concerns about social safeguards, others critiqu- staff, which compels them to ignore risks of wider ing environmental safeguards. On environment, import. They felt that this limited their ability the main concern was “mission creep,” which to adapt their work to different contexts and was reflected in overcategorization of projects forced them to rely on the two social safeguard such as some technical assistance loans13 and policies as the only entry point for engaging 93 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Indigenous peoples policy Box 5.2: Sector Manager Comments on Bank The Indigenous Peoples policy was developed Safeguards originally to address social exclusion of disadvan- taged ethnic groups in Latin American countries. • “Safeguards do add value when done well. When done poorly they Subsequently, the policy was adopted and made increase transaction costs and lead to delays and problems.” applicable globally, although the term “indige- • “High level of risk aversion is resulting in extension of the [envi- nous” does not translate easily elsewhere and ronmental] safeguards approach to areas where their immediate its relevance is more contested in other regions, value is questionable, for example, to technical assistance loans except in countries where national legislation and land cadastre projects.” recognizes the rights of indigenous peoples • “The disconnect between borrower-Bank policies is greatest on or ethnic minorities. The policy is particularly resettlement.” challenging in conflict countries and multi- • “Safeguards work is much stronger on environment than social ethnic countries in Asia and Africa grappling with because of the existence of an umbrella policy. Social safeguards challenges of national integration. are very different. The Bank does not have an umbrella policy on social analysis, unlike the IFC, which has a balanced policy on The richest experience in applying this policy social and environment.” remains in Latin America by both the Bank and • “The Indigenous Peoples policy is easier to implement in Latin the IDB. The experience and approach adopted America [than in other regions] as there is constitutional protec- by IDB is therefore directly relevant to the WBG. tion in most countries, which have moved from assimilation to Policies of the IDB (and EBRD) include provisions acceptance of diversity.” for adjusting the policy requirements to specific • “The Indigenous Peoples policy has resulted in more risk aver- contexts. sion [other than in Latin America and Caribbean region] than the resettlement or environmental policies. Projects have been modified The IDB classifies projects under its Indige- to avoid dealing with IPs.” nous Peoples Policy (IPP) into three categories: • “Safeguard policies are not well aligned to community-driven de- (i) independent projects for indigenous peoples, velopment and programmatic approaches. They have been applied (ii) mainstreaming projects, and (iii) projects to such projects with policy frameworks. But we often fail to ensure with safeguards. While all three types involve that site-specific supervision and design is done when frameworks consultations with indigenous peoples to ensure are used.” that their inputs and perspectives are integrated • “Environment policies are more prescriptive and are even less within project design, the intensity of studies aligned with programmatic approaches than the social safeguard varies. Independent projects are those that are policies, which include provisions for use of frameworks.” designed wholly to benefit indigenous peoples. Therefore, by design they are expected to “do Source: IEG interviews with sector managers. no harm.” Mainstreaming also does not involve harmful impacts; it seeks to increase the value added by the bank’s projects to ensure that indigenous people are also able to benefit from them. Projects with safeguards are those that clients. One of the main criticisms identified in have potential direct or indirect adverse impacts IEG’s client surveys and manager interviews was on the rights and assets of indigenous peoples. that the social safeguards are implemented in a The third type of IPP project is subject to much rigid, legalistic manner with insufficient adapta- greater scrutiny than the other two. tion to the operational context. IEG investigated how other IFIs were implementing their social The EBRD’s policy14 toward indigenous peoples safeguard policies in different contexts. While distinguishes between projects that are likely to not an exhaustive treatment of the subject, the result in adverse impacts on them and those that examples below from other IFIs seem relevant to are not, with more stringent requirements for WBG operations. projects causing adverse impacts. EBRD’s policy 94 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S makes a distinction between the types of impacts on indigenous peoples to determine the level of Box 5.3: Civil Society Comments on Bank risk assessment needed. As is the case for IFC, Safeguards EBRD’s policy also permits the assessment of indigenous peoples’ issues to be undertaken as • “The Bank safeguards were effective when adopted but have not part of a broader community impact assessment, been updated, and the world has changed in the meantime.” which may be more acceptable in countries • “Whereas certain Bank policies may once have been considered where other forms of vulnerability or social ’best practice,’ there are now institutions and accepted standards exclusion are also pervasive. that offer superior social and environmental protections: Several private banks, including Citigroup, Bank of America, Involuntary resettlement policy and HSBC, have stronger biodiversity conservation and forest protection policies than the WBG. In one African country, the government Corporations such Alcoa, DuPont, and Shell have eclipsed the decided to fund a housing project involving Bank in setting greenhouse gas emissions reduction targets. slum clearance and redevelopment from its On climate change the ADB is more stringent than the Bank.” own funds rather than Bank financing. In • “IBRD/IDA needs to broaden safeguard coverage of social impacts, the government’s view, the Bank’s safeguard at least to match IFC/MIGA.” policies would be unworkable in the high-density • “Environmental/social assessment framework for development urban environment where it would be unable policy lending must be strengthened.” to control the huge influx of rent-seeking illegal • “Performance Standards do not go far enough to protect indigenous occupants demanding compensation. peoples.” —Country director interview • “Special provisions are needed in respect of safeguards in conflict areas or where there is poor governance (the Chad-Cameroon As previously identified from interviews with Pipeline project is a good example of this).” clients and managers, involuntary resettlement is • “The issue of human rights should not be avoided.” perceived as contentious because of the rigidity in the WBG’s policy requirements on land acquisi- Source: IEG evaluation survey findings. tion regarding treatment of untitled persons. Unless untitled persons have customary tenure rights they are considered illegal occupants in all countries, yet WBG policies entitle them improve the quality of lives of affected persons, to compensatory assistance for involuntary and those projects where displacement is an resettlement and economic rehabilitation. The adverse consequence of a broader operation not policy does not distinguish among displacement designed to benefit them.15 In particular, when induced by projects purposively designed to a project is designed to move people from areas benefit affected people from those that merely that are unfit for human habitation, or to provide induce adverse impacts on them, and the policy basic infrastructure in urban upgrading projects, requirements apply regardless of the significance or resolve land tenure problems, resettlement is of impact. In one instance “an extension of a treated in a different manner from those projects small pumping station by a few meters,” which where the project is not designed to benefit the did not even cause physical displacement of any people it is affecting. The distinction between household, led to a whole overlay of safeguard operations designed to benefit locally affected compliance issues. These rigidities are being persons and those that only affect them adversely, addressed by other MDBs by considering the provides much more flexibility in adapting policy nature and significance of resettlement impact. to the operational context. In their policy of Involuntary Resettlement, the Impoverishment Risk Analysis was pioneered IDB similarly makes a distinction between projects at the World Bank, however, the Bank does not where resettlement is a project objective to use this instrument routinely in its resettlement 95 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D work to assess potential social impacts and paradigm for the private sector. The two share evaluate project outcomes. IDB has adopted it similar objectives “to avoid, mitigate, or minimize and requires an Impoverishment Risk Analysis16 adverse environmental and social impacts of for projects inducing involuntary resettlement. projects supported by the Bank” and to ensure Impoverishment Risk Analysis serves to remind that they are “environmentally sound and sustain- task teams that the primary risk of resettlement able.” IFC applies the Performance Standards is impoverishment. Systematic risk analysis “to manage social and environmental risks and enables task teams to identify which of these impacts and to enhance development opportu- risks are more acute and need to be addressed. nities in its private sector financing.” By linking the lack of clear title to housing, liveli- hoods, and impoverishment, IDB also offers The safeguards paradigm prescribes assessments a way to deal with the claims of speculators or and normative requirements for clients with politically influential land grabbers who seek mitigation measures designed prior to project to capture development benefits. Adapting approval, just as traditional projects were fully safeguards measures to the poverty context designed at appraisal. Although many countries provides a useful guide to mitigation actions and have adopted similar environmental legislation, also provides an opportunity to link safeguard whenever national regulations differ from the outcomes to the Bank’s mission of poverty Bank’s safeguard policies those are overridden reduction. by the requirements of the Bank’s safeguard policies. Supervision focuses essentially on These illustrations are provided to encourage compliance with the mitigation plan rather than introspection and debate within the Bank on on monitoring outcomes. The exceptions made how to improve implementation and ownership in recent years for financial intermediary projects of these policies. The examples provided do not and other decentralized projects that replaced imply that the policies should be restricted to risk assessments with policy frameworks have had cases where adverse impacts are induced, nor do even worse results than other projects because of they imply an elimination of the need to consult weak follow-up during implementation. with affected people or indigenous peoples. But they do suggest the need for further refine- The Performance Standards paradigm is based ment of the policies to differentiate projects that on an expectation that the clients, who are need mitigation plans to address adverse social private business entities, will voluntarily adhere impacts from those projects where the consulta- to Performance Standard requirements, with tion process is aimed at enhancing the project’s loan covenants that provide remedies if they benefits. Adopting such measures to customize do not. IFC places greater responsibility for safeguards policies to the operational context implementation and monitoring of specified would require a revision to existing performance indicators on the client and Keeping the focus on policies to permit this sort of flexibil- supports this with supervision and documenta- poverty outcomes and ity during implementation. Keeping tion of performance. However, lack of indepen- benefits accruing to the focus on poverty outcomes and dent verification and disclosure remain a affected people can help benefits accruing to affected people fundamental lacuna. ensure that the ends can help ensure that the ends justify justify the means. the means. The introduction of the Performance Standards is too recent to compare results with those WBG Paradigms for Environmental and under the safeguard policies. However, there Social Management is evidence that the instruments and practices The WBG is using two different paradigms for introduced by IFC in parallel with the Perfor- its environmental and social work, the Bank’s mance Standards have improved implemen- safeguards paradigm, largely for the public tation. IFC’s systems include a balanced sector, and the IFC’s Performance Standards environmental and social thematic coverage, 96 S A F E G U A R D S A N D S U S TA I N A B I L I T Y F R A M E W O R K S : P O L I C Y I S S U E S including more relevant social standards. It relies ment, and appraisal of the proposed mitigation on the responsibility of the private sector client plans to ensure compliance with safeguard or partner for annual monitoring and achieve- policy requirements. They have highlighted and ment of Performance Standards but lacks disclo- covered high-risk projects adequately but have sure of monitoring information, third-party lacked the flexibility to address a broader array of verification or community monitoring. It has risks. Supervision and monitoring receive far less more intensive supervision and review of perfor- attention affecting implementation quality and mance, including use of explicit performance leading to highly uneven results. indicators, an annual environmental and social performance review by IFC, as well as integra- Ideally, policy reform should draw on the tion of environmental and social effects as one strengths of both the safeguards and Perfor- of the four dimensions of the project’s develop- mance Standards in an integrated fashion. A ment objective. paradigm based on more relevant thematic coverage, flexibility in procedures ensuring The World Bank’s systems are frontloaded, stress- the integrity of standards, and client responsi- ing requirements during appraisal, and individual bility for monitoring would likely lead to more policies have more specific requirements (see client ownership, and verification and disclo- table 5.2). They emphasize screening, risk assess- sure would ensure better results. The quality Table 5.2: Paradigms for Achieving Social and Environmental Results Safeguard policies Performance standards Client refers to borrower, agency, and local Client refers to private sector firms or business partners, but not communities beneficiaries Environmental Assessment Policy provides Performance Standard 1: Social and Environmental Assessment and Management Systems an entry point for the country dialogue on provides an entry point for dialogue with the client environmental (but not social) safeguards Thematic coverage of social safeguards