Report No. 91854-ET Ethiopia: Promoting Basic Services Phase III Project (P128891) Investigation Report Nov 21, 2014 The Inspection Panel Acknowledgements The preparation of this Report would not have been possible without the support and valuable contributions of many people. The Panel wishes to thank the Anuak Requesters who met with the Panel team in South Sudan and Kenya; the Anuak communities the Panel met with in Gambella and Addis Ababa, Ethiopia, and the Requesters’ representatives. The Panel expresses its gratitude to the office of the Executive Director representing Ethiopia in the Board of Executive Directors of the World Bank for their invaluable guidance and support during the investigation. The Panel also thanks national government officials in Ethiopia with whom the Panel met during the eligibility visit and regional government officials in Gambella, who provided valuable information to the Panel team, and to all those that agreed to meet with the Panel in Ethiopia and provided information and comments. The Panel expresses its appreciation Bank Management and Staff for their assistance in obtaining documents, providing the Panel with information, and responding promptly to written requests. The Panel also wishes to thank the World Bank Staff in Washington D.C. and in the Ethiopia office for assisting with logistical arrangements during the Panel team’s visits to the Project area. The Panel would also like to thank representatives of the PBS Development Partners for meeting with the Panel team during its visits and sharing their insights. The Panel is grateful for the expert advice provided by Dr. Alison Evans, Prof. Eisei Kurimoto, Prof. Stephen Peterson, and Mr. Roberto Mosse and appreciates the professionalism they exhibited at all times. The Panel wishes to express its deepest gratitude Eimi Watanabe, former Panel Chairperson, who led ably this investigation and provided expert and invaluable guidance to the Panel team. Finally, the Panel wishes to convey its gratitude and appreciation to the members of its Secretariat for their professional and resourceful handling of this investigation.     Abbreviations and Acronyms AF Additional Financing AfDB African Development Bank ARRA Administration for Refugee and Returnee Affairs BoFED Bureau of Finance and Economic Development BG Benishangul-Gumuz BP Bank Procedure CDP Commune Development Programme CPS Country Partnership Strategy CSO Civil Society Organization DA Development Agent DAG Development Assistance Group DFID Department for International Development (UK) DP Development Partner EIO Ethiopian Institution of the Ombudsman EPRDF Ethiopian People’s Revolutionary Democratic Front EU European Union FRFT Federal-to-region fiscal transfer FY Fiscal Year GoE Government of Ethiopia GPDP Gambella People’s Democratic Party GPLM Gambella Peoples’ Liberation Movement GRM Grievance Redress Mechanism GTP Growth and Transformation Plan HRW Human Rights Watch ICR Implementation Completion and Results Report IDA International Development Association IDI Inclusive Development International IEG Independent Evaluation Group IGFT Intergovernmental fiscal transfer IL Investment Lending ISDS Integrated Safeguard Data Sheet JBAR Joint Budget and Aid Review JRIS Joint Review and Implementation Support KfW Kreditanstalt fuer Wiederaufbau (Germany) LIG Local Investment Grant M4R Managing for Results MDG Millennium Development Goal MOFA Ministry of Federal Affairs MoFED Ministry of Finance and Economic Development MOP Memorandum of the President NGO Nongovernmental Organization OMS Operational Manual Statement ii OP Operational Policy ORAF Operational Risk Assessment Framework P4R Program for Results PAD Project Appraisal Document PBS Promoting Basic Services PDO Project Development Objective PFM Public Financial Management PID Project Information Document PSIA Poverty and Social Impact Assessment PSNP Productive Safety Net Project ROC Regional Operations Committee RPF Resettlement Policy Framework RWFT Region-to-woreda Fiscal Transfer SA Social Accountability SIL Specific Investment Loan SNNPR Southern Nations, Nationalities, and People’s Region SPLA Sudan Peoples’ Liberation Army TOR Terms of Reference TS Treasury System UN United Nations UNHCR United Nations High Commissioner for Refugees UNICEF United Nations Children's Fund VECOD Vision Ethiopia for Congress of Democracy VP Villagization Program WoFED Woreda-level Office of Finance and Economic Development       iii Table of Contents Acknowledgements .......................................................................................................................... i  Abbreviations and Acronyms ......................................................................................................... ii  Executive Summary ..................................................................................................................... viii  Chapter 1: Introduction and Background ........................................................................................ 1  1.1. Outline and Organization of the Report ............................................................................... 1  1.2. Description of the PBS Program and the PBS III Project .................................................... 2  1.3. Issues Raised in the Request for Inspection ......................................................................... 5  1.4. Management Response ......................................................................................................... 5  1.5. Project Context ..................................................................................................................... 7  1.5.1. The Gambella Region .................................................................................................... 7  1.5.2. The Anuak People ......................................................................................................... 8  1.5.3. Conflicts Involving the Anuak .................................................................................... 10  1.5.4. Recent Developments .................................................................................................. 11  1.5.5. Description of the CDP................................................................................................ 12  1.6. The Panel’s Investigation Process ...................................................................................... 14  Chapter 2: PBS III and CDP - Operational Interface and Risk..................................................... 16  2.1. The Operational Link between PBS III and CDP, and Allegations of Harm..................... 16  2.1.1. Requesters’ Claims and Management Response ......................................................... 16  a)  Requesters’ Claims ......................................................................................................... 16  b)  Management Response ................................................................................................... 17  2.1.2. Assessment of the Link ............................................................................................... 17  2.1.3. The Four Sets of Allegations of Harm......................................................................... 18  a)  Involuntary Taking of Land and Use of Force and Intimidation ................................... 19  b)  Harms Related to Lack of Services at Relocation Sites and Deterioration of Livelihoods 21  Agriculture ............................................................................................................................. 21  Education ............................................................................................................................... 24  Health..................................................................................................................................... 25  Water Supply and Sanitation ................................................................................................. 26  2.2. Risk Assessment Analysis in PBS III Appraisal and Implementation ............................... 27  iv 2.2.1. Requesters’ Claims and Management Response ......................................................... 27  a)  Requesters’ Claims ......................................................................................................... 27  b)  Management Response ................................................................................................... 27  2.2.2.  Applicable Bank Policies ........................................................................................ 28  2.2.3. PBS III Risk Assessment ............................................................................................. 30  a)  During Project Design and Appraisal............................................................................. 30  b)  During Project Implementation ...................................................................................... 33  2.2.4.  Panel’s Analysis of Risk Assessment and Compliance .......................................... 34  a)  Adequacy of the Risk Analysis undertaken during PBS III Preparation ....................... 34  b)  Adequacy of the Risk Analysis undertaken during PBS III Implementation ................ 37  Chapter 3: Application of Safeguard Policies and Distribution of Services in Gambella ............ 41  3.1. Application of the Safeguard Polices to PBS III ................................................................ 41  3.1.1. Application of Safeguards to Projects Focused on Recurrent Expenditures ............... 41  3.1.2. The Involuntary Resettlement Policy .......................................................................... 42  3.1.3. The Indigenous Peoples Policy.................................................................................... 44  a)  Requester’s Claims ......................................................................................................... 44  b)  Management Response ................................................................................................... 44  c)  Indigenous Peoples in Ethiopia’s Constitution .............................................................. 44  d)  Background and Chronology relating to the Application of OP 4.10 in Ethiopia ......... 45  e)  The Concept of Functional Equivalence ........................................................................ 46  f)  The Panel’s Analysis ...................................................................................................... 47  3.2. Delivery of Services and Livelihoods ................................................................................ 49  3.2.1. Summary of the Requesters’ Claims and Management Response .............................. 49  a)  Requesters’ Claims ......................................................................................................... 49  b)  Management Response ................................................................................................... 49  3.2.2. The Panel’s Analysis on Agriculture and Livelihoods ................................................ 50  3.2.3. Project Appraisal ......................................................................................................... 52  3.2.4. Monitoring during PBS III Implementation ................................................................ 54  Chapter 4 ....................................................................................................................................... 56  Financial and Human Resources Link Analysis ........................................................................... 56  4.1. Financial Risk Assessment and Management .................................................................... 56  v 4.1.1. Requesters’ Claims and Management Response ......................................................... 56  a)  Requesters’ Claims ......................................................................................................... 56  b)  Management Response ................................................................................................... 57  4.1.2. Panel’s Analysis .......................................................................................................... 58  c)  Diversion of Financial Resources .................................................................................. 59  Potential for Diversion........................................................................................................... 60  Evidence of Diversion ........................................................................................................... 63  Detection of Diversion........................................................................................................... 