Expanded thematic coverage includes labor and working conditions, and community impacts restricted to involuntary resettlement and (including health, safety, and security) indigenous peoples Mandatory Bank requirements for project Preparation and disclosure requirements rest with private sector clients or partners; loan preparation and disclosure; loan covenants covenants included for compliance with Performance Standards during implementation included largely for higher-risk projects No monitoring requirement from clients and no Client submits its Annual Monitoring Report and IFC staff prepare annual Environmental and reporting requirement in supervision reports, Social Review Document with indicators for each Performance Standard except for ISR safeguard ratings Identification and supervision of category A Identification of category A inadequate, but supervision process strengthened for all adequate but not of category B Supervision reports not disclosed until AMRs and ESRD not disclosed and environmental and social outcomes not independently fiscal 2011 and outcomes not measured or verified, except for high-risk projects, weakening accountability independently verified, except for high-risk projects Rare use of EHS guidelines EHS guidelines routinely used ICR reporting weak; no link to results Environmental and Social Effects rated in the Extended Project Supervision Report as one of framework and the project’s development four dimensions of the project’s development outcome objective Source: IEG. 97 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Without investing in of implementation and monitoring, Implications for the World Bank Group these elements the IFC's which depends on client capacity and The Bank’s 2001 Environment Strategy Performance Standards commitment, needs to be adequate, highlighted the need to address a number of paradigm would be a and checks and balances need to safeguard issues that are still relevant today. riskier option than the be in place to ensure that intended A two-pronged approach was outlined to first previous safeguards social and environmental outcomes address a number of short-term priorities, compliance model with are achieved. In this context, four such as strengthening safeguard compliance, respect to environmental elements are vital: (i) investment in “mainstreaming” environmental concerns across and social outcomes. the clients’ social and environmen- Bank sectors, and improving results by strength- tal management system; (ii) integra- ening the Bank’s internal review, monitoring, tion of adequate environmental and social and tracking system. The strategy also sought to performance indicators in the project’s results move safeguard considerations “upstream” in the framework; (iii) effective instruments for decision-making process through use of strate- monitoring by the client, reinforced by indepen- gic and country environmental assessments. It dent verification of environmental and social also encouraged development of a risk-based outcomes; and (iv) regular supervision, perfor- management system to address the intrinsic mance review, and disclosure of monitoring and risk of a given project and the client country’s supervision reports. Without investing in these capacity to manage that risk, and move toward elements the IFC’s Performance Standards greater in-country ownership of and capacity to paradigm would be a riskier option than the implement safeguards in accordance with their previous safeguards compliance model with own policies and procedures. respect to environmental and social outcomes. The analysis does not provide sufficient basis The limitations of the safeguards paradigm to prescribe specific policy changes, but it does become even more visible in the UCS pilots. make a compelling case for updating the Bank’s The mandatory requirements spelled out in an safeguards framework to achieve better environ- Operational Policy have prevented meaningful mental and social outcomes. That would involve dialogue and the experimentation needed for the refining Bank environmental and social policies to pilots to succeed. And the piecemeal approach strengthen implementation, borrowing elements to examining individual safeguards policies at the of the changes in M&E and the clients’ role from project level has reduced the likelihood that any IFC, as well as policy innovations of other MDBs, Bank borrower will be able to adopt the entire and redesigning the Bank’s country systems suite of safeguards policies. Adoption of country approach. At the same time, implementation of systems will only be feasible if the Bank starts high-risk projects and the enhanced disclosure from national laws and regulations, rather than policy adopted by the Bank provide lessons for from individual projects and individual safeguard other MDBs. Harmonizing thematic coverage policies, and invests in strengthening client of the policies, shared systems for risk-based institutions and capacities, analogous to IFC’s categorization of projects, and performance focus on the social and environmental manage- indicators to track outcomes across the WBG are ment systems of its clients. This would require highly desirable. This will need leadership from a shift in emphasis from mandatory procedural senior management to indicate the priorities requirements at entry to monitoring implemen- for stronger environmental and social results, tation and outcomes, adopting a results-oriented and mobilization of a technical team dedicated paradigm for its operational work, even if to seeking out opportunities for innovation and the overall objectives of the Bank safeguards reform to strengthen social and environmental framework are retained. sustainability. 98 Chapter 6 Community discussion in India. Photo by Curt Carnemark, courtesy of the World Bank Photo Library. Conclusions and Recommendations Conclusions I n the 1980s, the WBG began developing a series of Operational Policies to safeguard people and the environment against undue harm from the development operations it finances. In 1997, ten existing policies were labeled safeguard policies and have since been treated as the environmental and social standards to which the WBG has committed and is publicly ac- countable. These policies have been revised over the past decade but always in a piecemeal fashion, without examination of the safeguards policies as an integrated framework. While the World Bank still relies on these safeguard policies, IFC and MIGA, two other members of the WBG, replaced them with the Policy and Performance Standards on Social and Environmental Sustainability in 2006 and 2007, respectively. Cumbersome procedures to revise Bank policies have created a disincentive against modernizing the safeguard policies, leading in several instances to rigidities and difficulties in implementation. IEG found strong support among WBG staff, IEG has conducted several evaluations of management, and clients for the underlying individual safeguard policies, such as Involuntary principles of the safeguards and Performance Resettlement, Indigenous Peoples, and of related Standards. Three-fourths of the respondents themes, including the Environment Strategy, to an IEG survey of NGOs said that the WBG’s Social Development in World Bank Operations, performance in dealing with environmental and Environmental Sustainability, and Climate social impacts is significantly better than in the Change. This evaluation is the first assessment 1990s, while less than 10 percent thought WBG of the entire suite of environmental and social performance had worsened. However, substan- safeguards and sustainability policies. Its purpose tial implementation and capacity challenges is to go beyond evaluation of compliance with remain. requirements of individual safeguard policies 101 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D or standards but of the relevance, efficacy, and social policies does not necessarily imply an effectiveness of the safeguards and Performance additional burden on the Bank and its clients. Standards as integrated policy frameworks. Consolidation under one umbrella social policy that provides for an integrated social assessment The evaluation also attempts to compare environ- and management of relevant, project-induced mental and social performance at IFC and MIGA social risks and vulnerabilities would increase under the safeguards policies with that under the efficiency and enhance synergies among them. A Performance Standards. IEG found IFC’s instru- social policy umbrella would allow a determina- ments and procedures for monitoring performance tion of the most significant social risks relevant under the PPSSES to be an improvement over to each project, just as OP 4.01 allows the identi- those under the safeguard policies, although still fication of the most significant environmental lacking mechanisms for disclosure and indepen- risks relevant to each project. Further consoli- dent verification. And given the short period since dation of the environmental and social themes the adoption of the PPSSES, IEG is unable to arrive under one umbrella is not advisable because the at conclusions regarding performance of projects environmental institutions, which the Bank has prepared under the new policy framework. helped put in place in borrower countries, have neither the mandate nor the expertise to address Nonetheless, IEG’s assessment reveals that the most of these social risks. Expecting environ- PPSSES cover a wider spectrum of environmen- mental agencies to expand their mandate to tal and social effects,1 including labor issues, cover social issues runs the risk of continued such as retrenchment and workers’ rights, and marginalization of social issues, as has been the community impacts relevant to WBG operations experience within the Bank to date. than that covered by the Bank’s safeguard policies. In addition, IFC uses EHS guidelines as These policy changes would improve the relevance binding covenants for clients, covering a number and efficiency of the Bank’s environmental and of high-risk industries with specific guidance and social policies. The evaluation has highlighted requirements to prevent pollution. IFC applies the urgent need to strengthen implementation, specific guidelines to address occupational health supervision, monitoring, and reporting to improve and safety aspects, an important social area that the effectiveness of the World Bank’s social and has been relevant in 85 percent of IFC projects environmental policy framework. but the use of EHS guidelines by the Bank is rare, even in projects with obvious health and safety The changing nature of the portfolio requires risks, as in power plants and road construction. special attention. IEG found a growing number of projects whose impacts cannot be appraised prior IEG found the Bank’s safeguard policies to be to project approval. In the Bank, these include imbalanced, with several notable gaps on social projects managed by financial intermediaries, risks, which are addressed more systematically community-driven development projects, and by other MDBs. The Environmental Assessment projects with multiple subprojects not identified policy (OP 4.01) functions as an effective umbrella at appraisal. For these projects, task teams rely on for the environmental safeguards, which could policy frameworks,2 which describe the actions be consolidated under one combined environ- to be taken by the client to assess impacts, and mental policy. However, the current safeguards develop and implement mitigation plans, during framework has led to the exclusion from scrutiny project implementation. Uneven supervision has of some project-induced risks already included affected the WBG’s ability to ensure quality in in IFC/MIGA Performance Standards, other these projects. existing Bank policies, and EHS guidelines. Their inclusion will make it more likely that At the Bank, staff capacity is unevenly distrib- priority risks more relevant to client and project uted across the regions, with Africa suffering needs will be addressed. The addition of these the most from skill shortages, particularly on 102 C O N C L U S I O N S A N D R E C O M M E N D AT I O N S social safeguards. The skills deficit is further exposed to environmental and social risks, and compounded by the increasing separation leaves MIGA unable to provide assurance that the between environmental and social staff who projects it supports have adequately prevented work on safeguards and those who work on and mitigated their adverse environmental and other environmental and social issues, which are social impacts. perceived as being professionally more satisfy- ing. The relocation of safeguards clearance and Recommendations compliance functions with the transfer of the The following recommendations to the Quality Assurance and Compliance Unit from World Bank, IFC, and MIGA are made to help ESSD to OPCS, and the Regional Safeguard maintain the objectives of safeguards and Advisors from regional ESSD departments to sustainability policies; strengthen compli- regional departments with OPCS functions, ance, implementation, and accountability; appears to have resulted in more systematization and improve clients’ and the WBG’s ability to of clearances and appraisal but has also inadver- promote positive social and environmental tently increased the separation between environ- results. These objectives are critical to realiz- mental and social staff working on safeguards ing the WBG’s goal of poverty reduction in the and those engaged in own-managed operations, context of increasingly mature clients and a whose work programs are not always well aligned rapidly evolving portfolio. with each other. Responsibilities and incentives need to be realigned to address these constraints. The WBG’s safeguards and Performance Standards play a key role in ensuring adequate Overall, IFC is demonstrating improved compli- attention to environmental and social outcomes. ance with its policy framework since introduc- Given the changing nature of its clients and tion of the PPSSES. IFC is using its leverage in portfolios, the challenge is to ensure the contin- applying the Performance Standards effectively ued relevance of the WBG’s environmental and within project and corporate finance with identi- social policies, complementing the emphasis fied use of proceeds, but this leverage is limited on compliance with effective implementation in listed equity projects and trade finance. The to increase the efficiency and effectiveness of introduction of IFC’s on-line ESRD document, its practices. The transaction costs of safeguards with specific performance indicators and a rating and rigidities in policy interpretation, particularly system, has strengthened monitoring under for the social safeguards, have led to risk aversion the PPSSES framework. However, its efficacy and, in some cases, to avoidance of projects or depends largely on the quality of the AMRs components that would benefit the poor. submitted by clients and IFC’s environmental and social staff resources to maintain the ESRD To ensure that vulnerable groups are not excluded documents. Supply chains are not well identi- from development benefits, there is a need to fied and managed, especially in agribusiness and adopt differentiated approaches for interven- trade finance in IFC-financed operations. IFC’s tions designed to benefit affected persons from local disclosure practices and monitoring are those likely to induce adverse effects. The evalua- weak and communities do not have sufficient tion points to the need for a systems approach information on projects impacts. that balances up-front risk assessment with implementation support to increase effective- Overall, MIGA is managing the limited resources ness; policy consolidation with more compre- devoted to its sustainability function efficiently, hensive, balanced thematic coverage to ensure with positive results in comparison to the adequate up-front regulations, while providing absence of MIGA’s intervention. But the very for better supervision, monitoring and evalua- limited extent to which MIGA can monitor and tion, verification, and disclosure; and partnership ensure the projects’ compliance with the applica- with clients, third parties, and local communities ble Performance Standards leaves the projects to enhance ownership and results. 