63  d)  Fiduciary Risk Management .......................................................................................... 64  e)  Compliance with Bank Operational Policies and Procedures ........................................ 65  Relevant Bank Policies .......................................................................................................... 65  Project Appraisal ................................................................................................................... 66  Supervision ............................................................................................................................ 68  4.2. Human Resources Link Analysis ....................................................................................... 70  4.2.1. Requesters’ Claims and Management Response ......................................................... 70  a)  Requesters’ Claims ......................................................................................................... 70  b)  Management Response ................................................................................................... 70  4.2.2. Panel’s Analysis .......................................................................................................... 70  Chapter 5 ....................................................................................................................................... 72  Conclusions ................................................................................................................................... 72  Annex A: Table of Findings ........................................................................................................ 78  Annex B: About the Panel ........................................................................................................... 80  Annex C: Biographies .................................................................................................................. 81    Figure 1 Timeline of Events ......................................................................................................... 30 Figure 2 Gambella - Budget Performance (in million birr) .......................................................... 61   Box 1: Panel Interviews in the Field - Land and Agriculture, Food And Housing ..................................... 22  Box 2: Panel Interviews in the Field - Education .......................................................................................  24  Box 3: Panel Interviews in the Field - Health .............................................................................................  25  Box 4: Panel Interviews in the Field – Water Supply and Sanitation ......................................................... 26  Box 5: Financial Management in Gambella (March 2013) ........................................................................  63  vi vii Executive Summary 1. A Request for Inspection of the Ethiopia - Promoting Basic Services (PBS) Phase III Project was submitted on September 24, 2012 by two representatives on behalf of 26 Anuak people from the Gambella region of Ethiopia. The Requesters currently live in refugee camps in South Sudan and Kenya, and asked for their identities to be kept confidential. The PBS Program and the PBS III Project 2. The PBS Program is a nationwide multi-donor program contributing to expanding access and improving quality of basic services in five sectors (education, health, agriculture, water supply and sanitation, and rural roads). It consists of three phases, the first of which (the PBS I Project) started in 2006. The PBS III Project, which is the focus of this investigation, was approved in September 2012 for USD 600 million. PBS III contributes to the higher-level objective of expanding access and improving the quality of basic services by funding block grants that ensure adequate staffing and operations, and by strengthening the capacity, transparency, accountability and financial management of government at regional and local authorities’ levels. PBS III has three components: a) Basic Service Block Grants; b) Strengthening Local Accountability and Transparency Systems (Citizen’s Engagement, Local Public Financial Management and Procurement, and Managing for Results), which includes a Social Accountability Program and strengthening or establishing grievance redress mechanisms, and c) Results Enhancement Fund. PBS III serves approximately 84 million people in Ethiopia. Key Claims and Management Response 3. The Requesters state that they have been severely harmed by the PBS Program as a result of the Bank's non-compliance with its policies and procedures. They argue that the PBS Program is contributing directly to the Ethiopian Government's Villagization Program (officially known as the Commune Development Program or CDP1) in Gambella, that the Anuak indigenous people are being forced to leave their ancestral land under the CDP Program, and that the land is then leased to investors. The Request states that this is carried out under the pretext of providing better services and improving the livelihoods of communities. The Requesters further claim that the new relocation sites provided infertile land, no schools, clinics or other basic services. According to the Requesters, the safeguard policies on Indigenous Peoples and Involuntary Resettlement should have been applied. The Requesters also argue that PBS funds were diverted to finance the implementation of the CDP, and that government workers paid under the PBS Program were implementing the CDP. 4. Bank Management states that PBS III2 neither finances the CDP, nor depends on it to achieve its objectives; therefore, in the management’s view, the harm described in the Request does not arise from PBS III or a Bank failure in applying its policies. Management argues that there is no scope for the application of safeguards policies to CDP. Management explains that the                                                              1 This Report uses the terms Villagization, “Commune Development Program,” and CDP interchangeably. 2 Management focuses its response on those aspects of the Request that relate to PBS III, explaining that Bank funding for PBS I was closed and PBS II more than 95% disbursed at the time the Request was submitted. viii Government of Ethiopia (GoE) did not consult or advise the Bank or the other Development Partners (DPs) about the CDP, and that the Bank became aware of the CDP during an unrelated field visit in October 2010. Management explains that the Bank and other DPs have tried to achieve influence through policy dialogue with the GoE. Regarding the alleged human rights abuses, Management Response refers to the missions of DPs, which did not find evidence of forced relocations or systematic human rights abuses in the implementation of the CDP. Management also explains that allegations of misuse of funds have been raised since 2005, but were of a general nature and did not provide a basis for revising PBS implementation arrangements. Furthermore, local government officials often play multiple roles and it would not be feasible to limit their work to functions supported by PBS. Management argues that the PBS III arrangements enable close monitoring of the Project and tracing of resources. Project Context 5. Gambella, one of nine regions in Ethiopia, is located in the West of the country bordering South Sudan. According to the 2007 national census, Gambella’s population included 307,000 people of the following ethnicities: Nuer (46%), Anuak (21%), highlanders (20%), Majangir (7%), Komo (3%), and Opuo (3%). 6. The Anuak belong to the Nilotic group, and live in the lowlands of Gambella and in the Pochalla and Akobo counties of South Sudan. They depend primarily on agriculture, fishing, hunting and gathering. The Anuak have a strong sense of territoriality: even though their territory is sparsely populated, there is no “empty” or “unused” land, as every inch of land is identified with distinct communities. Land has been significant to the Anuak’s security for a long time, as they have used the forest to hide when threatened. The Anuak have undergone changes in their lifestyle and culture over the past decades, and many of their traditional institutions and customs have changed. 7. The Anuak have experienced a history of conflict. These conflicts also expand beyond the country’s borders: political instability in South Sudan and a large influx of refugees contribute to a difficult environment for Gambella’s ethnic groups. Questions of security impact the livelihoods and vulnerability of Gambella’s women and children. Another area of contention which has received increasing attention in the past years is large-scale agricultural investment in Gambella. 8. According to Bank Management, the GoE began implementing the CDP in 2010 in four historically underserved regions, among them Gambella. CDP is a three-year program with the objective of facilitating more effective delivery of services through the voluntary congregation of dispersed populations around village centers. In the Gambella region, the regional government’s strategy for improving access to basic services is described in the Villagization Program Action Plan of 2010, and planned for the resettlement of 45,000 households, which is close to 70% of all households in Gambella. According to information by the DPs, the regional government reported that the program formally ended in Gambella in mid-2013, after resettling a total of 37,883 or approximately 60% of households there. ix PBS III and CDP: Operational Interface and Risks 9. In its Report and Recommendation submitted to the Board on February 8, 2013, the Panel indicated that this investigation would have a specific focus by examining the risks of the concurrent implementation of the PBS and CDP programs in Gambella, Management’s analysis of these risks, and subsequent Management actions. The Panel also stated that the investigation would not seek to verify allegations of specific human rights abuses linked to CDP, nor the underlying purposes of CDP as these are not within its mandate.3  10. Assessment of the Link between PBS III and CDP. The Panel notes that both PBS and CDP have the ultimate objective of providing improved basic services to the population. There was a chronological and geographic overlap in the two programs’ implementation, as PBS and CDP were implemented concurrently in Gambella from 2010 to 2013, covering the same woredas.4 CDP is a program that aims at restructuring settlement patterns, service infrastructure and livelihoods, including farming systems, and as such constitutes a significant part of the context in which PBS operates. In this sense, and from a development perspective, the two programs may influence each other’s results. The Panel is of the view, therefore, that there is an operational link (interface) between the CDP and PBS in the woredas where there is concurrent implementation. 11. Allegations of Harm. The Request raised issues of harms that covered four broad aspects: (i) taking of people's customary land without their free and informed consent, (ii) use of force and intimidation to get people to relocate, (iii) lack of public services and assistance at relocation sites, and (iv) deterioration of livelihoods due to less fertile land and lack of other income generating opportunities at relocation sites. 12. As mentioned earlier, the investigation report does not include findings on the first two issues of harm. Nevertheless, in the course of the Panel’s review of documents and interviews in the field, the Panel came across information regarding those allegations. Given that these issues were raised in the Request for Inspection, the Panel recorded this information in this Report but without attempting to verify it or otherwise. 13. The Panel’s investigation and analysis instead focused on the third and fourth claims with operational links to PBS: lack of basic services in new villages, and resulting deterioration of people’s livelihoods. The Report presents an account of information aligned with the PBS sectors gathered by the Panel during the field visit and provided in DP’s reports following visits to Gambella and other regions where CDP was implemented. Overall, the Panel observed a mixed picture in the status of basic services, with some deterioration (e.g. agriculture) and some improvement (e.g. education), but in general noted that the level of basic services in the villages visited remained inadequate. 14. Risk Assessment during Project Appraisal. Following a number of missions to the regional states where CDP was being implemented, DPs indicated that they found no evidence of                                                              3 Report and Recommendation: 21. 