103 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 1. Revise the policy frameworks to harmo- • Increase the capacity of the Environmental nize thematic coverage and guidance and Social Unit to the level needed to provide across the WBG and enhance the relevance credible assurance on performance against the of those frameworks to client needs. standards for every project. Should MIGA be unable to increase its resources devoted to • IFC, MIGA, and the World Bank should jointly implementation of Performance Standards, adopt and use a shared set of objective crite- it should revise its Policy on Social and Envi- ria to assess social and environmental risks ronmental Sustainability to disclaim any re- to ensure adequacy and consistency in proj- sponsibility for monitoring the projects’ social ect categorization across the WBG, using the and environmental performance and ensuring more inclusive criteria for category A, and re- that they comply with the standards. Under fining the categorization system to address the this option, MIGA’s role would be limited to bunching of higher- and lower-risk projects reviewing the client’s assessment of the proj- within the current category B. ect’s environmental and social risks against the standards, identifying corrective actions as The World Bank should: needed, and securing the client’s commitment to implement these actions. • Ensure adequate coverage of social effects— • Require that category-B Small Investment Pro- integrating community and gender impacts, gram projects follow the same disclosure re- labor and working conditions, and health, quirements as for regular category-B projects. safety, and security issues not currently cov- ered by its safeguard policies—by consoli- 2. Enhance client capacity, responsibility, dating existing social safeguards with other and ownership. WBG policies on social risks as requirements under one umbrella policy on social sustain- The World Bank should: ability. • Consolidate the environmental policies as • Increase the synergies between safeguards requirements under one umbrella policy on work and broader Bank engagement on en- environmental sustainability. vironmental and social sustainability by in- • Revise the current approach to safeguards vesting in upstream analytical work, technical pilots on use of country systems to focus on assistance, and lending to strengthen country strengthening country institutions and sys- and sector institutions and capacities in client tems to manage environmental and social countries. risks. • Require regular reporting by the borrower on implementation and outcomes of safeguards IFC should: in Bank-supported projects, and work with clients to develop instruments and indicators • Strengthen the provisions on sustainability to help in such monitoring. to address emerging issues, notably climate change and supply chains and their commod- IFC should: ity certification. • Develop more robust approaches to the im- • Develop incentives for investment officers plementation of the Performance Standards in to share ownership of the Performance Stan- financial intermediary projects, listed equities, dards and mainstream their implementation. and trade finance. • Use advisory services to build social and • Strengthen policies and practices on disclo- environmental management systems and sure, including at the local levels. implementation capacity, especially among small and medium enterprises, financial in- MIGA should: termediaries, and clients in countries and 104 C O N C L U S I O N S A N D R E C O M M E N D AT I O N S sectors with weak environmental and social ing for higher-risk projects, and disclosure of management. monitoring and supervision reports. • Mobilize resources at appraisal for energy and clean production audits, using auditors with IFC should: relevant sector knowledge. • Define areas of influence and requirements • Enhance the supervision of financial interme- to better address supply chain risks and op- diaries at the subproject level by developing portunities, particularly related to biodiversity clear guidelines for applying the Performance and forestry, expanding the application of ma- Standards at the subproject level and by adopt- terial biodiversity along the supply chain for ing a systematic approach to environmental suppliers. and social specialists’ site visits to selected subprojects. MIGA should: • Use loan covenants, including Conditions of Disbursement to enforce compliance with • Focus the due diligence reviews of financial environmental and social requirements and sector projects on the Social and Environ- reporting if the clients lack commitment and mental Management Systems of developing are continuously out of compliance. country subsidiaries the project supports, rather than the corporate policies of the par- 4. Strengthen safeguards monitoring, evalu- ent banks. ation, and completion reporting. • Expand the size and eligibility of the Trust Fund for Addressing Environmental and Social The World Bank should: Challenges to all low-capacity clients on the basis of need. • Include performance indicators on environmen- tal and social outcomes in project results frame- 3. Revise guidelines, instruments, and in- works and ensure systematic collection of data to centives to strengthen supervision ar- monitor and evaluate safeguards performance. rangements. • Ensure that Implementation Completion Re- ports and IEG reviews of those reports rate The World Bank should: and report effectively on the outcomes of safe- guards and, for all projects with significant • Assign responsibility and budget for safeguards environmental and social effects, ensure the oversight and reporting to environmental and results are incorporated as an essential dimen- social units in each operational region—in line sion when assessing achievement of the proj- with IFC practice—in place of the delegation ect’s development objective, as has already of safeguards processing and supervision to been done for IFC and MIGA. sector management units. • Introduce a certification program to expand IFC should: the pool of staff qualified to undertake social and environmental preparation and supervi- • Disclose project-level environmental and so- sion while ensuring quality and consistency, cial information from monitoring and supervi- and provide orientation training on environ- sion reports. mental and social sustainability to all task team • Make use of independent, third-party, or com- leaders. munity monitoring and evaluation for its proj- • Develop and implement an action plan to en- ects, particularly for projects with involuntary sure regular supervision of financial interme- resettlement and higher-risk financial interme- diary projects and investment projects that use diary and agribusiness projects. social and environmental policy frameworks through third-party or community monitor- MIGA should: 105 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D • Disclose project-level environmental and so- • Seek greater symmetry in the structure of WBG cial information from supervision reports. accountability and grievance redress mecha- • Develop a credible mechanism to ensure that nisms. For the World Bank this would entail Performance Standards are adhered to by fi- creation of a grievance redress and conflict nancial sector projects. resolution mechanism to complement the IPN. For IFC and MIGA this would entail a more 5. Improve systems and instruments for independent compliance review process, en- accountability and grievance redress. suring that the CAO submits its audits directly to the Board. IFC, MIGA, and the World Bank should: 106 Appendixes 107 APPENDIX A: SAFEGUARD AND SUSTAINABILITY POLICIES Table A.1: World Bank Safeguard Policies OP/BP 4.01 The Bank requires environmental assessment (EA) of projects proposed for Bank financing to help ensure Environmental Assessment that they are environmentally sound and sustainable, and thus improve decision making OP/BP 4.04 To promote environmentally sustainable development by supporting the protection, conservation, Natural Habitats maintenance, and rehabilitation of natural habitats and their functions. OP 4.09 To minimize and manage the environmental and health risks associated with: Pest Management • Pesticide use; and • Promote and support safe, effective, and environmentally sound pest management. OP/BP 4.10 To design and implement projects in a way that fosters full respect for indigenous peoples’ dignity, human Indigenous Peoples rights, and cultural uniqueness, so that they: • Receive culturally compatible social and economic benefits; and • Do not suffer adverse effects during the development process. OP/BP 4.11 To assist in preserving physical cultural resources and avoiding their destruction or damage. Physical Physical Cultural Resources cultural resources include resources of archaeological, paleontological, historical, architectural, religious (including burial sites), aesthetic, or other cultural significance. OP/BP 4.12 To avoid or minimize involuntary resettlement and, where that is not feasible, to assist displaced persons Involuntary Resettlement in improving or at least restoring their livelihoods and standards of living, in real terms, relative to predisplacement levels or to levels prevailing prior to the start of project implementation, whichever is higher. This policy covers direct economic and social impacts that both result from Bank-assisted investment projects and are caused by the involuntary taking of land resulting in: • Relocation or loss of shelter; • Loss of assets or access to assets; or • Loss of income sources or means of livelihood, whether or not the affected persons must move to another location. OP/BP 4.36 This policy seeks to: Forests • Realize the potential of forests to reduce poverty in a sustainable manner; • Integrate forests effectively into sustainable economic development; and • Protect the vital local and global environmental services and values of forests. OP/BP 4.37 To ensure quality and safety in the design and construction of new dams and the rehabilitation of existing Safety of Dams dams, and in carrying out activities that may be affected by an existing dam. OP/BP 7.50 To ensure that Bank-financed projects affecting international waterways would not affect relations Projects in International between: Waterways • The Bank and its borrowers and between states; and • The efficient utilization and protection of international waterways. OP/BP 7.60 To ensure that projects in disputed areas are dealt with at the earliest possible stage, so as not to affect Projects in Disputed Areas relations between the Bank and its member countries, or between the borrower and neighboring countries; so as not to prejudice the position of either the Bank or the countries concerned. 109 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table A.2: IFC Policy and Performance Standards Key Content and Comparison with IFC’s 1998 Safeguards Performance Standard 1: Social and Environmental Assessment and Management System Relevant existing • OP 4.01 Environmental Assessment safeguards and IFC • OP 7.50 International Waterways documents Differences in IFC (not the client) categorizes projects. approach or Documentation and processes are driven by risks and impacts, not project categorization. interpretation; Clients can conduct the assessment themselves or outsource it. Moves away from a requirement to consult “at other least twice” during the assessment process, to an ongoing and iterative consultation process throughout the life observations of the project. International Waterways (OP 7.50) is no longer a freestanding policy because all transboundary impacts are considered as part of the assessment process. The Performance Standard does not require the client to compare project alternatives to the “without project” situation. Performance Standard 2: Labor and Working Conditions Relevant existing • Forced Labor and Harmful Child Labor safeguards and IFC • Occupational Health and Safety Guidelines documents Differences in Refers to child labor instead of “harmful” child labor. approach or interpretation Performance Standard 3: Pollution Prevention and Abatement Relevant existing • No directly equivalent policy safeguards and IFC • Related policies/documents: documents OP 4.01 Environmental Assessment OP 4.09 Pest Management Pollution Prevention and Abatement Handbook IFC Environmental, Health, and Safety Guidelines Differences in Sets out the principles of the Pollution Prevention and Abatement Handbook (PPAH) and EHS Guidelines as a approach or freestanding policy. interpretation Pest Management (OP 4.09) is no longer a freestanding policy because pesticides are addressed as part of a more inclusive approach to pollution prevention and abatement established by Performance Standard 3. Addresses climate change through quantification and monitoring of significant greenhouse gas (GHG) emissions more than 100,000 tons per annum. Performance Standard 4: Community Health, Safety, and Security Relevant existing • No directly equivalent policy safeguards and IFC • Related policy: documents OP 4.37 Safety of Dams Differences in The Performance Standard builds on OP 4.37 Safety of Dams and addresses structural safety through a risk- approach or based approach to the design, construction, and operation of all project equipment and infrastructure that may interpretation pose risks—not just dams. No longer applies the 15-meter threshold for dam height. 