4 Woredas constitute the third level (after regions and zones) in the country’s decentralized administrative structure. Each woreda is composed of kebeles (villages), the smallest administrative units in Ethiopia. x systematic human rights abuses connected with the villagization process, and the World Bank determined that the CDP program was not linked to the PBS block grants. The Bank therefore considered more appropriate to address the issues posed by the implementation of the CDP at the country and portfolio levels. At the World Bank Regional Operations Committee (ROC) meeting to review the proposed PBS III Project documents prior to Board approval, it was acknowledged that there was an overlap between the PBS objectives of promoting improved access to basic services and the Government’s objectives for its CDP. A re-evaluation of the risks was thus recommended with a focus on managing reputational risks for the institution. 15. Considering the magnitude of the operation, the nature of block grant financing, and the overlapping implementation between PBS III and CDP, the Panel finds that Management did not carry out the required full risk analysis, nor were its mitigation measures adequate to manage the concurrent roll-out of the villagization program in four PBS III regions. The Panel finds that Management’s approach did not meet the standards of a systematic or holistic assessment of risks, as called for in the Operational Risk Assessment Framework (ORAF) Guidance, which is aimed among other objectives at identifying adequate risk management measures for affected communities. The Panel finds these omissions in non- compliance with OMS 2.20 on Project Appraisal. 16. The Panel notes that a social assessment would have been justified both to verify the robustness of the results chain that links the block grant mechanism to PBS results, and to assess any changes in the local governance context, which might put future results achievement at risk. Such an exercise might have identified a series of risks to watch during implementation, or an alternative set of risk management measures designed to avoid negative impacts from CDP. 17. Risk Assessment during Project Implementation. The issues raised by the Requesters were prominently discussed during the Board meeting for the approval of PBS III in September 25, 2012. In response to concerns raised by Executive Directors, Management committed to undertake a Poverty and Social Impact Assessment (PSIA) of PBS III to further verify that increasing funds, transparency, and accountability through citizens’ engagement provide proportionately greater returns to marginal areas and vulnerable groups. The Panel appreciates that the PSIA analysis on resources allocation has shown that historically disadvantaged areas, including four Anuak woredas, are favored under the PBS spending framework. The Panel notes, however, that while higher allocation of resources is positive in itself, it does not automatically translate into better quality or coverage of services, nor does it address the issue of interface between PBS and CDP and their mutual impacts. 18. The Panel recognizes Management’s efforts to address social accountability during PBS III implementation and appreciates that the establishment of the Grievance Redress Mechanism (GRM) and the implementation of the social accountability component were fast-tracked through the opening of the Ombudsman office in Gambella. The Panel understands and appreciates that more recently, Management has recognized the issue of the operational interface, and is currently supporting a study on the interface between World Bank-funded and nationally-funded programs to ensure coherence in their implementation. xi 19. Notwithstanding these facts, it is the view of the Panel that the lack of recognition and analysis during appraisal of the operational interface between PBS III and CDP as required by the ORAF and described above, meant that the resulting risks were not adequately taken into account, neither were they properly managed and mitigated during PBS III implementation. Application of Safeguards Policies 20. Application of Safeguards to Recurrent Expenditures. The World Bank operational policies and procedures do not exclude the application of safeguard policies in Investment Lending (IL) operations that provide support for recurrent expenditures. The Panel finds that to the extent that one or more safeguard policies are found to be relevant to the areas of operation of the proposed IL operation, Management should trigger and apply them. 21. Application of the Policy on Involuntary Resettlement (OP 4.12). The Panel notes that the nationwide PBS was initiated in 2006 and is expected to close in 2018. Meanwhile, the CDP began in Gambella in 2010 and terminated in 2013. Therefore, while CDP in Gambella was contemporaneous with PBS, the latter was being implemented four years prior to the commencement of CDP, and is continuing after the conclusion of CDP. Furthermore, PBS was designed to cover the entire population of Gambella, irrespective of whether they relocated under CDP. Thus, the Panel does not consider CDP necessary to achieve the objectives of PBS III. The Panel finds that paragraph 4 of OP 4.12 is not applicable, and that Management acted consistently with the provisions of OP/BP 4.12 by not triggering it. 22. Application of the Policy on Indigenous Peoples (OP 4.10). Management indicated to the Panel that OP 4.10 was not applied to Bank operations in Ethiopia before 2013, including the PBS III project, because of the GoE’s concern that the policy was not compatible with the Ethiopian Constitution and the country context. The Panel acknowledges the difficult context in which PBS III was prepared and the concerns expressed by the GoE regarding the definition of Indigenous Peoples and the application of the related policy in the Ethiopian context. The Panel also notes Management’s indication that prior to December 2012, Bank operations, including PBS III, followed the approach known as “functional equivalence” with the policy. 23. The Panel, however, found no evidence in Project documents that the functional equivalence approach was applied to PBS III. In contrast, the PAD clearly noted that safeguard policies are not triggered as project financing is limited to recurrent expenditures for basic services. The Panel finds that, barring the triggering of OP 4.10, Management should have adopted the “functional equivalence” approach in the design of PBS III. The relevance and need of adopting such approach to the Anuak stem from the distinct characteristic of these groups that the Bank Policy OP 4.10 well lays out. The Panel notes that livelihoods, well-being and access to basic services, which are closely tied to the Anuak’s access to land and natural resources was not taken into account in the design of PBS III, in non-compliance with OP 4.10. xii Agriculture and Livelihoods 24. PBS supports the agriculture sector by paying the salary of trained development agents that provide agricultural extension services for crops, livestock and natural resource management. Progress towards attaining the PBS Project Development Objectives in the agriculture sector is measured by the number of agriculture sector beneficiaries (direct beneficiaries); the intermediate results indicator is the number of specialized Development Agents (DAs) in villages; while the higher level objective of PBS III is increased agricultural productivity. 25. Availability of land and land-related issues (agricultural production and food) are of the greatest concern to people affected by CDP. The Panel observed a general trend in many villages where land was provided to resettled people, but they could not work it because the land had not been cleared and little or no support was provided for clearing it. As DPs also noted, livelihood options were limited and in some cases, lack of access to fishing and riverside mango trees worsened the situation. The effectiveness and quality of services in the agriculture sector is closely connected and influences the quality of livelihoods of villagers largely dependent on agriculture. Access to appropriate agricultural services is particularly important for indigenous groups such as the Anuak, whose livelihood is closely tied to the land. 26. The Panel concludes that access to services in the agriculture sector in Gambella, along with the possibility of achieving the higher level objectives of the Project (increasing agricultural productivity), was adversely affected by the CDP as it was implemented. The availability of workable land at the new resettlement sites is an aspect where CDP had an impact on the delivery of results under PBS. While the PBS results, indicated by the number of DAs and the number of direct beneficiaries were achieved, little can be said about the effectiveness and quality of the service delivered in the agriculture sector, let alone the high level objective of agricultural productivity. As noted by villagers that the Panel met, access to the services provided by DAs had minimal effect since beneficiaries could not make efficient use of such services for lack of workable land nearby. 27. The Panel finds that, in accordance with Bank Policies, the operational interface between CDP and PBS should have been taken into account at the PBS project level, both during the appraisal and implementation phases, especially in a region such as Gambella where 60% of households, which are also PBS beneficiaries, were resettled as part of the Government’s CDP. The Panel finds that Management’s approach has not enabled PBS to mitigate or manage the harms described in the Request for Inspection with respect to access and quality of basic services in the agricultural sector and livelihoods of affected people in Gambella. 28. The Panel finds that, while monitoring was expected to follow the principle of effectiveness and results, the quality of services in Gambella was not in alignment with actual conditions on the ground. In particular, there is no indication that the agricultural extension services was monitored and aligned to deliver appropriate results. It is the Panel’s view that this lacuna in monitoring and reporting about impacts on the ground of the CDP, and how these impacts affected the quality and effectiveness of the services delivered under PBS, resulted from the failure to conduct a full risk analysis at Project appraisal and from a “thin” results framework for PBS, which “put little xiii emphasis on the quality of or impact of basic service delivery,”5 as noted in an Independent Evaluation Group (IEG) evaluation of PBS I. Following a review of the results indicators for PBS I, II and III, it is the Panel’s view that such conclusion on the results framework is relevant to PBS III as well, especially with respect to indicators related to the agriculture sector. 29. Since PBS III began implementation, three “Joint Review and Implementation Support” (JRIS) missions were undertaken, but the resulting reports are silent on the issues noted above. The Panel finds that this is not consistent with the supervision provisions of the Investment Lending Policy (OP/BP 10.00). Financial Risk Assessment and Management 30. The Inspection Panel reviewed the claim of diversion of PBS funds to CDP activities at three levels: potential, detection and evidence. The Panel notes that the un-earmarked and discretionary nature of block grants raises the potential for diversion, because the transfers from the federal level to regions, and from regions to woredas (FRFT and RWFT) are not specified, and there is discretion as to their use. In addition, the potential for diversion of funds from PBS depends on the quality of the financial systems of the GoE. The Treasury System (TS) has been rated as high risk for PBS III. According to the PAD, there are weak internal audit controls at the federal and regional levels and continued inadequate follow up on audit reports. The Panel finds that the weakness of internal controls supports the possibility that funds could have been diverted (“potential”). It also means that diversion cannot be verified, and that the World Bank’s assertion that it can fully track expenditures cannot be supported. 31. The Panel did not have access to Government financial records to find evidence of diversion. Therefore, the Panel cannot present findings on this issue. In terms of detection of possible diversion, the Panel notes that the use of the TS, which is weak, compromises the Bank’s ability to detect diversion. One key means of detection that the Bank uses is timely and good- quality external audit reports; meanwhile for Gambella, the 2013 supervision mission found a five- year backlog in audits and an absence of internal audit. It is the Panel’s view that the Bank’s principal means of detecting diversion in Gambella lacks timeliness and quality. The Panel finds that Management did not comply with the requirements of OMS 2.20 and OP/BP 10.02 in the design and appraisal of PBS III. The Panel notes that the Bank’s assertion that the funds can be tracked at the woreda level cannot be sustained. 32. Monitoring of Fiduciary Risks. The March 2013 PBS III supervision mission found serious deficiencies in the Gambella region’s performance in financial management and on the Public Financial Management (PFM) Project Development Objectives (PDO) The starting point for both appraisal and monitoring of fiduciary risk is the adequacy of PDOs. The Panel finds that, since PDO results indicators that directly address fiduciary risks were inadequate in the initial planning, and subsequently have not been adjusted, the supervision of those risks is not in compliance with Bank policy OP/BP 10.00. Given the Panel's findings on the situation in Gambella, going forward, it is hoped that particular attention may be paid to the woredas in                                                              5 Independent Evaluation Group (IEG) (2013): Project Performance Assessment Report. Ethiopia. Protection of Basic Services Project (IDA H2240-ET, IDA H3470-ET) (Report No.: 82528): x. xiv developing regional states such as Gambella, to strengthen their capacity to ensure adequate implementation and monitoring of Project’s financial management arrangements. Human Resources Link Assessment 33. The information gathered by the Panel is not sufficient to determine whether woreda workers financed by PBS were forced to participate in the implementation of the CDP against their will; if so, what kind of activities they carried out in this respect; and whether their salaries were decreased to finance activities under the CDP. Importantly, the Panel was also not able to find in JRIS or other reports any information from Gambella or from the other regions where CDP was being implemented, regarding the extent to which workers financed under PBS were undertaking additional or alternative work in relation to CDP, which may have impacted the achievement of PBS PDOs. The Panel notes that it does not have sufficient information to draw a firm conclusion on this issue, but again notes how the operational interface between the CDP and PBS, and mutual impacts affecting the achievement of results under PBS, were not adequately considered at appraisal or during implementation. Conclusions 34. The Panel acknowledges the important contribution that the PBS projects (all three phases), together with other related basic services projects, have made towards supporting Ethiopia’s achievement of the MDGs, as widely acknowledged by the international community. It is also considered a positive example of development partnership, collaboration and support by using an innovative mechanism of block grants which are in line with the “Paris – Accra – Busan” development effectiveness principles based on mutual accountability. 35. While fully appreciating the context of this Project, the Investigation has highlighted various ramifications of applying the block grant mechanism in terms of its limitations for monitoring the flow of resources, both financial and human, and for monitoring and assessing results to ensure that PBS is on the right course, or to take corrective actions as soon as possible when results are found to be “off track.” 36. It is the Panel’s view that when reports of villagization emerged, Management responded together with other DPs in the context of the DAG by recognizing the serious implications of CDP and pursuing high level policy dialogue with the GoE. The Panel appreciates the proactive leadership role that Management played in these delicate policy dialogues and in a difficult context, including through its leadership in the DAG. In the view of the Panel, however, Management distanced itself from recognizing the operational interface between CDP and PBS and as a result, from recognizing the impact that CDP was having and will continue to have on PBS results. The impact of CDP was not taken into account during the appraisal or supervision of PBS III. 37. The Panel understands that more recently, Management is currently supporting a study on the interface between WBG-funded and nationally-funded programs to ensure coherence in their implementation. The Panel welcomes this development. xv 38. As noted by IEG, PBS presents key features of the Bank’s Program for Results (P4R) approach, linking disbursements to defined results with a special focus on strengthening institutions. The Panel does not question the application of the “quasi-P4R” PBS modality, which is in line with contemporary principles of aid effectiveness. Fundamental to the effective delivery of results and successful outcomes, however, the system of delivery needs to be robust and meet accepted environmental and social safeguards principles. This investigation has highlighted key challenges for effectively implementing a system in the context of sub-national levels in Gambella that requires capacity strengthening, and in an operating environment where an alternative program is having a significant impact on PBS operations. xvi Chapter 1: Introduction and Background 1.1. Outline and Organization of the Report 39. This report presents the findings and analysis of the Request for Inspection of the Ethiopia - Promoting Basic Services Phase III Project (P128891), hereinafter the “Project” or “PBS III,” received by the Inspection Panel on September 24, 2012.6 40. The Request was submitted by two representatives on behalf of 26 Anuak people from the Gambella Region of Ethiopia (hereinafter the “Requesters”). Both the representatives and the Requesters have asked the Panel to treat their identities confidentially “due to grave concerns about our personal security and that of our relatives in Ethiopia.”7 The Requesters no longer reside in Ethiopia, but live in refugee camps in South Sudan and Kenya. 41. In line with its mandate, the Panel investigation focuses solely on allegations of harm that may result from non-compliance by the Bank with its operational policies and procedures. The report includes five Chapters:  Chapter 1 (this Chapter) describes the Bank-supported PBS Program and the PBS III Project, summarizes the claims made by the Requesters and the Bank’s Management Response, provides the Project context by describing the Gambella region and the specific historical and current situation of the Anuak people, and describes the Ethiopian government’s Commune Development Program (hereinafter “CDP” or “Villagization”)8. The Chapter concludes with a description of the Panel’s investigation process and framework.  Chapter 2 addresses the potential links between PBS and CDP, and describes the four sets of harms alleged in the Request. The Chapter records information on claims related to forcible evictions but without an accompanying assessment. It also includes the Panel’s analysis of the Bank’s risk assessment during appraisal and implementation.  Chapter 3 includes the Panel’s findings on the application of safeguard policies to PBS III, including the policy on Involuntary Resettlement, the specificities of the Indigenous Peoples policy in the context of this program, and the Panel’s analysis of the allegations of harm on access to basic services in the villages and livelihood conditions.  Chapter 4 includes the analysis of possible financial and human resource linkages between PBS and CDP, and the Panel’s compliance analysis on these issues                                                              6 The Request also refers to Protection of Basic Services Program Phase II - Additional Financing (P121727). However, at the time the Request for Inspection was received, PBS II-AF was around 97% disbursed and thus did not qualify for the Panel’s investigation, which therefore focused on PBS III. The Request for Inspection is available at the Panel’s website at http://ewebapps.worldbank.org/apps/ip/Pages/ViewCase.aspx?CaseId=88 (accessed November 10, 2014) 7 Request for Inspection: 1. 8 The terms “CDP” and “Villagization” refer to the same program and are used interchangeably in this report. For detailed information about the Program, see Chapter 1.5.5.  Chapter 5 presents the Report’s main conclusions. 1.2. Description of the PBS Program and the PBS III Project 42. PBS is a nationwide program contributing to expanding access and improving quality of basic services defined as education, health, agriculture, water supply and sanitation, and rural roads9. It consists of three continuous operations and two Additional Financing (AF) components: PBS I, PBS I-AF, PBS II, PBS II-AF and PBS III10. PBS is a multi-donor program, involving the Government of Ethiopia (GoE), the World Bank, the African Development Bank (AfDB), the UK Department for International Development (DFID), the European Union (EU), and the Governments of Austria and Italy. The social accountability activities are funded by DFID, KfW, Irish Aid and the EU.11 43. Basic Services and the MDGs. According to the Project Appraisal Document (PAD) for PBS III, “Ethiopia has achieved impressive development results in recent years. That progress is evidenced through rapid and significant improvements in basic service delivery indicators.”12 The government’s poverty reduction strategy paper, the Growth and Transformation Plan13 (2010/11- 2014/15), envisages the achievement of all Millennium Development Goals (MDGs), and the government recognizes the provision of greater access to quality decentralized services as a key to achieving the MDGs. According to Bank Management, the GoE is committed to increasing resources to woreda14 authorities for the improvement of those services.15 A 2010 study by the Overseas Development Institute which measures progress across countries in achieving the MDGs shows that Ethiopia ranks third globally in “absolute progress” (i.e., the largest positive change regardless of initial conditions16). The 2013 MDG Report lists Ethiopia among six African countries which are making especially impressive progress towards reaching the MDGs.17 According to Management’s Poverty and Social Impact Assessment (PSIA) of PBS III, the Bank- supported PBS Program is the primary means by which the Bank assists Ethiopia in the achievement of the MDGs.18                                                              9 PBS III Project Appraisal Document (PAD): 10. Available online: http://documents.worldbank.org/curated/en/2012/08/16711203/ethiopia-third-phase-promoting-basic-services- project (accessed November 12, 2014) 10 This report pertains only to PBS III. 11 In this Report, the PBS financiers are also referred to as PBS Development Partners (DPs). 12 PBS III PAD: 1. 13 Ethiopian Ministry of Finance and Economic Development (2010): The Federal Democratic Republic of Ethiopia: Poverty Reduction Strategy Paper: Growth and Transformation Plan 2010/11–2014/15 – Volume I. Available online: https://www.imf.org/external/pubs/ft/scr/2011/cr11304.pdf (accessed April 18, 2014) 14 Woredas constitute the third level (after regions and zones) in the country’s decentralized administrative structure. Each woreda is composed of kebeles (villages), the smallest administrative units in Ethiopia. 15 Management Response: 3. Available online: http://ewebapps.worldbank.org/apps/ip/Pages/ViewCase.aspx?CaseId=88 (accessed November 12, 2014) 16 Overseas Development Institute (2010): Millennium Development Goals Report Card: Measuring Progress Across Countries: 9. Available online: http://www.odi.org/publications/5027-millennium-development-goals-mdg- report-card-measuring-progress-across-countries (accessed June 17, 2014) 17 African Union Commission, UNECA and others (2013): MDG Report 2013: Assessing Progress in Africa toward the Millennium Development Goals. Food security in Africa: Issues, Challenges and Lessons. Available online: http://www.zw.undp.org/content/dam/undp/library/MDG/english/MDG%20Regional%20Reports/Africa/MDG%20 Report2013_ENG_Fin_12June.pdf (accessed June 17, 2014) 18 Khan, Q., Faguet, J., Gaukler, C., Mekasha, W. (2014): Improving Basic Services for the Bottom Forty Percent. Lessons from Ethiopia. A World Bank Study. (Poverty and Social Impact Assessment – PSIA): 5. 