110 A P P E N D I X A : S A F E G U A R D A N D S U S TA I N A B I L I T Y P O L I C I E S Table A.2: IFC Policy and Performance Standards (cont'd) Key Content and Comparison with IFC’s 1998 Safeguards Performance Standard 5: Land Acquisition and Involuntary Resettlement Relevant existing • World Bank OP 4.30 Involuntary Resettlement (1990 version) safeguards and IFC documents Differences in Articulates private sector responsibilities under government-managed resettlement to complement government approach or activities, if permitted by the responsible government agency, to achieve outcomes consistent with the interpretation Performance Standard. Requires clients to monitor government-managed resettlement activities through completion. Performance Standard 6: Biodiversity Conservation and Sustainable Natural Resource Management Relevant existing • OP 4.04 Natural Habitats safeguards and IFC • OP 4.36 Forestry documents Differences in The Performance Standard’s requirements draw from OP 4.36 Forestry and OP 4.04 Natural Habitats and approach or establish a comprehensive approach to biodiversity conservation and sustainable natural resource management. interpretation Performance Standard 7: Indigenous Peoples Relevant existing • OP 4.20 Indigenous Peoples (1991 World Bank version) safeguards and IFC documents Differences in Tailors the requirements to the responsibilities of the private sector. approach or Moves away from a requirement for a freestanding Indigenous Peoples Plan to a more flexible and broader interpretation community development plan with components for indigenous peoples, where indigenous peoples are integrated into larger affected communities. Performance Standard 8: Cultural Heritage Relevant existing • World Bank OPN 11.03 Cultural Property (1986 version) safeguards and IFC documents Differences in Uses the term “cultural heritage” rather than either “cultural property” (as in OPN 11.03) or “physical cultural approach or resources” in order to capture the concept of intangible cultural heritage (addressed in the project’s use of interpretation cultural heritage). 111 APPENDIX B: PORTFOLIO EVALUATION METHODOLOGY Desk Review of the Portfolio confidence interval at the level for most subcat- World Bank egories is less robust than at the aggregated Portfolio review objective sample level. Among the regions, the sample A desk review was undertaken of a representative is more robust for Africa, East Asia and Pacific, sample of projects with significant safeguard risks Europe and Central Asia, and Latin America and for the period, fiscal years 1999–2008. The purpose the Caribbean and yielded a confidence interval of the review was to assess the extent to which the of ±13 percent (see table B.4). The results for Bank’s projects fulfilled the process requirements South Asia and Middle East and North Africa and objectives of the safeguard policies. should be treated with caution. Among the EA categories, the sample is more robust for category Project universe for the portfolio review A and B and yielded a confidence interval of ±10 The gross population addressed by the review percent (see table B.3). The results for category was 2,495 projects approved during the 10-year FI should also be treated with caution. period beginning in fiscal 1999 (when OP 4.01 was introduced) and ending in fiscal 2008. From IFC this population, projects that were assigned EA Portfolio review objective category C (minimal impacts) and U (unclassi- The objective was to assess the extent to which fied) were excluded, yielding a universe of 1,404 IFC’s projects met safeguard and sustainability projects that were assigned EA category A, B, or FI. framework process and performance objectives and to compare the pre-Performance Standard Sample size and robustness projects (fiscal 1999–April 2006) with post-Perfor- From the universe of 1,404 projects, a sample mance Standard projects (May 2006–fiscal 2008). of 252 projects was chosen, giving a confidence interval of ±5.6 percent at a confidence level of 95 Project universe for the portfolio review percent. This sample was drawn from the universe The gross population addressed by the review by applying random numbers. A complete list of was 403 pre-Performance Standard projects sampled projects is available on request. and 220 post-Performance Standard projects. The population included A, B, and FI projects Distribution of the portfolio review sample but excluded dropped projects, C projects, and The distribution of the universe and sample, by projects without categorization. region and network, is shown in tables B.1 and B.2. In order to improve the confidence interval Sample size and robustness for EA category-A projects, a higher percentage From the population of 623 projects, a sample of this group was sampled relative to its share size of 63 projects (39 non-FI and 24 FI projects) in the population (table B.3). The distribution was chosen, giving a confidence interval of of the sample with respect to the universe is ±11.7 percent, at a confidence level of 95 shown by region (table B.4). In each case, the percent. The sample was stratified to mimic the distribution of projects in subcategories is similar population, based on region, industry sector, for the universe and the sample. However, the and environmental category (A, B, FI). The 113 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table B.1: Bank Portfolio Review Universe Number of EA Category A, B, and FI Projects Approved during FY99–08, by Region and Network Network Environmentally and Socially Financial and Human Poverty Reduction Sustainable Private Sector Development and Economic Development Development Network Infrastructure Management Total Region Number % Number % Number % Number % Number % Number % AFR 112 6 24 0 80 5 152 9 17 3 385 24 EAP 70 5 3 0 14 1 134 13 0 221 19 ECA 116 8 14 1 23 2 136 8 9 0 298 19 LCR 88 6 8 1 46 3 97 8 3 0 242 17 MNA 30 1 0 19 2 49 4 0 98 8 SAR 52 3 11 1 31 4 65 5 1 0 160 12 Total 468 30 60 3 213 16 633 47 30 4 1404 100 Table B.2: Bank Portfolio Review Sample Number of EA Category A, B, and FI Projects Approved during FY99–08, by Region and Network Network Environmentally and Socially Financial and Human Poverty Reduction Sustainable Private Sector Development and Economic Development Development Network Infrastructure Management Total Region Number % Number % Number % Number % Number % Number % AFR 16 8 1 2 13 6 23 11 7 1 60 27 EAP 13 5 0 2 1 34 10 0 49 16 ECA 21 8 3 1 4 2 20 10 1 1 49 21 LCR 14 6 2 1 8 3 19 7 1 0 44 17 MNA 3 2 0 5 1 11 3 0 19 7 SAR 8 4 2 1 9 2 12 5 0 31 11 Total 75 33 8 4 41 15 119 45 9 2 252 100 stratified pre-Performance Standard sample mental and Social Effects ratings of XPSRs. The was drawn from the randomly sampled XPSRs.1 reports provide information and ratings on This sampling methodology allowed IEG to use individual performance indicators, including environmental and social review reports that compliance with Safeguard Policies and EHS had been prepared to validate the Environ- Guidelines. 114 A P P E N D I X B : P O R T F O L I O E VA L U AT I O N M E T H O D O L O G Y Table B.3: Bank Portfolio Review Population and Sample by EA Category EA category Population (no.) % of population Sample (no.) % of sample Confidence interval (%) A 219 16 60 24 10.8 B 1076 77 176 70 6.8 FI 109 8 16 6 22.7 Total 1404 100 252 100 5.6 Table B.4: Bank Portfolio Review Population and Sample by Region Population % of Sample % of Confidence Region (no.) population (no.) sample interval (%) AFR 385 27 60 24 11.6 EAP 221 16 49 19 12.4 ECA 298 21 49 19 12.8 LCR 242 17 44 17 13.4 MNA 98 7 19 8 20.3 SAR 160 11 31 12 15.9 Total 1404 100 252 100 5.6 The Safeguard Evaluation Questionnaires Distribution of the portfolio review sample (SEQs) for the portfolio review were designed The distribution of the population and sample primarily for non-FI projects. Since Annual by region, industry department, and Environ- Environmental Performance Reports (AEPRs) mental Screening Category (ESC) is shown for were available for only 10 post-Performance post-Performance Standard (tables B.5 and B.6) Standard FI projects, and only 5 post-Perfor- and pre–Performance Standard projects (tables mance Standard FI projects had requirements B.7 and B.8). In each case, the distribution of to make sure that their subprojects complied projects in subcategories is similar for the popula- with Performance Standards, the sample for tion and the sample. FI evaluation was expanded to 136 invest- ments—94 pre-Performance Standard projects MIGA from the XPSR database that contains compli- IEG-MIGA undertook a portfolio review of a ance information regarding IFC’s standard stratified sample of 35 MIGA projects approved requirements 2 for FIs, and 42 randomly during fiscal years 2000–09. More recent projects sampled post-Performance Standard projects were oversampled to facilitate findings on with AEPRs submitted. The expanded FI sample MIGA’s implementation of its current policies of 136 projects from the population of 250 and standards. Therefore, projects processed yields a confidence interval of ±5.7 percent, at by MIGA under the 2007 PPSSES account for a confidence level of 95 percent. 40 percent (14/35) of the sample. Projects 115 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table B.5: IFC Portfolio Study Population for Post-Performance Standard Projects Region ESC Total Department code Africa Asia ECA LAC MENA World A B FI CAG 6 12 11 13 1 1 42 2 44 10.9% CFN 8 10 5 5 3 31 31 7.7% CGF 21 24 30 34 15 4 128 128 31.8% CGM 15 17 29 12 12 2 2 85 87 21.6% CHE 5 6 3 3 5 18 4 22 5.5% CIN 1 19 7 12 7 1 2 41 4 47 11.7% CIT 3 2 1 6 6 1.5% COC 3 9 7 6 4 4 4 29 33 8.2% CSF 1 2 1 1 5 5 1.2% Total 62 100 95 86 48 12 8 226 169 403 100.0% 15.4% 24.8% 23.6% 21.3% 11.9% 3.0% 2.0% 56.1% 41.9% 100.0% Table B.6: IFC Portfolio Study SEQ Sample for Post-Performance Standard Projects Region ESC Total Department code Africa Asia ECA LAC MENA World A B FI CAG 1 2 1 4 4 9.8% CFN 1 1 1 3 3 7.3% CGF 3 1 2 7 13 13 31.7% CGM 1 1 4 2 1 9 9 22.0% CHE 1 1 2 2 4.9% CIN 3 2 4 1 5 12.2% CIT 1 1 1 2.4% COC 1 1 1 1 2 3 7.3% CSF 1 1 1 2.4% Total 6 10 10 9 4 2 1 23 17 41 100.0% 14.6% 24.4% 24.4% 22.0% 9.8% 4.9% 2.4% 56.1% 41.5% 100.0% supported by the MIGA-Japan Trust Funds were For each project in the sample, the study also oversampled (4 projects). Figures B.1 and evaluated the effectiveness of MIGA’s B.2 highlight the sample composition by region framework for preventing and mitigating and sector. adverse environmental and social impacts 116 A P P E N D I X B : P O R T F O L I O E VA L U AT I O N M E T H O D O L O G Y Table B.7: IFC Portfolio Study Population for Pre-Performance Standard XPSR Projects Region ESC Total Department code Africa Asia ECA LAC MENA World A B FI CAG 3 3 3 8 16 1 17 7.7% CFN 2 4 2 1 1 1 11 11 5.0% CGF 7 12 28 14 6 1 66 67 30.5% CGM 5 13 19 18 7 1 61 62 28.2% CHE 5 7 12 12 5.5% CIN 1 3 4 20 1 2 26 1 29 13.2% CIT 1 3 2 2 7 1 8 3.6% COC 4 3 3 4 6 7 1 14 6.4% Total 23 41 66 74 15 1 9 130 81 220 100.0% 10.5% 18.6% 30.0% 33.6% 6.8% 0.5% 4.1% 59.1% 36.8% 100.0% Table B.8: IFC Portfolio Study SEQ Sample for Pre-Performance Standard Projects Region ESC Total Department code Africa Asia ECA LAC MENA World A B FI CAG 1 1 1 1 2 9.1% CFN 1 1 1 4.5% CGF 1 2 1 2 1 1 6 7 31.8% CGM 1 2 2 1 6 6 27.3% CHE 1 1 1 4.5% CIN 1 2 3 3 13.6% CIT 1 1 1 4.5% COC 1 1 1 4.5% Total 2 4 7 7 2 1 13 8 22 100.0% 9.1% 18.2% 31.8% 31.8% 9.1% 4.5% 59.1% 36.4% 100.0% and enhancing client capacity for their applicable standards have been implemented. management. The evaluation was based on The examination was structured around a an examination of the extent to which MIGA’s project-specific evaluation questionnaire, policies and processes for reviewing environ- which was completed based on a review of mental and social risks, monitoring project relevant documents for each project, such performance, and ensuring compliance with as Contracts of Guarantee, Environmental 117 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Figure B.1: MIGA Sample by Region Figure B.2: MIGA Sample by Sector East Asia Middle Infrastructure Oil, Gas & 9% East/North 26% Mining South Asia Africa 5% 11% 6% Latin America/ Carribean 14% Africa 40% Agriculture, Manufacturing, Services & Tourism Europe and 43% Central Asia Financial 20% 26% Impact Analyses, Environmental and Social projects for study included adequate coverage of Review Summaries, Environmental and Social all safeguard policies, reflecting the distributions Clearance Memoranda, monitoring and other of the population of projects in terms of project reports, as available in project files. The result- status (active/completed); EA category; regions; ing portfolio review data are the main basis for and sectors. Of the field study sample, 14 this report. projects were active and 15 were completed at the time of the field visits. Twelve projects Field Case Studies belonged to EA category A, 15 to category B, World Bank and 2 to category FI. The largest number in any The study draws upon findings from 31 field- region was Africa (9), followed by East Asia and based, project case studies. Criteria for selecting Pacific (7), Europe and Central Asia and Latin Table B.9: World Bank Purposeful Sample for Field Visits Region EA category Sector AFR EAP ECA LAC MNA SAR A B C FI Unassigned Total Energy 5 2 2 4 3 2 9 Environment 1 2 1 1 1 1 1 4 Health 1 1 1 Rural 2 1 1 4 4 Transport 2 1 1 1 4 1 5 Urban 2 1 2 1 2 4 6 Water 2 2 2 Total 11 7 4 4 2 3 14 13 2 1 1 31 Note: The count for total number of projects is 29 if Chad-Cameroon pipeline is taken as one project. 118 A P P E N D I X B : P O R T F O L I O E VA L U AT I O N M E T H O D O L O G Y America and the Caribbean (4 each), South Asia selected for field visits to obtain in-depth (3), and Middle East and North Africa (2). In findings from industry sectors and regions that terms of sectors, Energy, Mining and Telecom were expected to provide rich lessons for the (9) and Transport ( 7) had the largest numbers, study, see table B.10. One mining project in while Environment and Urban Development Africa was also included in the random sample, had 3 each, followed by Social Development and making a total of 18 portfolio and field study Water (2 each), and Health and Social Protection projects. (1 each). The field study teams included experi- enced local safeguards consultants. The field As shown in the table, five category-A projects case study teams filled questionnaires similar to were sampled, including three in Oil, Gas, those used by the portfolio review for compari- Mining, and Chemicals (COC), one in Infrastruc- son purposes and supplemented them, to the ture (CIN), and one in Global Manufacturing extent possible, with face-to-face interviews with (CGM). multiple stakeholders—beneficiaries, govern- ment officials, Bank staff, project officials, and Background Papers civil society representatives. Three background papers were commissioned as follows: IFC To complement the stratified random sample, • Safeguards Aspects of Using Country Systems. a purposeful sample of 18 projects were This paper examines the Bank pilot program Table B.10: IFC Purposeful Sample for Field Visits Department code Region ESC Pre-Performance Standard Africa Asia ECA LAC A B FI Total CAG 1 1 1 CFN 1 1 1 CGF 1 1 2 2 CGM 1 1 1 CIN 1 1 1 COC 1 1 1 Total 1 1 2 3 2 2 3 7 Post-Performance Standard CAG 3 1 4 4 CGN 1 1 1 1 2 CIN 2 2 2 COC 1 1 1 2 1 3 Total 2 7 2 3 8 11 Grand total 3 8 2 5 5 10 3 18 119 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table B.11: Distribution of Screened Literature Survey Documents Type of document Number screened IEG documents (Sector Reports, Country Assistance Evaluations, Background Papers) 304 Bank Working Papers 195 Country Assistance Strategies (CAS) 130 Board Reports 129 Publications 91 Articles from NGO sources 51 Peer-reviewed articles and others 9 Country Environmental Analysis (CEA) 5 to promote the use of country systems for en- The key findings from the papers have been vironmental and social safeguards. These pilot included within the evaluation report. projects were intended to increase country ownership, build institutional capacity, and Literature Review increase cost effectiveness. The literature review on environmental and • Social Safeguards. This paper evaluates the social safeguard policies drew upon documents Bank social safeguards in terms of their effec- from the Quality Assurance and Compliance tiveness in preventing and litigating adverse Unit (QACU) of the Operational Services depart- environmental and social impacts and their ment of OPCS; research and analytical literature impact on client capacity. generated by the World Bank and IEG; as well • Accountability Mechanisms. This paper looks as external sources such as academic journals, at the Bank Inspection Panel and the IFC/ civil society organizations, and news sources. MIGA Compliance Advisor Ombudsman to Based on an appropriate keyword search, determine how these mechanisms affect the documents were selected for review as indicated implementation of safeguard and sustainability in Table B.11. policies. 120 , APPENDIX C: SURVEY, INTERVIEW AND FOCUS GROUP METHODOLOGIES IEG Staff Survey Task team leaders that were no longer listed in World Bank and IFC the Bank’s directory were excluded. All current An Internet-based staff survey was conducted for identifiable environmental and social safeguards Bank (i) project task team leaders and (ii) environ- specialists were covered. In all, the survey was mental and social safeguard specialists, and for IFC administered to 1,195 task team leaders, 81 (iii) investment officers and (iv) environmental environmental safeguards specialists, and 53 and social specialists. The purpose of the survey social safeguard specialists. After excluding was to draw upon the project-level experiences incomplete and invalid responses, the response of the task team leaders, investment officers, and rates were 50 percent for task team leaders specialists regarding the process and efficacy of and 79 percent for specialists. At these levels the Bank’s safeguard policies and IFC’s Perfor- of response, the survey results are valid within mance Standards and areas for improvement. ±2.8 percent for the population of team leaders The survey was administered by a third-party and specialists taken together; ±3.2 percent for firm to ensure the anonymity of the respon- the team leaders alone; and ±4.6 percent for dents. The respondents were asked to complete the specialists alone, all at a confidence level of the survey, based on the last project for which 95 percent. The salient figures for the survey they participated to the greatest extent during are summarized in table C.1.Similarly for IFC, both the preparation/appraisal and implemen- the survey included 500 investment officers tation stages. Separate survey instruments were and 51 specialists. The survey results are valid used for each of the team leader/investment within ±3.8 percent confidence interval for the officer and specialist groups in keeping with their population of investment officers and special- respective roles in the environmental and social ists, at a confidence level of 95 percent (see appraisal and supervision process. table C.2). The survey covered team leaders of all projects MIGA approved during fiscal years 1999–2008, after Because of the small number of staff involved, a excluding EA category-U (unclassified) projects. similar survey was not conducted for MIGA. Table C.1: Staff Survey on Bank Safeguard Policies: Population and Sample Size Population Neta response rate (%) Confidence interval (%) Project task team leaders 1,195 50 ±3.2 Environmental and social 135 79 ±4.6 ±2.8 safeguard specialists a. After excluding incomplete and invalid responses. 121 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Table C.2: Staff Survey on IFC Performance Standards: Population and Sample Size Population Neta response rate (%) Confidence interval (%) Investment officers 500 51 ±4.3 Environmental and social 51 86 ±5.5 ±3.8 specialists a. After excluding incomplete and invalid responses. IEG NGO Survey IEG Discussions and Focus Group An Internet-based survey was conducted for Meetings with WBG Staffs nongovernmental organizations (NGOs) in all Face-to-face interviews using structured regions of the world, covering those operating at questionnaires were also held with 55 Bank the local, subnational, national, and global levels. managers. The questions varied slightly The survey covered the experience and opinions depending on the responsibility areas of the of NGOs on the content, conduct, and efficacy respondents. Managers interviewed included of Bank safeguard policies and the Performance the safeguard consumers (Country Directors Standards of IFC and MIGA. The survey was and Non-Environmental and Social Sector administered using the updated list maintained Managers), and the safeguard service provid- by the World Bank’s External Affairs Department, ers (Regional Safeguard Advisors and Environ- comprising NGOs that have shown interest in the mental and Social Sector Managers). Additional developmental efforts of the WBG. Of the listed interviews were conducted with Operations NGOs, 174 (40 percent of those with active email Directors/Quality Managers. IEG conducted addresses) completed the survey. 26 interviews in parallel with IFC management (Country/Regional Directors, Sector Directors, IEG Interviews with Clients and CES/Advisory Services). A number of instruments were used to validate the results of the portfolio review and to provide Discussions were held with the Sustainable additional perspectives and information. As Development Network Council at the concept part of the field study missions, supplementary stage and also (at their request) with a group of interviews with structured questionnaires were forestry staff. held with Bank and IFC clients/stakeholders during the field visits. This included represen- Three formal facilitated focus group discussions tatives from both financial intermediaries and were also arranged. In each case a brief presenta- nonfinancial intermediaries. The feedback tion was made by IEG, and certain questions were obtained from the interviews with project proposed. In all three cases, however, partici- management unit directors, representatives of pants were then encouraged to put forward any government planning and finance ministries, and issues they believed to be pertinent. One group other interested stakeholders was supplemented covered social specialists, one covered environ- with additional telephone interviews with mental specialists, and one was for IFC environ- implementing agency staffs to obtain feedback mental and social specialists. on client experience and perceptions regarding safeguards implementation. 122 APPENDIX D: FEEDBACK FROM NONGOVERNMENTAL ORGANIZATIONS Feedback was received from over 200 NGOs, (ii) The impacts associated with supply either in the open questions in the electronic chains should be assessed in all projects. survey, through a focus group discussion or in (iii) In the environmental and social process the form of documents sent to the IEG team. for category-A projects, independent ex- Typical comments are recorded below, grouped perts should be retained. by theme. (iv) Action plans and monitoring reports should be translated into languages un- Compliance, Enforcement, and derstandable and accessible to affected Monitoring peoples. • “Self-compliance by clients and private firms (v) Environmental and social monitoring should be banned.” (progress) reports should be disclosed • “Independent monitoring of resettlement to affected people and the public. plans should be mandatory.” (vi) Impacts of climate change due to defor- • “Performance Standards are much better on estation should be assessed in the social paper than in practice since not all staff treat and environmental assessment process. them with equal seriousness. Enforcement can (vii) Free, prior, and informed consent be quite weak depending on the commitment should be obtained from the indigenous of local staff. While the IFC has done a better peoples. job than the Bank on educating its staff and (viii)Areas having biodiversity that support partners on its standards, there seems to be an local communities’ basic needs should institutional bias against correcting violations be a criterion in Performance Standard 6. [of standards].” (ix) In areas of critical habitat, the client • “The Bank needs to make assessments of should not significantly convert or de- whether environmental/social conditions grade such habitat. actually improved over time after Bank in- (x) Environmental and social monitoring terventions, because too much assessment reports should be disclosed on IFC’s is devoted to measuring adequacy on inputs/ website. outputs rather than outcomes.” (xi) IFC’s monitoring results including site • “Performance Standards are over-discretionary visit reports should also be disclosed on and rely too much on client-generated infor- IFC’s website.” mation and client self-monitoring.” • “One report claims that IFC has not complied Transparency properly with its post-2006 Performance Stan- • “In a poor governance context, total access to dards and makes 11 recommendations: information should be compulsory.” (i) Performance Standard 1: Restoring the livelihoods and living standards of proj- Use of Country Systems ect affected persons. Non-land-related • “IBRD/IDA would make a colossal mistake to displacement is not covered and should declare the era of safeguards to be over (that be added. is, to rely on country systems to deliver on the 123 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Bank’s sustainability and poverty reduction Changing World mandates).” • “The Bank safeguards were effective when • “National systems should only be seen as adopted but have not been updated, and the equivalent if the appropriate national policy is world has changed in the meantime.” in operation. Planning to have a new national • “Despite the importance of the safeguard poli- policy is not enough.” cies and their achievements, the World Bank’s • “Governments should not decide, themselves, policy framework has come under increasing in assessing whether national systems are pressure since the late 1990s. The Bank has equivalent to Bank systems.” failed to comprehensively implement and con- • “The World Bank has initiated a number of sistently update its safeguard policies based specific processes supposedly to simplify the on the latest best-practice standards and the safeguard policies. While some of these pro- findings of multistakeholder reviews.” cesses, such as the IFC’s policy review and • “Whereas certain Bank policies may once have the Bank’s country systems approach, are on- been considered “best practice,” there are now going, the intended outcome is clear: a shift institutions and accepted standards that offer from explicit mandatory policies, to which superior social and environmental protections. the WBG can be held accountable, to flexible For example, several private banks, including principles permitting the investor and/or bor- Citigroup, Bank of America, and HSBC, have rowing government to determine the project’s stronger biodiversity conservation and forest social and environmental requirements. The protection policies than the WBG. EBRD, the Bank’s proposed country systems approach UK export credit agency, and others explicitly does not include any concrete measures to address principles of democracy and human effectively strengthen national capacities.” rights in their policy requirements. Corpora- , tions such Alcoa, BP DuPont, and Shell have Social Impacts also eclipsed the Bank in setting greenhouse • “IBRD/IDA needs to broaden safeguard cover- gas emissions reduction targets.” age of social impacts, at least to match IFC/ • “On climate change, the ADB is more stringent MIGA’s.” than the Bank.” • “Environmental/social assessment frame- • “The safeguard policies need to take into ac- work for development policy lending must count the imperatives of investment lending be strengthened as DPLs are accounting for reform.” an increasing share of lending.” • “Gaps in standards between IBRD/IDA and Implementation IFC/MIGA should be harmonized upwards, • “Implementation of safeguard policies by while providing flexibility for different busi- the WBG has weakened over the last decade. ness models.” Much WBG investment is no longer covered • “Performance Standards do not go far enough by the safeguards, such as program and DPL to protect indigenous people.” lending and lending to intermediaries.” • “Special provisions are needed in respect of • “In practice, the implementation of safeguard safeguards in conflict areas or where there is policies leaves much to be desired.” poor governance (the Chad-Cameroon Pipe- • “Internal processes need to be looked at; what line project is a good example of this).” are the incentives for staff to comply with safe- • “The issue of human rights should not be guard policies?” avoided.” 124 APPENDIX E: MAIN MESSAGES FROM FOCUS GROUP DISCUSSIONS WITH WORLD BANK GROUP STAFF World Bank Environmental Specialists should focus on achieving results by revising • Consistency of safeguard application: There is them to be more flexible, living guidance tools. a need for more transparency and consistency Greater harmonization of safeguard policies in applying safeguards, avoiding reversal of with other international financial institutions safeguards decisions. The current process is and development partners is needed. arbitrary and unpredictable. • Delegation of lower-risk projects to regional World Bank Social Development sector management units has reduced the in- Specialists fluence of the Regional Safeguards Advisor. In Compliance approach. Effectiveness of Africa, delegation has helped safeguard units emphasis on compliance depends on the focus on highest-risk projects. country context. In middle-income countries it is • Client capacity building: Country ownership less useful and sometimes counterproductive. In is crucial, and the key to that is focusing on low-income countries with weaker capacity it can using safeguards as a risk management tool to be a useful entry point to help the client address strengthen a country’s institutions, to manage social issues. its own environmental and social risks, not on strict procedural compliance with Bank Delegation to sector managers. Unanimous safeguards policies. Regional capacity-building view that this attempt to streamline safeguards initiatives might be more effective and cost- work has not worked. Sector managers are efficient. often reluctant to take on this responsibil- • Staff incentives: There is a serious shortage ity. Delegated projects are more vulnerable to of safeguards specialists and their work is not reputational risk. recognized as important or rewarded as an attractive career path. Safeguard outcomes Social safeguard policies. The Involuntary should be an element of task team leader’s Resettlement policy needs customization to the and project’s performance rating. context and faces implementation challenges. • Inspection Panel is well-intentioned but in- The Indigenous Peoples policy has more volves high costs and lowers staff morale; an problems with the policy itself because the social intermediate, complaint resolution mecha- context varies enormously across regions and nism is needed to resolve issues initially, when- countries. If the policy was based on