2 44. PBS III Development Objective. PBS III, which is the focus of this investigation, was approved on September 25, 2012 for USD 600 million, and structured as a five-year Specific Investment Loan (SIL). The Program Development Objective of PBS III, as identified in the Financing Agreement, is “to contribute to the higher-level objective of expanding access and improving the quality of basic services, by funding block grants that ensure adequate staffing and operations, and by strengthening the capacity, transparency, accountability and financial management of government at Regional and Local Authorities levels”19. PBS III serves approximately 84 million people in Ethiopia.20 45. The PBS III includes three components: - Sub-program A: Basic Service Block Grants (USD 6.2 billion total; World Bank USD 555 million); - Sub-program B: Strengthening Local Accountability and Transparency Systems (Citizen’s Engagement, Local Public Financial Management and Procurement, and Managing for Results) (USD 114.6 million total; World Bank USD 40 million). This component includes a social accountability program and strengthening or establishing Grievance Redress Mechanisms; - Sub-program C: Results Enhancement Fund (USD 78.2 million, DFID-administered; no World Bank funds).21 46. Recurrent Expenditures and Decentralized Disbursement. The Basic Service Block Grants under Sub-Program A provide for recurrent expenditures (salaries, operations and maintenance) in the five basic service sectors at the local level to ensure adequate staffing and operations. The funds are disbursed from the federal level to the regions and from the regions to the woredas.22 The federal government provides “un-earmarked block-grant financing”23 through the regional governments, which subsequently provide block grants to the woreda administrations. Each woreda allocates the resources to the five sectors as needed.24 A large part of the Block Grant resources are used to fulfill salary obligations: 85 percent of the Block Grants finance recurrent expenditures, of which 80 percent is allocated to salaries.25 47. Core PBS Principles. PBS III is based on several interrelated core principles that are used to track progress, and the PAD26 describes them as follows: - Development Effectiveness/Results: The Effectiveness Review addresses the issues of adequate inter-sectoral resource allocation, balanced intra-sectoral allocation and results achieved in terms of access, quality and inclusiveness. - Sustainability: The main objective of the Sustainability Review is to ensure that financing of basic services is sustainable in the long term, even without PBS. The main areas of focus                                                              19 PBS III Financing Agreement: 4. 20 Management Response: 3. 21 PBS III PAD: 12ff. 22 PBS III PAD: 12. 23 PBS III PAD: 2. 24 PBS III PAD: 12. 25 PBS III PAD: 2f. 26 PBS III PAD: 45ff. 3 for the review are the sources of financing for basic services and the cost of service delivery. - Additionality: The main objectives of this principle are ensuring medium-term commitments to increase overall financing for federal Block Grants at the sub-national level, and the flow of resources in a predictable manner. The Additionality Test monitors federal Block Grants as part of total federal spending and sub-national spending on basic services. - Fairness: This principle aims at ensuring that resource allocations between the different government levels are rule-based and transparent, and that federal block grants to the regions and from regions to woredas are actually disbursed. - Equity: This Review’s main objective is to track and assess access to basic services among different regions, woredas and between genders. Possible interventions in the case of discrepancies should be identified. - Fiduciary Probity and Transparency: The PBS Program requires regular submission of financial reports and audits by the GoE and meetings to ensure adherence to standards and a timely follow-up on weaknesses. Another core principle aims at the promotion of greater transparency at all levels. - Predictability: PBS III follows a two-pronged approach to monitor the parties’ short and long term commitments. In the short term, the GoE and PBS Development Partners (DPs) conduct reviews through the Joint Review and Implementation Support (JRIS), and the Joint Budget and Aid Reviews (JBAR). In the long term, they engage in dialogue to develop a financing strategy for decentralized basic services.27 As explained in the Management Response, the Bank assures government compliance with these Core PBS Principles through semi-annual JRIS missions, and the further disbursement of the Basic Service Block Grants depends on the outcome of these missions.28 48. Implementation Arrangements. Ethiopia’s Ministry of Finance and Economic Development (MoFED) is the government agency that is responsible for project implementation. Within it, a Coordinating Unit has been established to coordinate daily PBS activities across the ministries and sub-national government entities that are responsible for basic services. At the regional level, the Bureaus of Finance and Economic Development (BoFEDs) are the responsible agencies and at the local level, the woreda and urban administrations. Additionally, the DPs established a PBS Secretariat which facilitates and coordinates the dialogue on the program implementation, and which is funded through contributions to a Multi Donor Trust Fund.29 49. Safeguard Policies. According to the PAD, the safeguard policies were not triggered for PBS III, and the project was assigned an environmental category of “C.” For PBS I and II the Environmental Assessment and Involuntary Resettlement policies were triggered because of small infrastructure works through a pilot Local Investment Grant (LIG), a specific purpose grant for capital investments at the woreda level. The LIG component, however, was not included in PBS III. With regards to the Involuntary Resettlement Policy, the PBS III Integrated Safeguard Data                                                              PBS III PAD: 45ff. 28 Management Response: 5. 29 PBS III PAD: 24ff. 4 Sheet (ISDS) at the appraisal stage reads that “this project does not involve land acquisition leading to involuntary resettlement or restrictions of access to resources or livelihoods”30. 1.3. Issues Raised in the Request for Inspection 50. The Requesters state that they have been severely harmed by the PBS Program as a result of the Bank's non-compliance with its policies and procedures. They argue that PBS is "contributing directly to the Ethiopian Government's Villagization Program in Gambella Region."31 The Requesters claim that Anuak indigenous people are being forced to leave their ancestral land under the CDP,32 and that the land is then leased to investors. The Request letter states that “[t]hese mass evictions have been carried out under the pretext of providing better services and improving the livelihoods of the communities.”33 The Requesters further claim that the new relocation sites, which they were forcibly moved to, provided infertile land, no schools, clinics or other basic services. According to the Requesters, the Safeguard Policies OP 4.10 (Indigenous Peoples) and OP 4.12 (Involuntary Resettlement) should have been applied. 51. The Request also states that the choice of lending instrument was inappropriate considering the design and implementation environment of the Project. According to the Requesters, PBS funds were diverted, including deductions from salaries or non-payment of salaries to finance the implementation of the CDP. It also alleges that government workers who receive their salaries out of the PBS budget were implementing the CDP. 52. The Requesters question the Bank’s diligence with regards to its social and environmental risk analysis.34 They argue that the Bank’s risk assessment did not take into account social, historical, legal and other specific features of Ethiopia that would influence institutional performance at both the regional and woreda level, and decision-making processes did not consider sociological aspects of PBS activities.35 The Request claims that the Bank should have engaged in communication with the different levels of government when it first became aware of the CDP, and should have conducted a risk assessment with respect to all affected people, including indigenous groups.36 1.4. Management Response 53. Bank Management focuses its response on those aspects of the Request that relate to PBS III, explaining that Bank funding for PBS I was closed and PBS II more than 95% disbursed at the                                                              30 PBS III Integrated Safeguard Data Sheet (ISDS) - Appraisal Stage: 3. 31 Request for Inspection: 1. 32 The Requesters allege that “villagization has been carried out by force and accompanied by gross violations of human rights.” The Anuak “have been victims of inter alia threats and harassment; arbitrary arrest and detention; beatings and assault in some cases leading to death; torture in custody; rape and other sexual violence; forced displacement from traditional lands, homes and livelihoods; destruction of property including housing and crops; and inhumane conditions at the new villages including a lack of access to food and livelihood opportunities, in some cases leading to starvation.” (Request for Inspection: 7f.) They argue that the Bank did not fully consider the effects of the PBS Program on the people in the project area and did not ensure consistency with international agreements, such as human rights treaties. 33 Request for Inspection: 1. 34 Request for Inspection: 13. 35 Request for Inspection: 14. 36 Request for Inspection: 18. 5 time the Request was submitted. Management states that PBS III “does not finance villagization and does not depend in any way on villagization in order to achieve its objectives”37 that it “does not build upon villagization”38 and that it “is not synchronized with villagization.”39 Management therefore argues that the harm described in the Request does not arise from PBS III or a Bank failure in applying its policies. Nevertheless, Management is “deeply troubled by the reports”40 and “takes these allegations of harm very seriously and has undertaken an extensive review of the allegations raised in the Request.”41 54. Management explains that the GoE did not consult or advise the Bank or the other DPs about the CDP. It states that “[w]hen the Bank became aware of the villagization program in late 2010, the Bank quickly recognized that this program was an important element of the GoE’s development strategy in FY11 to FY13 for the Developing Regional States and that it was likely to face implementation challenges.”42 The Bank and other DPs have tried to “achieve influence”43 through policy dialogue with the GoE. Management explains that “[t]he Bank and other Development Partners concluded that the best approach to engaging in a dialogue on the villagization program would be through a separate and direct approach to the topic by DAG members, rather than through attempting to expand the scope of any partner-funded program.”44 55. With regards to the allegation of human rights abuses, Management refers to missions of the DAG45 which included Bank staff and which did not find evidence of forced relocations or systematic human rights abuses in the implementation of the CDP. The Response also states that local government officials often play multiple roles and it would not be “feasible nor desirable”46 to pursue arrangements where officials work exclusively on functions supported by Bank funds. Furthermore, the implementation support arrangements for PBS III enable a close monitoring of the Project. 56. Management sees no scope for the application of the safeguard policies on Indigenous Peoples (OP 4.10) and Involuntary Resettlement (OP 4.12) to the CDP, as the Bank does not finance this Program. Nevertheless, Management explains that it “continues to have a close regard for environmental and social issues as they relate to the achievement of the project’s development objective.”47 It also states that PBS III supports strengthening of environmental and social assessments at the local level through assessing capacity, targeting capacity building and encouraging best practices in planning and implementation. Management also explains that the                                                              37 Management Response: 16. 38 Management Response: viii. 39 Management Response: viii. 40 Management Response: viii. 41 Management Response: viii. 42 Management Response: 19. 43 Management Response: 10. 44 Management Response: 19. 45 The DAG was established in 2001 as a forum to share and exchange information, and comprises 27 development agencies active in Ethiopia, which are also referred to as Development Partners. Its main objective is to ensure more effective delivery and utilization of development assistance. For more information, please see: http://www.dagethiopia.org (September 1, 2014) 46 Management Response: 15. 47 Management Response: 18. 6 Bank is conducting analytical work on social and environmental issues specifically relating to PBS III.48 57. Extensive implementation support is also part of the Project’s approach: under PBS III, this includes JRIS missions with DPs and Government officials, extensive financial management review, monthly meetings with the DPs reviewing implementation progress, and constant close coordination with the Ministry of Finance and Economic Development (MoFED), regional governments, other relevant ministries, and DPs. The Task Team Leader in the country and a donor-financed Secretariat with a coordination mandate ensure extensive field-based presence.49 1.5. Project Context 58. The Panel gathered the information presented in this section from a number of sources, including research work by Dr. Eisei Kurimoto, an expert anthropologist on the Anuak people and their history, who assisted the Panel in its investigation. 