vulnerabil- ever possible. ity rather than ethnicity, or was integrated within • Supervision: To overcome the limitation of a broader assessment of community impacts, insufficient supervision budgets for adequate governments would be more willing to engage safeguard support, establish regional rosters with it. of qualified local consultants as an alternative. • Focus on results, not process: Safeguards are Advice to senior management perceived as a compliance issue, not as an • Need a stronger policy framework for social element of the development process. Bank assessment parallel to the EA policy, which should not drop or weaken safeguards, but looks at risks and opportunities and includes 125 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D community impacts, gender, and labor re- • The Environmental and Social Review Docu- trenchment. ment is not very useful for risk management. It • Need to train a core of social specialists to serves mainly to provide information to man- harmonize approach to safeguards implemen- agers. The tool is necessary, but the format tation. can be improved. • Need to integrate social agenda across sector • On supply chains, more use of third-party veri- managers and task teams to create broader fication would be preferable to burdening the ownership of social issues. client. There is a need to differentiate types of supply chain as the potential leverage with IFC Environmental and Social Specialists clients differs. • Positive impact on clients but sometimes • Advisory Services cannot currently be used to time-consuming. The market (consumer pref- help clients comply with Performance Standards; erences, public image, etc.) helps drive the it could be of use especially for weaker clients. culture shift in favor of Performance Standards. • IFC stamp of approval helps clients gain Suggestions to improve implementation of credibility with host governments and attract Performance Standards additional finance. Greater acceptance of en- • Move from compliance to a risk-based ap- vironmental and social standards through IFC proach. demonstration effect and adoption of Equator • Provide clearer indicators for effectiveness in Principles. Performance Standards are rarely the Social and Environmental Management an issue in project approval. System. • Value added on environmental and social as- • Financial Intermediaries need more attention pects comes from greater IFC focus on im- to ensure that they are paying sufficient atten- plementation and outcomes. (Performance tion to community engagement issues. Standard 1 lays the foundation). But also • Consider integration of new, emerging issues greater value through Performance Standard 2 (e.g., climate change, human rights) into exist- (Labor and Health & Safety) and Performance ing Performance Standards. Standard 4 (Community Impacts). 126 ENDNOTES Management Response 2002). IEG prepared a process review of Environmental 1. Paris Declaration on Aid Effectiveness: Assessments and National Environmental Action Plans Ownership, Harmonization, Alignment, Results and in 1996. Furthermore, some IEG sector evaluations Mutual Accountability. Second High-Level Forum on [e.g., Extractive Industries (2003), Community-Driven Aid Effectiveness. Paris, March 2005. Development, Low Income Countries under Stress 2. Accra Agenda for Action. Third High-Level (2006)], and some Country Assistance Evaluations have Forum on Aid Effectiveness. Accra, September 2008. included background papers or significant discussions on safeguard issues. At the project level, numerous IEG Chapter 1 Project Performance Assessment Reports have covered 1. The World Bank consists of two complementary safeguards issues where relevant. institutions—the International Bank for Reconstruc- 5. For more on the rationale of the evaluation, see tion and Development (IBRD) focuses on middle IEG 2009b. income and creditworthy poor countries, while the 6. Also see Picciotto, van Wicklin, and Rice 2001. International Development Association (IDA) focuses 7. See also IEG 2009b. on the poorest countries in the world. Together they 8. Investment Lending Reform—Concept Note. provide low-interest loans, interest-free credits and 9. Beginning in 2003 IFC advised 10 leading banks grants primarily to the public sector in developing from seven countries on the adoption of the Equator countries. The International Finance Corporation Principles, a voluntary set of guidelines based on IFC (IFC) provides investments and advisory services policies. A revised set of principles was launched in to build the private sector in developing countries, July 2006, reflecting IFC’s 2006 Policy and Performance while the Multilateral Investment Guarantee Agency Standards on Social and Environmental Sustainability. (MIGA) provides political risk insurance. The World By February 2009, 68 banks representing the majority Bank Group (WBG) includes a fifth institution, the of project investments in developing countries, includ- International Centre for Settlement of Investment ing major institutions such as Citigroup and HSBC, had Disputes (ICSID), which does not provide financial declared their adherence to the Principles and Perfor- support to its clients. For the purposes of this report mance Standards. the term WBG refers only to the World Bank, IFC, and 10. See, for example, “Rome Declaration on MIGA. Harmonization,” Rome, Italy; February 25, 2003, on 2. See, for example, the World Bank website at an “effort to harmonize the operational policies, http://www.worldbank.org. procedures, and practices of our institutions [IFIs]” 3. See annex 1, available on this report’s website at and “A Common Framework for Environmental Assess- http://www.worldbank.org/ieg. ment: A Good Practice Note,” Multilateral Financial 4. The Quality Assurance Group (QAG) has used Institutions Working Group on Environment, February sample project assessments and surveys to assess 28, 2005, prepared with contributions from the World the safeguard policies, as part of the wider canvas Bank Group. of its Quality-at-Entry Assessments and Quality-of- 11. For example, Environmental Defense Fund Supervision Assessments since 1999. Three Reviews of (www.edf.org), Shannon Lawrence, Retreat from the Environmental Assessment have also been prepared by Safeguard Policies, 2005, and Oxfam International the Bank’s Environment Department (1993, 1997, and (www.oxfam.org). 127 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 12. World Bank 2005. generated at least 18 months of operating revenues 13. World Bank 2008b. Also see “First Year Review (covered by at least two sets of company annual of Implementation of Incremental Scaled-Up Program audited accounts). Additional performance indicators to Pilot Use of Country Systems to Address Environ- were sourced from earlier 348 IEG’s environmental mental and Social Safeguard Issues in Bank-Supported and social reviews for the XPSR projects appraised Projects,” July 14, 2009. between fiscal years 1999 and 2004. To analyze environ- 14. In the text the term “WBG safeguard policy mental and social performance in FIs, the FI sample framework” refers to the Bank safeguard policies and was expanded by 42 randomly sampled post-Perfor- IFC/MIGA Policy and Performance Standards on Social mance Standard projects that had received first Annual & Environmental Sustainability. Environmental Performance Reports. 15. “Development policy lending is rapidly disburs- 21. The evaluation includes findings up to the ing policy-based financing, which the Bank provides in third quarter of fiscal year 2009. the form of loans or grants to help a borrower address 22. IEG-MIGA’s sample included category A (4 actual or anticipated development financing require- projects), B (21), C (5), and FI (4). Prior to the PPSSES, ments that have domestic or external origins. … They financial sector projects were classified by MIGA as “C.” typically support a program of policy and institutional 23. The World Bank has introduced a new policy actions, for example, to improve the investment on Access to Information, which went into effect on climate, diversify the economy, create employment, July 1, 2010, and replaces the Information Disclosure and meet applicable international commitments.” (OP Policy (2002). 8.60 Development Policy Lending). 24. See Involuntary Resettlement Sourcebook 16. Category-A projects are likely to have significant (World Bank 2004, p. 3). adverse impacts that are sensitive, diverse, or unprec- 25. The Environmental and Health & Safety (EHS) edented; category-B projects can have potentially Guidelines that IFC uses as covenants in investment adverse impacts; category-C projects are likely to have agreements are important parts of IFC’s sustainabil- minimal or no adverse impacts; Financial Intermediary ity framework. In all, 41 environmental guidelines projects involve on-lending typically by national banks, were collected under WBG “Pollution Prevention credit institutions, and other financial intermediaries. and Abatement Handbook” in 1998. In addition, IFC The financial intermediary screens each subproj- developed its own General Health & Safety Guideline ect proposed for financing and classifies it into the and 29 industry sector guidelines. appropriate category (A, B, or C). 26. Also see IFC’s Sustainability Policy and Perfor- 17. The Bank portfolio review included 108 mance Standards—Key Differences with IFC’s 1998 completed projects and 144 active projects. In Safeguard Policies (appendix table A.2). addition, field assessments of 29 purposely selected 27. These procedures do not apply to develop- projects (15 completed, 14 active) were also carried ment policy lending (DPL) under OP 8.60, but the out for the evaluation. policies do apply to other instruments except DPLs. 18. Regional and networks samples were drawn They also apply to emergency operations under in proportion to their distribution in the universe, OP 8.00, but may be exempted when compliance with yielding a CI of ±12 percent or better for most regions, safeguards requirements would prevent the effective except for South Asia Region (±16 percent CI) and and timely achievement of the objectives of the Middle East and North Africa (±20.3 percent CI). The emergency operation. CI was ±8.1 percent for Sustainable Development 28. The Regional Safeguard Advisor confirms the Network–INF, ±10.4 percent for Sustainable Develop- applicable Environmental Category of the project at a ment Network–ESSD, and ±13.8 percent for Human safeguards review meeting with the task team leader, Development Network. which is recorded in the Integrated Safeguards Data 19. The IEG evaluation included field visits to 18 Sheet. Bank environmental and social specialists often projects. advise the borrower on Terms of Reference for relevant 20. Every year IEG selects a random sample for assessments. The borrower is responsible for prepara- XPSRs, which covers 51 percent of projects approved tion of the Environmental Assessment (EA), Environ- five calendar years prior to the current year and have mental Management Plan (EMP), Resettlement Action 128 ENDNOTES Plan (RAP), or Indigenous Peoples Plan (IPP) and 4. The NGO survey received 174 responses, consultation with affected parties. Safeguards aspects yielding an effective response rate of 40 percent. of the draft Project Appraisal Document are reviewed 5. The social assessments include socioeconomic and cleared by the Regional Safeguard Advisor, and surveys and other studies of resettlement impact the assessments and plans are publicly disclosed prior undertaken to prepare the RAP, or the assessments to appraisal. The implementation completion report of impact on indigenous peoples to prepare the IPP prepared at the end of the project assesses compliance (formerly known as the Indigenous Peoples Develop- with safeguard policy and procedural requirements ment Plan). and is subject to review by the Independent Evalua- 6. OP 4.01 does not explicitly provide for similar tion Group (IEG). policy frameworks but some regions, such as the 29. The Project Data Sheet – Early Review includes Africa Region, frequently rely on Environmental and a description of environmental and social aspects, the Social Management Frameworks (ESMFs) where a full client’s management capacity, provisional categoriza- impact assessment cannot be prepared prior to Board tion and applicable environmental and social require- approval. In other regions, such as Europe and Central ments and standards. Asia, the use of policy frameworks to address environ- 30. At appraisal the need for and content of a Peer mental issues has been discouraged, Strategic Environ- Review Meeting (PRM) as well as evidence of free prior mental Assessment being the instrument of choice. and informed consultation are considered. 7. Chi-square test independence probability p was 31. Disclosure time before a Board Meeting for much below 5 percent, with 95 percent confidence the Summary of the Proposed Investment, ESRS and level. Environmental and Social Action Plan is 60 days for 8. IFC’s quality at appraisal was 88 percent satisfac- category-A and 30 days for category-B projects. tory for 347 XPSR projects appraised in 1999–2004 and 32. The Environmental and Social Risk Rating evaluated by IEG between 2004 and 2009. score is based on operational and reputational risks, 9. EBRD Environmental and Social Policy, client management capability and compliance with approved on 12 May 2008, appendix 1. environmental and social requirements. 10. IFC Annual Report FY09, original committed 33. Since 2004 IEG has also evaluated IFC’s portfolio by industry sector. environmental and social work quality at appraisal 11. ESRP 2009, paragraph 7.2.7. They were and supervision, and role and contribution, as well as previously also required to comply with host-country performance in meeting Safeguard Policies and EHS laws according to ESRP 1998 and annex 2. Guidelines, and other requirements (such as effluents, 12. The IFC Exclusion List defines the types of air emissions, waste management, health and safety, projects that IFC does not finance such as production annual reports, etc.). It commenced evaluating compli- or trade in weapons or tobacco (ESRP 2009, p. 16). ance with Performance Standards in January 2007. 13. Production-specific indicators define pollution 34. See paragraph 13 in MIGA 2007. loads and energy consumption per unit production, for example, kilograms of SO2 emissions per ton of Chapter 2 product (SO2 kg/t) and power and heat consump- 1. The portfolio review contained projects from tion per ton of product (kWh/kg, MJ/kg). Such indica- categories A, B, and FI only. Reported results from the tors can be easier benchmarked against industry best portfolio review in the evaluation report are based on practices compared with the traditional concentra- projects with satisfactory ratings as a percentage of this tion–based indicators like milligrams of SO2 in a norm subset of the portfolio universe. cubic meter (SO2 mg/Nm3). 2. The staff surveys yielded a response rate of 50 14. The AMR templates should include the most percent (470 Bank task team leaders; 202 IFC invest- important results-based H&S indicator, namely Lost ment officers), and 80 percent from environmental Time Accidents per million working hours (LTA/ and social specialists (92 from the Bank; 38 from IFC). mill hrs), making it possible for IFC and the client to 3. Face-to-face, semi-structured interviews were benchmark health and safety performance against conducted with 50 Bank country and sector managers, publicly available industry statistics, available from and 21 IFC managers. major industrial countries. 129 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D 15. Both category-A projects, one before and one 1 (available on this report’s website at http://www. after the Performance Standards, were extractive worldbank.org/ieg). industry projects in Africa. 26. Previous IEG evaluations have also found that 16. IEG Environmental Sustainability Study 2008 OMS 2.20 was often ignored by Bank staff even though and MSME study 2008. IFC’s common requirements in it is part of the Operational Manual. ESRP 1998 for all FI projects included Training, Process 27. For example Clark, Fox, and Treakle, 2003. (policy, procedure, responsible staff), and Annual 28. The accountability mechanisms at the WBG reports. have also encouraged other IFIs to develop their 17. In the ESRP 2009 linked to the PPSSES own such mechanisms. The Inspection Panel at 15 framework, the requirements to apply Performance Years World Bank 2009. Standards for subprojects are based on the tenor, 29. For a more detailed breakdown of requests transaction sizes and the industrial sector risks in the received see 2008–09 data in annex 3, table X3.5 at portfolio, and IFC’s leverage and capacity to carry out http://www.worldbank.org/ieg. a reasonable review of the project’s risks. 18. MIGA’s 1999 Environmental Assessment Policy, Chapter 3 which applied until 2007, did not have a category FI for 1. These outcome indicators are based on the financial intermediary projects. intent of OP 4.01 Environmental Assessment which 19. The confidence interval is ±12 percent or states that “The [World] Bank requires environmen- better for most regions, except for South Asia Region tal assessment (EA) of projects proposed for [World] (±16 percent) and Middle East and North Africa Bank financing to help ensure that they are environ- (±20.3 percent). The confidence interval was ±8.1 mentally sound and sustainable, and thus to improve percent for the Sustainable Development Network- decision making.” INF, ±10.4 percent for Sustainable Development 2. See forthcoming IEG Results and Performance Network–ESSD, and ±13.8 percent for Human Report (RaPR 2010). MIGA has begun to implement Development Network. self-evaluation in FY10 and will use the same indicators 20. This is similar to the supervision performance in assessing the development outcome of guarantee of other projects using policy frameworks instead of projects. fully prepared environmental/social management 3. Outcome data from self-assessments by staff are plans. not available for MIGA due to the recent introduction 21. Number of environmental and social specialists of self-evaluation in MIGA beginning in FY10. supervising FI projects increased from one in 2005 to 4. These include BP 4.01 Environmental Assess- five in 2009. ment, BP 4.04 Natural Habitats, and BP 4.11 Physical 22. Knowledge Gap is the share of active projects Cultural Property. under supervision where monitoring information has 5. The relevant social safeguard policies are BP not been obtained in the past two or more years. 4.10 Indigenous Peoples, and BP 4.12 Involuntary 23. In IEG’s XPSR statistics 2004–2009, the perfor- Resettlement. mance in submitting satisfactory AMRs is 62 percent 6. These differences are statistically significant. for real sector projects and 60 percent for FI projects in Results for Middle East and North Africa, although regard to their EMS compliance, and only 41 percent in also good, are based on a very small sample and are regard to subprojects’ compliance with IFC’s environ- therefore not reported. mental and social requirements. 7. IEG solicited feedback from borrower 24. CAO Audit of MIGA’s Due Diligence of the representatives, project staff, and local NGOs in the Dikulushi Copper-Silver Mining Project in the countries where field studies were undertaken. The Democratic Republic of the Congo, Final Report, client feedback is based on those semi-structured November 2005. Office of the Compliance Advisor/ interviews. Ombudsman, IFC/MIGA. 8. The variation between high and low performing 25. The findings on the IPN/CAO are extracted regions was wide enough to be statistically significant. from a background paper on the WBG’s account- 9. Overall quality of client implementation in 13 ability mechanisms which is summarized in annex post-Performance Standard projects was 85 percent in 130 ENDNOTES contrast to 60 percent of 15 pre-Performance Standard Acquisition and Involuntary Resettlement, PS7: Indige- projects. nous People, and PS8: Cultural Property. The last two 10. Management of hazardous materials was are not evaluated here, since they were not applicable satisfactory in 93 percent of in 14 post-Performance to any of the MIGA projects in the review sample. Standard projects and 87 percent of 187 XPSR projects 20. As of December 31, 2009. appraised 1999–2004. 11. Implementation of EHS Guidelines was satisfac- Chapter 4 tory in 62 percent of 13 post-Performance Standard 1. Under the Pelosi Amendment, EIAs for projects projects and 67 percent of 84 XPSR projects appraised with major environmental impacts, category-A projects 1999–2004. are required to be disclosed 120 days prior to Board 12. Only 56 percent of 71 pre-Performance approval or the U.S. Executive Director is required to Standard projects and 50 percent of 38 post-Perfor- abstain or vote against a project. In practice, category-A mance Standard projects implemented the ESMS in a projects are disclosed 120 days prior and category-B satisfactory manner projects 30 days prior. IFC discloses category-A 13. Manager perceptions are from the projects 60 days prior, and category-B projects 30 days semi-structured interviews conducted by IEG with prior to the Board date. 50 Bank managers. Regional Safeguard Advisers and 2. Category-C projects were not included in the environmental and social sector managers were more portfolio review; consequently the extent of underclas- pessimistic about safeguards results than were country sification between categories B and C has not been managers and sector managers from other sectors. assessed in this evaluation. 14. The outliers were local NGOs from Latin 3. Eight risk gradations were specified: low-low, America and the Caribbean, 71 percent reported low-high, moderate-low, moderate-high, substantial- high or substantial client capacity building, and NGOs low, substantial-high, high-low, high-high. from Africa who rated effectiveness of environmental 4. That is, environment risk is low-high while social mitigation as 72 percent, and NGOs from South Asia risk is high-low, and vice versa. Region who rated effectiveness of social mitigation as 5. This estimate does not include costs incurred 70 percent. by Bank management and the Legal Department on 15. This move was motivated by the desire to project preparation and supervision or costs incurred avoid perceptions of conflict of interest when the Vice in addressing requests for investigation filed with the Presidency for Environmental and Socially Sustainable Inspection Panel. Development (ESSD) was merged with the Infrastruc- 6. Bank management informed IEG that the costs ture Vice Presidency to create a new Sustainable Develop- for category-B projects with limited impacts and risks ment Network. The transfer of QACU from ESSD to are significantly lower. OPCS was accompanied by the transfer of the Regional 7. Two projects in the extreme upper-right corner Safeguard Advisors from Regional ESSD Departments to of table 4.5 were classified as category A: one from Regional Departments with OPCS functions. the extractive industries sector and one from the 16. It should be noted that an EMS was not formally transport sector; the third, from the urban sector, was required for these XPSR projects since requirement classified as category B. for the client to develop and implement a social and 8. Given that, of the more costly half of the sample, environmental management system (SEMS) became 63 percent (12/19) are in the high cost/high benefit a formal requirement when Performance Standards quadrant, and 27 percent (7/19) in the high cost/low were approved in April 2006 benefit quadrant. 17. Environmental and Social Effects ratings FY2007–09 are based on IEG’s Environmental and Chapter 5 Social Reviews that validate XPSRs prepared by project 1. The portfolio review excluded category C and teams. unclassified projects, consequently OP 4.01 (Environ- 18. As of June 30, 2009. mental Assessment) affected all 252 projects in the 19. These are PS2: Labor and Working Conditions, review sample, OP 4.12 (Involuntary Resettlement) PS4: Community Health, Safety and Security, PS5: Land 49 percent, OP 4.10 (Indigenous Peoples) and OP 4.11 131 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D (Physical Cultural Property) 22 percent and OP 4.36 support severance. In investment operations support (Forests) 18 percent. The rest occurred in less than 12 is provided through financing of severance pay. percent of the sample. 9. This crowding-out effect has previously been 2. Prior to 2004, safeguard policies included identified by QAG assessments of Bank supervi- sectoral adjustment loans which are now subsumed sion, described in the background paper on social under DPLs and governed by their own environmen- safeguards. tal and social requirements. Sector adjustment loans 10. OP 4.12 Involuntary Resettlement, and OP 4.10 triggered the environmental assessment policy only. If Indigenous Peoples provide relatively better guidance DPLs were included from the portfolio the relevance than the other safeguard policies on supervision, of environment would decrease to 60 percent and monitoring and completion reporting. involuntary resettlement to 25 percent of the portfolio. 11. See for example, Shannon Lawrence, “Retreat 3. The pilot program is governed by a new from the Safeguard Policies,” Environmental Defense Operational Policy 4.00 “Piloting the Use of Borrower Fund, January 2005; Dana Clark, “Comments on IFC Systems to Address Environmental and Social Policy and Performance Standards,” International Safeguard Issues in Bank-Supported Projects”. The Accountability Project November, 2005. policy applies only to the World Bank, not to IFC and 12. The goal of the SPS is to promote the sustain- MIGA, consequently the UCS section of this chapter ability of project outcomes by protecting the environ- only pertains to the World Bank. ment and people from projects’ potential adverse 4. This section of the chapter summarizes the impacts by: (i) avoiding adverse impacts of projects main findings from the background paper commis- on the environment and affected people, where sioned by IEG on the use of country systems, possible; (ii) minimizing, mitigating, and/or compen- augmented by additional information gleaned from sating for adverse project impacts on the environment the staff surveys. and affected people when avoidance is not possible; 5. Before agreeing to use country systems, the and (iii) helping borrowers/clients to strengthen their Bank assesses the strengths and weaknesses of safeguard systems and develop the capacity to manage borrower safeguard systems and identifies targeted— environmental and social risks. or gap-filling—measures to strengthen such systems. 13. There is a difference between technical The Bank has developed a tool to do this analysis: assistance provided for project preparation studies or the Safeguards Diagnostic Review (SDR). The SDR project implementation, and technical assistance for evaluates the equivalence of the borrower’s system studies unrelated to actual project implementation. (the extent to which it is designed to achieve the Managers acknowledged the relevance of safeguards same objectives and adhere to the same principles as to the former but expressed concerned about the the Bank’s safeguard policies) and the acceptability practice of classifying all technical assistance projects of borrower implementation practices, track record, as category B. and capacity. Measures to achieve and sustain equiva- 14. See European Bank for Reconstruction and lence and acceptability are identified, included in the Development, Environmental and Social Policy, May legal agreement for the project, and are then actively 2008. “In projects where Indigenous Peoples are likely supervised during implementation. to be affected, the client is required to carry out an 6. That is, systems based on implementing agency/ assessment … of the impacts on Indigenous Peoples. utility corporate policies in the Fifth Power Systems Depending upon the outcome of this, the client is Development Project in India and Eskom Investment expected to first avoid adverse effects and where this Support Project in South Africa. is not feasible, to prepare an Indigenous Peoples’ 7. The first attempt to apply the UCS to a country Development Plan … so as to minimize and/or mitigate was in Mauritius. However, as of December 2009, any potential adverse impacts and identify benefits” when the second round of inquiry was undertaken by (p. 52). “The client’s proposed actions to minimize, IEG, there was insufficient progress for the Mauritius mitigate and compensate for adverse effects and to case to be assessed by IEG. identify and share benefits are required to be developed 8. Operations that include downsizing (retrench- with the informed participation of affected Indigenous ment) in the public sector or public enterprises may Peoples and contained in a time-bound plan, such as 132 ENDNOTES an Indigenous Peoples Development Plan (IPDP), or a Chapter 6 broader community development plan with separate 1. The two Performance Standards in the PPSSES components for Indigenous Peoples” (p. 53). that are not among the Bank’s safeguard policies are 15. “When the primary objective of an operation is on “Labor and Working Conditions” and “Community to move people from areas that are unfit for human Health, Safety and Security.” Additionally, IFC is habitation or, as in urban upgrading projects, to provide currently reviewing its PPSSES, and may integrate basic infrastructure or resolve problems of land tenure, additional environmental and social effects within its the guiding principle will be to minimize the disruption policy. of the affected population. The views of the affected 2. The current safeguard policies at the Bank population will be taken into account in the design and provide for Resettlement Policy Frameworks and execution of the resettlement plan, and where feasible, Indigenous Peoples Planning Frameworks. While voluntary procedures will be established to determine there is no such provision within the environmen- which households will be relocated. The plan will also tal safeguard policies, an increasing number of ensure that those who are displaced will have access projects, particularly in Africa, are making use of joint to equivalent or better employment opportunities and Environmental and Social Management Frameworks urban services.” Inter-American Development Bank, (ESMFs). Sector Operational Policy: Involuntary Resettlement, GN-1979–3, July 1998. Appendix B 16. “When the baseline information indicates that a 1. Every year IEG selects a random sample of significant number of the persons to be resettled belong XPSRs that cover projects approved five calendar years to marginal or low-income groups, special consider- prior to the current year and have generated at least 18 ation will be given to the risks of impoverishment to months of operating revenues (covered by at least two which they may be exposed as a result of resettlement, sets of company annual audited accounts). through: (i) loss of housing, land, access to common 2. IFC’s standard requirements for FIs are: property or other rights to real property, due to lack of (i) process (policy, procedure, responsible staff); clear title, economic pressure or other factors; (ii) loss (ii) appraising and supervising subprojects; (iii) IFC of employment; (iii) loss of access to means of produc- training; (iv) exclusion list; (v) host-country laws; tion; (iv) food insecurity, increased morbidity and (vi) IFC policies; (vii) category-A and/or -B subproj- mortality; (v) disarticulation of social networks; and (vi) ects meet applicable guidelines; (viii) IFC clearance, loss of access to education. A detailed analysis will be public consultation, and EA of category-A subprojects; carried out at the earliest opportunity, covering gender, (ix) annual report (AEPR) on process (EMS, appraisal, ethnicity, income and other socioeconomic factors, in screening); and (x) annual report summarizing order to determine the risks and design preventive subprojects’ performance. Requirements (i)–(vi) are measures to minimize them.” Inter-American Develop- applied to all projects, and some of requirements (vii)– ment Bank, Sector Operational Policy: Involuntary (x) are applied depending on the type of IFC invest- Resettlement, GN-1979–3, July 1998. ment and the portfolio risks of FIs. 133 REFERENCES ADB (Asian Development Bank). 