1.5.1. The Gambella Region 59. Ethiopia has a population of 94 million and is divided into nine regions, of which Gambella is one of the smaller ones, covering about 25,000 sq. km., and located in western Ethiopia about 780 kilometers from Addis Ababa, bordering South Sudan. According to the Panel’s experts, the center-periphery structure of the Ethiopian state is an important feature for its societal and political dynamics. The Ethiopian population is divided into so-called highlanders and lowlanders. Social anthropologist Dereje Feyissa explains that the term highlander is “as much a geographic description as it is a metaphor for power.”50 60. The PSIA states that areas including today’s Gambella region, as well as Benishangul- Gumuz, Southern Nations and Nationalities, Afar, Oromia and Somali were incorporated into the Ethiopian empire in the late 19th century, and “[r]elations between the newly integrated areas and the historic center of the empire were troubled.”51 The PSIA explains that “[o]wing to the structural weakness of the center, successive Ethiopian governments did not command effective control over the periphery”52 and “exploitive economic policies promulgated by the center resulted in visible marginalization and relative underdevelopment at the periphery of the empire. This dynamic persisted through the socialist era.”53 According to Dr. Feyissa, “the Ethiopian state has been introduced through, identified with and is represented by the highlanders.”54 Also, as explained in Management’s Response, “the Gambella region was subject to neglect – infrastructure and social services were virtually nonexistent”55 until the end of Haile Selassie’s rule in 1974.                                                              48 Management Response: 18. 49 Management Response: 8f. 50 Feyissa, D. (2005): Land and the politics of identity. The case of Anywaa-Nuer relations in the Gambella region. In: Evers S., Spierenburg, M., Wels, H. (2005): Competing Jurisdictions: Settling Land Claims in Africa. Brill, Leiden. Available online: https://openaccess.leidenuniv.nl/bitstream/handle/1887/20610/ASC-075287668-172- 01.pdf?sequence=2#page=211 (accessed August 10, 2014): 204. 51 PSIA: 39. 52 PSIA: 39. 53 PSIA: 39. 54 Feyissa (2005): 204. 55 Management Response: 13. 7 61. According to the Panel’s experts, the inclusion of Gambella into the Ethiopian empire by the end of the 19th century marked the beginning of a series of violent conflicts. Armies from the outside invaded the region, elephant tusks were exchanged for rifles, and some Anuak nobles became militarily powerful, controlling the trade and flow of arms. In the early 1900s, Gambella became a prosperous trading place. In the 1950s, and also after the socialist revolution of the 1970s, Anuak repeatedly revolted against the Ethiopian central government in the highlands. 62. According to Dr. Feyissa, the distinction between highlanders and lowlanders is also marked by a discourse on color: “[t]he ‘black indigenous’ are contrasted with the ‘red highlanders.’”56 In the 1980s, the Derg regime resettled 60,000 highland farmers from northern and southern Ethiopia to the Gambella region. Today’s Gambella is ethnically mixed: as of the 2007 national census, out of a total of 307,096 people, 46% are Nuer, 21% Anuak, 20% highlanders, 7% Majangir and 3% each Komo and Opuo.57 63. In 1991, the Ethiopian People’s Revolutionary Democratic Front (EPRDF) overthrew the socialist Derg regime (1974-1991) and established a new federal government. At this time, Gambella, which was formerly a district, became a regional state.58 The federal government also adopted a new system - often referred to as ethnic federalism. According to the PSIA, “[a] primary objective of this new government was to reverse the policy of ethnic homogenization and promote ‘equitable and broadly shared’ economic growth among all regions. Achieving this objective involved affirmative action in the form of increased budget allocations to historically less developed regions. Preferential treatment of previously disadvantaged ethnic groups within Ethiopia is helping to create a more equitable base for development.”59 Accordingly, Gambella is designated as one of the four developing regional states. 1.5.2. The Anuak People 64. The Anuak (or Anywaa) people belong to the Nilotic group and live in the lowlands of Gambella and in the Pochalla and Akobo counties of South Sudan. The exact Anuak population is unknown as data is unreliable, but the majority of Anuak live in Ethiopia.60 Linguistically, they belong to the Luo (or Lwo) group of the Western Nilotic languages, while the overwhelming majority of Ethiopians speak Afro-Asiatic languages. Unlike the Nuer or Dinka, who also belong to the Western Nilotic group and who are cattle pastoralists, the Anuak mostly depend on agriculture, fishing, hunting and gathering.61 65. According to the Panel’s experts, the Anuak mainly rely on crops, such as sorghum and maize, for their livelihoods; crop production dominates the rainy season, while fishing in the Baro and Akobo rivers is crucial during the dry season. The territory of the Anuak is sparsely populated,                                                              56 Feyissa (2005): 204. 57 For more information see: http://www.csa.gov.et/index.php/2013-02-20-14-51-51/2013-04-01-11-53-00/census- 2007 (accessed September 2, 2014) 58 Feyissa (2005): 205, 213. 59 PSIA: 39. 60 Kurimoto, E. (1996): People of the River: Subsistence Economy of the Anywaa (Anuak) of Western Ethiopia, In: Essays in Northeast African Studies. Senri Ethnological Studies 43, 1996: 29f. Available online: http://camel.minpaku.ac.jp/dspace/bitstream/10502/784/1/SES43_003.pdf (accessed July 14, 2014) 61 Kurimoto (1996): 30f. 8 with an estimated population density of between four and five people per sq. km. Their land might therefore seem “empty” to outsiders, but suitable land for the specific Anuak means of livelihood is very limited as they mainly depend on the “riverain belt” which extends about 50 to 100 meters on each side of the river. The Anuak have a strong sense of territoriality, and the land of their ancestors is an important element for their self-identification. Since the 1994 census, the Anuak population grew much less than other groups in the Gambella region, and they have thus increasingly become a minority on their ancestral lands. Picture 1:  Gambella Landscape  66. During its investigation visit, the Panel team learned that every inch of land that may appear “empty” and “unused” to outsiders is identified with a specific community who may forage, hunt, collect building materials, or perform other activities in these areas. The Anuak have a very clear and precise understanding of the boundaries of their customarily owned lands. The Anuak’s strong ties to their land go far beyond agriculture and hunting; according to Feyissa, “[t]he spiritual dimension of the earth is expressed in everyday forms of greetings. While enquiring about personal well-being an Anywaa asks, piny bede nidi (how is the earth?) to mean how are you? One replies piny ber jak (the earth is fine) or piny rac (the earth is bad) to mean everything is well or things are bad, respectively. In Anywaa spiritual imagination the relationship between earth and the people living on it is a very intimate one and this is mediated through the wat-ngomi, who is entitled to allow or refuse human intervention in nature.”62 The Panel team was also told that land is significant for Anuak’s security and for a long time the forest was used as a place of hiding when they found themselves under threat. 67. According to the Panel’s experts, the Anuak have a strong sense of ethnic identity, but do not have a single political system that encompasses everyone. The village is the central unit of their political and social organization. There are two types of political systems in the traditional Anuak society: that of the nobles, and that of the headmen. The Anuak have undergone drastic changes in their lifestyle and culture over the past decades, particularly since the 1974 socialist regime, when they were included into the state administration to a much larger extent than before.                                                              62 Feyissa (2005): 208. 9 Primary schools were constructed in remote areas, peasant, women and youth associations were formed, and many traditional institutions and customs were abolished, as they were “allegedly reactionary and anti-revolutionary.”63 The socialist regime deposed all nobles and headmen. After the fall of the regime in 1991, there was a move to restore the traditional rules, but this restoration remained quite limited. 1.5.3. Conflicts Involving the Anuak 68. The Anuak in Gambella have experienced a history of conflict, not only with the government and military, but also with other ethnic groups, particularly the Nuer. According to Bank Management, conflicts took place “both between groups and within groups, with a variety of victims and aggressors.”64 Gambella’s ethnic groups faced difficult situations not only within the region and within Ethiopia, but also beyond the state borders. According to Management, during the Derg era “Gambella was subject to armed conflict and instability as a by-product of military actions launched against Sudan by the Sudan Peoples’ Liberation Army (SPLA) from bases in Gambella. The Nuer mostly sought sanctuary in refugee camps, but the Anuak took up arms.”65 Dr. Jason Mosley, a Chatham House fellow, explains that “Ethiopia’s frontiers have played a long and significant role in the country’s political economy: as a location to escape central authority, and also (most recently) as the base of the revolutionary movements that overthrew the state and took power in Ethiopia (and Eritrea) in 1991.”66 69. The conflict between the Anuak and the Nuer can be traced back to the early 20th century.67 The different livelihoods pursued by the two groups are a central cause: traditionally, they had arrangements whereby the agriculturalist Anuak would provide grazing land to the pastoralist Nuer and in return, the Anuak would benefit from milk and other cattle provisions. The effective traditional conflict resolution mechanisms between the two groups lost their impact over time due to administrative reforms by the federal government, growing population pressure, and increasingly scarce land and water resources.68 70. In the 1980s, the socialist Derg regime abolished the Anuak’s traditional political system and their cultural practices and promoted the Nuer elites to higher offices.69 The Nuer dominated Gambella’s local administration until 1991.70 The Anuak established the Gambella People’s                                                              63 Kurimoto (1996): 47. 64 Management Response: 13. For more information on the different types of conflicts in Gambella, see: Feyissa, D. (2009): A National Perspective on the Conflict in Gambella. In : Proceedings of the 16th International Conference of Ethiopian Studies, ed. by Svein Ege, Harald Aspen, Birhanu Teferra and Shiferaw Bekele, Trondheim 2009. Available online: http://portal.svt.ntnu.no/sites/ices16/Proceedings/Volume%202/Dereje%20Feyissa%20- %20A%20National%20Perspective%20on%20the%20Conflict%20in%20Gambella.pdf (accessed November 12, 2014) 65 Management Response: 13. 66 Mosley, J. (2012): Gambella violence and land deals: a link? Available online: http://focusonthehorn.wordpress.com/2012/06/18/gambella-violence-and-land-deals-a-link/ (accessed September 2, 2014) 67 UN Emergencies Unit for Ethiopia (2002): Breaking the Cycle of Conflict in Gambella Region: 3f. 68 UN Emergencies Unit for Ethiopia (2002): 3f. 69 Feyissa (2005): 213. 70 UNICEF Addis Ababa (2006): Livelihoods & Vulnerabilities Study: Gambella Region of Ethiopia. Available online: http://www.consciousbeingalliance.com/2012/08/unicef-ethiopia-report-on-anuak-genocide/ (accessed August 10, 2014): 16. 