2009. Safeguard ment Working Paper 14958. Washington, DC: Policy Statement; Board Policy Paper. Manila. World Bank. ———. 2006. Environmental Safeguards; Davis, S., and A. Wali. 1993. Indigenous Territo- Special Evaluation Study. Operations Evalua- ries and Tropical Forest Management in tion Department; SST: REG 2006–13. Manila. Latin America. Washington, DC: World Bank. AfDB (African Development Bank). 2004. African Duke, G., and M. Aycrigg. 2000. Biodiversity and Development Bank Group’s Policy on the Environmental Assessment Toolkit. Working Environment. Tunis. Paper 24579. Washington, DC: World Bank. ———. 2000. Evaluation of the Performance EBRD (European Bank for Reconstruction and and Effectiveness of the AFDB’s Environmen- Development). 2008a. Environmental and tal Policy and Environmental Assessment Social Policy and Performance Require- Procedures. ADB/BD/WP/2000/207, Tunis. ments. London: EBRD. Bosshard, P. 2008. China’s Environmental ———. 2008b. 2003 Environmental Policy Review: Footprint in Africa. China in Africa Project; Achieving the Bank’s Mandate through Direct SA Institute of International Affairs. Investments. Special Study PE07–375S, Evalua- ———. 2004. “The World Bank’s Safeguard tion Department. London: EBRD. Policies under Pressure: A Critique of the ———. 2006. Achieving the Bank’s Mandate World Bank’s New Middle Income Country through Financial Intermediaries. Special Strategy”; International Rivers Network, www. Study PE-05–326S, Evaluation Department. internationalrivers.org. London: EBRD. Briscoe, J. 2009. “Reforming the World Bank— Egyptian Environmental Affairs Agency. 2007. Not so Easy.” Business Standard (India), June State of the Environment Report. Cairo. 10, 2009. www.business-standard.com/india. EIB (European Investment Bank). 2008. The EIB Cernea, M. 1997. African Involuntary Population Statement of Environmental and Social Princi- Resettlement in a Global Context. Working ples and Standards. Luxembourg: EIB. Paper 18174. Washington, DC: World Bank. Export-Import Bank of the United States. 2008. Cernea, M., and S. Davis (ed.). 2000. Risks and Environmental Procedures and Guidelines. Reconstruction; Experiences of Resettlers Washington, DC. and Refugees. Washington, DC: World Bank. Goodland, R., and J-R. Mercier. 1999. The Clark, D. 2005. Comments on the Second Draft Evolution of Environmental Assessment in of IFC Policy and Performance Standards. the World Bank from Approval to Results. International Accountability Project, Berkeley. Working Paper 24544. Washington, DC: World www.accountabilityproject.org. Bank. Clark, D., J. Fox, and K. Treakle. 2003. Demand- Hicks, R.,B. Parks, J.T. Roberts, and M. Tierney. ing Accountability: Civil Society Claims and 2008. Greening Aid? Understanding the the World Bank Inspection Panel. Lanham, Environmental Impact of Development Maryland: Rowman and Lillefield. Assistance. Oxford: Oxford University Press. Davis, S., L. Soeftestad. 1995. Participation and Horta, K. 2010. A Trajectory of Evaluation Reports: Indigenous Peoples. Environment Depart- Taking Stock as the Independent Evaluation 135 S A F E G U A R D S A N D S U S TA I N A B I L I T Y P O L I C I E S I N A C H A N G I N G W O R L D Group Completes its Evaluation of World Lawrence, S. 2007. “Nam Theun 2 Trip Report Bank Group Safeguards and Sustainability and Project Update.” International Rivers Policies (1999–2008). Briefing note prepared Network, Berkeley. www.irn.org. by Urgewald e.V Sassenberg. ., ———. 2005. Retreat from the Safeguard IDB (Inter-American Development Bank). 2009. Policies. Recent Trends Undermining Social Sustainability Report. Washington, DC. and Environmental Accountability at ———. 2006. Environment and Safeguards the World Bank. Environmental Defense. Compliance Policy. Washington, DC. www.environmentaldefense.org. IEG (Independent Evaluation Group). 2009a. Liu, Z., and C. Bennett. 2008. Using Involuntary The World Bank Group Program of Support Resettlement on Highway Projects to Allevi- for the Chad-Cameroon Petroleum Develop- ate Poverty. Working Paper 45445. Washing- ment and Pipeline Construction Project. ton, DC: World Bank. Report No 50315. Washington, DC: World Mani, M. 2009. Toolkit to Assess Environmen- Bank. tal Aspects of Development Policy Lending. ———. 2009b. Approach Paper: Evaluation Washington, DC: World Bank. of the World Bank Group’s Experience Mani, M., and A. Sears. 2006. Implementation with Safeguard and Sustainability Policy of Environmental and Natural Aspects in (1999–2008). Washington, DC: World Bank. Development Policy Lending: First Year ———. 2009c. Achieving Sustainable Develop- Review. Paper 36263. Washington, DC: World ment; Annual Review of Development Bank. Effectiveness. Washington, DC: World Bank. MIGA (Multilateral Investment Guarantee ———. 2008a. Doing Business: An Indepen- Agency). 2007. Policy on Social and Environ- dent Evaluation, Taking the Measure of the mental Sustainability. Washington, DC: World World Bank-IFC Doing Business Indicators. Bank. Washington, DC: World Bank. Millennium Challenge Corporation. 2010. ———. 2008b. Environmental Sustainabil- Guidelines for the Use of Country Systems in ity: An Evaluation of World Bank Group the Implementation of Millennium Challenge Support. Washington, DC: World Bank. Compacts www.mcc.gov. ———. 2005. Evaluation of the World Bank Multilateral Financial Institutions Working Group Group’s Activities in Extractive Industries. on the Environment. 2005. A Common Washington, DC: World Bank. Framework for Environmental Assessment ———. 2003a. “Implementation of Operational Good Practice Note, and Rome Declaration on Directive 4.20 on Indigenous Peoples: An Harmonization. www.aidharmonization.org. Evaluation of Results.” Norad (Norwegian Agency for Development ———. 2003b. Review of Implementation of Cooperation). 2007. Assessment of Sustain- Safeguard Policies of World Bank Extractive ability Elements/Key Risk Factors: Practical Industries Projects. IEG Background Paper. Guide. Oslo: Norwegian Ministry of Foreign Washington, DC: World Bank. Affairs. ———. 1998. “Recent Experience with Involun- OECD (Organization for Economic Cooperation tary Resettlement.” and Development). 2006. Applying Strategic IFC (International Finance Corporation). 2006. Environmental Assessment: Good Practice Performance Standards on Social and Environ- Guidance for Development Cooperation. DAC mental Sustainability. Washington, DC: World Guidelines and Reference Series. Paris: OECD. Bank. Park, S. 2007. “Assessing the Accountability of IFC and MIGA. 2004. Beyond Compliance? the World Bank Group.” Paper presented at a An External Review Team Report on the Conference on the World Bank, Bled, Slovenia. Compliance Advisor/Ombudsman Office Picciotto, Robert, Warren van Wicklin, and of IFC and MIGA. Board Report 27600. Edward Rice. (eds.). 2001. Involuntary Washington, DC. Resettlement: Comparative Perspec- 136 REFERENCES tives. World Bank Series on Evaluation and ———. 2008a. Forests Sourcebook: Practical Development, Volume 2. New Brunswick, Guidance for Sustaining Forests in Develop- NJ: Transaction Publishers. ment Cooperation. Paper 44640. Washington, Reseau de Lutte contre la faim. 2009. The impacts DC: World Bank. of the Chad-Cameroon Pipeline. www. ———. 2008b. Use of Country Procurement alternet.org/world/144303. Systems in Bank-Supported Operations: Tanabe, Y. 2009. Recommendations to IFC on Proposed Piloting Program. Board Report Policy and Performance Standards on Social No. 44363, World Bank, Washington, DC. and Environmental Sustainability, and Policy ———. 2007a. Culture and Sustainable on Disclosure of Information. Japan Center Development in the Bank: A Retrospective of for a Sustainable Environment; www.jacses. Our Work, 1996–2007. Working Paper 41191. org/en and Society and Friends of the Earth; World Bank, Washington, DC. www.foejapan.org/en, Tokyo. ———. 2007b. The World Bank Forest Strategy: Third High Level Forum on Aid Effectiveness, Review of Implementation. Washington, DC: 2008. Accra Agenda for Action, Ghana. www. World Bank. accrahlf.net. ———. 2005. Expanding the Use of Country World Bank. 2009a. Accountability at the World Systems in Bank-Supported Operations: Bank: The Inspection Panel at 15 Years. Issues and Proposals. Board Report 31734. Washington, DC: World Bank. Washington, DC. ———. 2009b. Annual Report: Compliance ———. 2004. Involuntary Resettlement Source- Advisor Ombudsman. Washington, DC: book; Planning and Implementation of World Bank. Development Projects. Washington, DC. ———. 2009c. Investment Lending Reform — ———. 2000. The World Bank Forest Strategy: Concept Note. Striking the Right Balance. Washington, DC: ———. 2009d. “Moving Ahead on Invest- World Bank. ment Lending Reform: Risk Framework and ———. 1994. The Bankwide Review of Projects Implementation Support,” Operations Policy Involving Involuntar y Resettlement and Country Services, World Bank, Washing- 1986–1993. Environment Department, April ton, DC. 8, 1994. Washington, DC. ———. 2009e. Environmental and Social Policy World Bank Group and United Nations Environ- and Procedural Guidelines for Projects ment Programme. 1999. Pollution Prevention Financed Jointly by Bank, IFC, and/or MIGA. and Abatement Handbook 1998: Toward Washington, DC. Cleaner Production. Washington, DC. 137 IEG PUBLICATIONS Analyzing the Effects of Policy Reforms on the Poor: An Evaluation of the Effectiveness of World Bank Support to Poverty and Social Impact Analyses Annual Review of Development Effectiveness 2009: Achieving Sustainable Development Addressing the Challenges of Globalization: An Independent Evaluation of the World Bank’s Approach to Global Programs Assessing World Bank Support for Trade, 1987–2004: An IEG Evaluation Climate Change and the World Bank Group—Phase I: An Evaluation of World Bank Win-Win energy Policy Reforms Debt Relief for the Poorest: An Evaluation Update of the HIPC Initiative A Decade of Action in Transport: An Evaluation of World Bank Assistance to the Transport Sector, 1995–2005 The Development Potential of Regional Programs: An Evaluation of World Bank Support of Multicountry Operations Development Results in Middle-Income Countries: An Evaluation of World Bank Support Doing Business: An Independent Evaluation—Taking the Measure of the World Bank–IFC Doing Business Indicators Egypt: Positive Results from Knowledge Sharing and Modest Lending—An IEG Country Assistance Evaluation 1999–2007 Energy Efficiency Finance: Assessing the Impact of IFC’s China Utility-Based Energy Efficiency Finance Program Engaging with Fragile States: An IEG Review of World Bank Support to Low-Income Countries Under Stress Environmental Sustainability: An Evaluation of World Bank Group Support Evaluation of World Bank Assistance to Pacific Member Countries, 1992–2002 Financial Sector Assessment Program: IEG Review of the Joint World Bank and IMF Initiative From Schooling Access to Learning Outcomes: An Unfinished Agenda—An Evaluation of World Bank Support to Primary Education Gender and Development: An Evaluation of World Bank Support, 2002–08 Hazards of Nature, Risks to Development: An IEG Evaluation of World Bank Assistance for Natural Disasters How to Build M&E Systems to Support Better Government IEG Review of World Bank Assistance for Financial Sector Reform An Impact Evaluation of India’s Second and Third Andhra Pradesh Irrigation Projects: A Case of Poverty Reduction with Low Economic Returns Improving Effectiveness and Outcomes for the Poor in Health, Nutrition, and Population Improving the Lives of the Poor through Investment in Cities Improving Municipal Management for Cities to Succeed: An IEG Special Study Improving the World Bank’s Development Assistance: What Does Evaluation Show: Maintaining Momentum to 2015: An Impact Evaluation of Interventions to Improve Maternal and Child Health and Nutrition Outcomes in Bangladesh New Renewable Energy: A Review of the World Bank’s Assistance Pakistan: An Evaluation of the World Bank’s Assistance Pension Reform and the Development of Pension Systems: An Evaluation of World Bank Assistance The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank’s Support Through 2003 The Poverty Reduction Strategy Initiative: Findings from 10 Country Case Studies of World Bank and IMF Support Poverty Reduction Support Credits: An Evaluation of World Bank Support Public Sector Reform: What Works and Why? An IEG Evaluation of World Bank Support Small States: Making the Most of Development Assistance—A Synthesis of World Bank Findings Sourcebook for Evaluating Global and Regional Partnership Programs Using Knowledge to Improve Development Effectiveness: An Evaluation of World Bank Economic and Sector Work and Technical Assistance, 2000–2006 Using Training to Build Capacity for Development: An Evaluation of the World Bank’s Project-Based and WBI Training Water and Development: An Evaluation of World Bank Support, 1997–2007 The Welfare Impact of Rural Electrification: A Reassessment of the Costs and Benefits—An IEG Impact Evaluation World Bank Assistance to Agriculture in Sub-Saharan Africa: An IEG Review World Bank Assistance to the Financial Sector: A Synthesis of IEG Evaluations World Bank Group Guarantee Instruments 1990–2007: An Independent Evaluation World Bank Engagement at the State Level: The Cases of Brazil, India, Nigeria, and Russia The World Bank’s Country Policy and Institutional Assessment: An Evaluation All IEG evaluations are available at http://www.worldbank.org/ieg. ISBN 978-1-60244-159-0