10 Liberation Movement (GPLM), which then allied with the EPRDF to overthrow the Derg regime. After the regime change in 1991, the Anuak were politically promoted by the EPRDF because of their contribution to ending the Derg era, while the Nuer were initially associated with the defunct regime.71 71. According to Dr. Kurimoto, the 1994 national census, the first since the EPRDP came to power, showed that the Anuak only represented a little more than a quarter of the total population in the region they claimed to represent, and that the Nuer were in the majority. The Nuer consequently demanded more power and also founded their own party, the Gambella People’s Democratic Party (GPDP). According to Management, the Anuak were “dominating the regional political space and occupying most managerial posts in regional bureaus. The Nuer sought more equitable political representation, but it was not until after a conflict between the Anuak and Nuer in 2003 that a new power-sharing arrangement and an ethnically-balanced form of zonal administration were put in place. Based on this arrangement, the Nuer, as the ethnic group with the largest population in Gambella, have the largest representation in local councils.”72 1.5.4. Recent Developments 72. In recent years, several major outbreaks of violence involving the Anuak have been reported. According to UNHCR, in December 2003, eight people travelling in a vehicle to a refugee site were killed in an ambush 18 km from Gambella town, among them three employees of the government’s Department of Administration for Refugee and Returnee Affairs (ARRA), UNHCR’s main implementing partner. This ambush sparked large-scale violence,73 although it was unclear who was responsible for the killings.74 Only a few weeks after these killings, another violent incident took place in Gambella.75 In April 2012, unknown gunmen attacked the premises                                                              71 Feyissa (2005): 213. 72 Management Response: 13. 73 World Organization against Torture (2004): Ethiopia and Sudan: “From today forward there will be no Anuak” – The attempted elimination of the Anuak people. Available online: http://reliefweb.int/report/ethiopia/ethiopia-and- sudan-today-forward-there-will-be-no-anuak-attempted-elimination-anuak (accessed September 2, 2014). The World Organization against Torture reports that “424 people were reported to have been killed, with over 200 wounded and approximately 85 people remain unaccounted for.” According to the Norwegian Refugee Council’s Global IDP Database compilation report for Ethiopia, “[i]n December 2003 as they were blamed for the death of eight people, 5,000 Anyuak sought refuge in a church and between 60 and 424 others were killed”. Available online: http://www.internal-displacement.org/assets/library/Africa/Ethiopia/pdf/Ethiopia-July-2004.pdf (accessed September 2, 2014) The Gambella Inquiry Commission found that “65 persons died while close to 75 were wounded and thousands were displaced and fled across the border”. Available online: http://www.ethioembassy.org.uk/news/press%20releases/RESOLUTION%20ON%20THE%20REPORT%20SUBM ITTED%20BY%20THE%20GAMBELLA%20INQUIRY%20COMMISSION.htm (accessed September 2, 2014) 74 UNHCR (2003): Ethiopia: Gambella situation still tense after killings. Briefing Notes. Available online: http://www.unhcr.org/3fdee7a34.html (accessed August 8, 2014) 75 It was reported that “one of Ethiopia's worst outbreaks of communal violence when about 200 people were killed in the Dima district of Gambella on January 30, 2004, most of them traditional miners. Reuters reported that, according to the Federal Affairs Ministry, the atrocity was conducted by over 200 men claiming to be Anuak leaders. Reuters in Addis Ababa (February 11, 2004): 200 killed in Ethiopian ethnic feud. Available online: http://www.theguardian.com/world/2004/feb/12/ethiopia (accessed August 8, 2014) 11 of an agricultural and irrigation business in the Awobo woreda in Gambella and according to different news and government sources, at least five persons were killed.76 73. A 2006 UNICEF research study found that “[i]t is impossible to separate the problems of livelihood and vulnerability of Gambella’s women and children from their problems of security.”77 The study notes that people interviewed uniformly noted that access to basic resources and their lives in general are profoundly affected by problems of security and protection.78 74. The influx of refugees has in the past decades caused stress to the political and economic situation in Gambella and continues today with increasing severity. According to a March 2014 article in The Guardian, when fighting erupted in Juba in December 2013, Gambella was home to more than 76,000 asylum seekers from South Sudan; UNHCR was preparing to accommodate an influx of 150,000 refugees, but the government was concerned that the actual number would be much higher.79 75. Agricultural investment in Gambella is another area that has received increasing attention in the past years. The Guardian reported in a March 2011 article that “[s]parsely-populated Gambella is at the center of the global rush for cheap land […]. Nearly a quarter of its best farmland, and 896 companies have come to the region in the last three years.”80 1.5.5. Description of the CDP 76. Introduction of the CDP. According to Management, “[t]he GoE began implementing the current villagization program in mid-2010 in the four Developing Regional States that are historically the most underserved areas of the country: Gambella, Afar, Benishangul-Gumuz and Somali. These regions collectively represent about 9 percent of the overall population. The Bank’s                                                              76 http://www.state.gov/j/drl/rls/hrrpt/2012/af/204120.htm, http://www.bloomberg.com/news/2012-05-30/saudi-star- offers-jobs-to-overcome-criticism-of-ethiopia-project.html, http://www.ethiomedia.com/2012_report/3750.html (accessed August 9, 2014) 77 UNICEF Addis Ababa (2006): 13. The study found that “the deteriorating vulnerability situation in Gambella in the last two years is primarily due to the protection problems. These protection problems have been caused by the heavy ENDF presence and their actions to target the civilian population often--but not always--in collaboration with Regional authorities, as well as the targeting of civilians by paramilitary rebel groups.” UNICEF Addis Ababa (2006): 12. 78 UNICEF Addis Ababa (2006): 12. 79 Thomas Reuters Foundation. UNHCR source. Available online: http://www.trust.org/item/20140714103912- f9ply/ (accessed August 8, 2014) The Guardian reported that 95% of the people seeking refuge in Ethiopia were women and children. So far, local communities have hosted refugees and it has helped that both groups are from the same ethnic Nuer group. However, “the absorption of a huge number of people into a region with a population of about 307,000 is bound to present problems.” Jobson, E. for The Guardian (March 24, 2014): Humanitarian crisis looms as refugees from South Sudan pour into Ethiopia. Available online: http://www.theguardian.com/global- development/2014/mar/24/humanitarian-crisis-refugees-south-sudan-ethiopia (accessed August 8, 2014). In July 2014, it was reported that the conflict is driving 883 refugees into Ethiopia every day and that “UNHCR and its partners have now revised the planning figure upwards to 300,000”. http://www.trust.org/item/20140714103912- f9ply/ 80 Vidal, J. for The Guardian (2011): Ethiopia at centre of global farmland rush. Locals move out as international contractors seize opportunities offered by government to lease farmland at knockdown rates. Available online: http://www.theguardian.com/world/2011/mar/21/ethiopia-centre-global-farmland-rush (accessed September 2, 2014) 12 understanding, based on official documentation available on the program and related discussions with Government, is that villagization has a three-year time horizon to transform the spatial distribution of willing, local populations in remote and under-served areas in the four regions to facilitate more effective delivery of services. Specifically, the program aims to congregate on a voluntary basis dispersed populations within kebeles around village centers in the same kebeles. It does not provide for any forced or long distance resettlement, nor for mixing people of different ethnic backgrounds. Villagization is not meant (despite the suggestion in the Request) to replace other development programs in the four regions in which it operates nor to provide all the services and facilities that are needed by the communities involved.”81 77. According to the DPs, the CDP involves the relocation of around four million people in the four Developing Regional States into new communities or “development centers.” The office of the Ethiopian Ombudsman in Gambella explains that the Program “would make social and economic services and modern technologies accessible and ensure sustainable food security and, hence, improve the living standards of the public at large.”82 According to a socio-economic baseline study and assessment of the impact of villagization commissioned by the Ministry for Foreign Affairs of Finland, 83 the participating communities would retain rights to their original lands, and people that moved could continue working on their old fields for two to three years. Land fertility and the availability of water resources were central criteria in the selection of villagization centers and, whenever possible, locations with already existing infrastructure were chosen rather than establishing entirely new villages. In cases where this was not possible, a plan was prepared on how to provide the needed infrastructure.84 78. CDP Framework in Gambella. In the Gambella region, the regional government’s strategy for improving access to basic services is described in the Villagization Program Action Plan of 2010,85 which aims at providing “access to basic socioeconomic infrastructure,”86 food security as well as “socioeconomic & cultural transformation of the people.”87 The Panel was informed during its visits to Ethiopia that the Action Plan is carried out under the authority of a Steering Committee chaired by the Regional President. The Regional Bureau of Agriculture coordinates the Plan’s activities and several sectoral bureaus, which form a Technical Committee, are responsible for implementing the Action Plan. The Ministry of Federal Affairs assists the regional government. At the woreda level, Steering and Technical Committees are responsible for implementation. 79. CDP Financing. During its eligibility visit, government officials in Addis Ababa informed the Panel team that the costs of the CDP’s physical infrastructure were covered by the regional capital budget and the MDG Fund, while the operational costs were covered by the regional recurrent budget.                                                              81 Management Response: 9. 82 Institution of the Ombudsman of the Federal Democratic Republic of Ethiopia (2013): Report on Self-Initiated Investigation upon Villagization Program and Investment Land Allocation in the Gambella Peoples National Regional State: 7. 83 Finnmap (2012): Socio-economic baseline study and assessment of the impact of the villagization: 1. The study was carried out between February and April 2012. 84 Finnmap (2012): 19. 85 Gambella Peoples’ National Regional State - Villagization Program Action Plan (2003 EFY) 86 Villagization Program Action Plan: 1. 87 Villagization Program Action Plan: 1. 13 80. CDP Implementation Strategy in Gambella. In Gambella, the CDP planned for the resettlement of 45,000 households, or close to 70% of the total of some 66,500 households88 in villagization centers. 60,000 hectares of land were to be delineated and provided to the target beneficiaries. Each beneficiary household of the program would be provided access to three to four hectares of arable land. Regarding the investment in socio-economic infrastructure, the Action Plan mentions 19 primary schools, 51 water schemes, 22 health posts, 18 veterinary clinics, 40 flour mills, 49 warehouses and 195 km of roads, all of which would be established in the settlement areas. Another program activity is capacity building on how to allocate lands through training of surveyors, zonal administrators, woreda steering and technical committee members, local administrators and community leaders. 89 81. The Action Plan describes the implementation strategies as: transparency, participation, timeliness, land certification, monitoring and evaluation. The Plan identified the following threats to the CDP: different cultural affiliations of the target beneficiaries that might change their opinion about the program, lack of resources and budget, and problems with project management and sequencing.90 82. Implementation Outcomes. According to the DPs, the regional government reported that the program formally ended in Gambella in mid-2013. The CDP had resettled a total of 37,883 or approximately 60% of households in Gambella, against the target of 45,000; the Program over-achieved its target of 15,000 in its phase I, when 20,243 households moved to new locations. Phase II, which aimed at 20,000, was behind target. In 2013, a DP mission noted that the government estimates that 3,000 to 4,000 households chose to return to their original lands. Across 95 resettlement areas, the government provided four hectares of land for each household, 46 schools, 46 health posts, 39 animal health posts, 370 water schemes, 85 grinding mills and warehouses, and 412 kilometers of rural roads. 1.6. The Panel’s Investigation Process 83. The Panel’s Recommendation to Investigate. The Panel reviewed the Request and Management Response and conducted a visit to Ethiopia, South Sudan and Kenya to meet with the government of Ethiopia, the Requesters, the DPs and staff at the Bank’s country office. Following the visit, the Panel judged that further consultations with Management were necessary. Following these consultations, the Panel decided to recommend an investigation to the Board of Executive Directors on February 8, 2013. 84. In its Report and Recommendation, the Panel indicated that the investigation would have a specific focus, as it would examine the risks of the concurrent implementation of the PBS and CDP in Gambella, Management’s analysis of these risks, and subsequent Management actions. The Panel also stated that the investigation would not seek to verify allegations of specific human rights abuses linked to the CDP, nor the underlying purposes of CDP, as these would exceed its                                                              88 http://www.csa.gov.et/newcsaweb/images/documents/surveys/Population%20and%20Housing%20census/ETH- pop-2007/survey0/data/Doc/Reports/National_Statistical.pdf (accessed November 5, 2014): 22. 89 Villagization Program Action Plan: 1f. 90 Villagization Program Action Plan: 4f. 14 mandate.91 The Board of Executive Directors approved the Panel’s recommendation to investigate on July 12, 2013. 85. Investigation Process and Methodology. The focus of the Panel investigation is to determine whether the harm alleged by the Requesters may result from a Bank’s lack of compliance with its policies and procedures in the design, appraisal, and implementation of the PBS III. 86. The Panel conducted a two-part investigation led by Panel Chairperson Eimi Watanabe. The first part involved detailed research into Bank records related to the Project, as well as an extensive review of relevant Project documents. The second part entailed a fact-finding mission to the Project area and interviews with Bank staff involved in the Project. The Panel retained four independent experts to assist with the investigation: Prof. Stephen Peterson and Mr. Roberto Mosse, both financial management experts; Dr. Eisei Kurimoto, an anthropologist and expert on the Anuak indigenous peoples; and Dr. Alison Evans, an expert on risk assessment. 87. A Panel team composed of Panel Chairperson Eimi Watanabe, Deputy Executive Secretary Dilek Barlas, Senior Operations Officer Tatiana Tassoni, and expert consultant Eisei Kurimoto visited Ethiopia from February 1 to 10, 2014. During its investigation visit, the Panel team visited a number of villages in three woredas in Gambella and met with people affected by the CDP and PBS. The team also met with World Bank staff in the country office in Addis Ababa, representatives of the DAG, including officials of the European Union (EU), the African Development Bank, the UK Department for International Development (DFID), and the Embassies of Finland and Germany. In Gambella, the Panel also met with the local Ombudsman and with the Vision Ethiopia for Congress of Democracy (VECOD), which is involved in the implementation of the social accountability component of PBS. 88. About a month after the approval of the Panel’s recommendation to investigate the claims alleged in the Request for Inspection (August 2013), the Panel published its Investigation Plan. This document indicated that the investigation would assess: a) whether the claim concerning links between PBS III and CDP are credible; b) whether the Bank’s assessment of risks and integration of risk mitigating measures from the concurrent implementation of PBS III and CDP is in compliance with applicable Bank policies and procedures; c) whether the Bank’s safeguard polices are applicable to PBS III, in particular OP 4.10 on Indigenous Peoples and OP 4.12 on Involuntary Resettlement; and in relation to these, specific issues of non-compliance, if any, that may have caused the harm alleged in the Request. 89. The investigation assessed whether the Bank complied with the following Operational Policies and Procedures: Project Appraisal – OMS 2.20 Indigenous Peoples – OP/BP 4.10 Involuntary Resettlement – OP/BP 4.12 Financial Management – OP/BP 10.02 Investment Project Financing (OP 10.00)                                                              91 Report and Recommendation: 21. 15 Chapter 2: PBS III and CDP - Operational Interface and Risk   90. The main claim of the Requesters is that the Government’s CDP and the Bank-financed PBS III project are linked, and that PBS III enables the implementation of the CDP, which in turn is causing the alleged harms. This Chapter analyzes the link between PBS III and CDP, and how this was appraised, followed by the risk assessment under PBS III. 91. This Chapter is based on information gathered by the Panel during its field visits and provided in project documents and other relevant sources. The Chapter’s analysis also draws from mission reports of the DPs, including the World Bank, which the Panel has carefully reviewed. These mission reports are not publicly available and as such the Panel also treats them as confidential information. While the Panel uses information included in these reports, it neither quotes nor provide references for them. 2.1. The Operational Link between PBS III and CDP, and Allegations of Harm 2.1.1. Requesters’ Claims and Management Response a) Requesters’ Claims 92. The Requesters argue that PBS is "contributing directly to the Ethiopian Government's Villagization Program in the Gambella Region."92 The CDP in Gambella is intended to improve access to basic services and its execution rests “with regional and woreda government” and has been under concurent implementation since “mid to late 2010 during phase II of PBS.” 93 The Requesters claim that “services and facilities supported through PBS are precisely the services and facilities that are supposed to be provided at new settlement sites under the Villagization Program.”94 They argue that it is therefore “apparent that villagization is the regional governments’ principle strategy for achieving the PBS objectives in Gambella.”95 The Request refers to the PAD for PBS II which explains that PBS is the major source of funding at the woreda level.96 They argue that PBS provides “the means to make villagization possible.”97 93. The Requesters also argue that PBS I, II and III constitute a single continuous program with only minor modifications at each phase, and that the manner in which PBS III is designed provides critical resources to woredas for their implementation of designated local level development activities, villagization being among them.                                                                92 Request for Inspection: 1. 93 Request for Inspection: 5. 94 Request for Inspection: 5. 95 Request for Inspection: 6. 96 Request for Inspection: 6. 97 Request for Inspection: 6. 16 b) Management Response 94. Management explains in its Response that PBS III “does not finance villagization and does not depend in any way on villagization in order to achieve its objectives”98 and also “does not build upon villagization”99 and “is not synchronized with villagization.”100 Management argues that “villagization is neither meant to replace other development progams in the four regions in which it operates nor provide all services and facilities that are needed by these communities.”101 Management concludes that “the harm described by the Requesters is unrelated to the Bank- supported PBS 3.”102 2.1.2. Assessment of the Link 95. To assess whether there is a link between PBS III and the CDP, the Panel examined the following features that would point to the existence of such a link: i. Commonality of objectives, ii. Concurrent implementation and scope, iii. Mutual impacts on the two programs’ respective results, and iv. Contribution to the implementation of CDP: whether PBS resources were used to implement activities under the CDP program. 96. The last element is analyzed in depth in Chapter 4 of this Report. The paragraphs below consider the first three elements. 97. Commonality of Objectives. The Panel notes that both PBS and CDP have the objective of providing improved basic services to the population, which is a main element of the GoE’s overall development strategy. In a recent letter to the DAG group, the Government stated that the CDP is one among several approaches pursued to improve livelihoods, expand basic services and build grassroots institutions of communities.103 Management outlines its understanding of the objective of the CDP as "to cluster dispersed populations in sparsely populated regions into 'commune centers' over a period of three years on a voluntary basis to enable more efficient provision of basic services."104 The Bank defines the objective of the PBS program as a "nationwide program which contributes to expanding access to and improving the quality of basic services."105 98. Concurrent Implementation and Scope. The Panel examined the geographic overlap of the implementation of these programs. The stated aim of CDP in the Gambella region was to relocate 45,000 households (around 70% of households) over a period of three years, and the establishment of social and economic infrastructure in new settlement areas, including schools, health services,                                                              98 Management Response: 16. 99 Management Response: viii. 100 Management Response: viii. 101 Management Response: 23. 102 Management Response: 14. 103 Letter from the GoE to the DAG, dated March 18, 2014. Subject: - Reply to DAG Findings and Recommendations on CDP and South Omo (DAG/OU/3/2014A, 18 March 2014). 104 Management Response: x. 105 Management Response: vii. 17 water and sanitation facilities and improved marketing facilities.106 The Panel was informed in meetings with the Ministry of Finance that over 70 villages were expanded or created under the CDP in Gambella. From the end of 2010, PBS and the CDP were implemented concurrently in the Gambella region covering the same woredas. In addition, the Panel notes that the civil servants who provide services under the CDP program, such as school teachers, health professionals and agricultural extension workers, are the same workers whose salaries are being paid under the PBS. Since its inception in 2010, CDP has been a priority development effort in most or all woredas in Gambella.107 99. Mutual Impacts. The results indicators linked to the higher-level objectives of PBS are the following: primary education enrolment rate, primary education completion rate, gender parity in primary schools, proportion of children vaccinated, pregnant women who received antenatal care, access to potable water, households with access to latrines, increase in agricultural productivity, and average time to nearest all-weather road.108 The Panel notes that the PBS indicators also reflect the stated objectives of CDP noted above. 100. The Panel notes that Management is of the view that “villagization is not directly or significantly related to PBS III, nor does PBS III depend on villagization in order to achieve the objectives described in the project document.”109 This statement seems to presuppose that the success or failure of either program, PBS and CDP, will not influence the other. In other words, that the PBS results indicators above are not dependent on the effects of CDP. The Panel is of the view that this is not a tenable position. CDP is a program that aims at fundamentally restructuring settlement patterns, service infrastructure and livelihoods, including farming systems in the Gambella region, and as such constitutes a significant part of the context in which PBS operates. In this sense, from a development perspective, the two programs may mutually influence each other’s respective results in the woredas where they are concurrently taking place. This point is illustrated further in Chapter 3. 101. The Panel is therefore of the view that there is an operational interface between the CDP and PBS in the woredas where there is concurrent implementation. This interface constitutes an operational risk for the Project, which required adequate analysis during Project appraisal, and this goes beyond the reputational risk that was recognized by Management. The adequacy of the PBS III risk assessment is discussed in Part C of this Chapter. 2.1.3. The Four Sets of Allegations of Harm 102. The Panel indicated in its Eligibility Report that the Request raised issues of harms that covered four broad aspects: (i) taking of people's customary land without their free and informed consent, (ii) use of force and intimidation to get people to relocate, (iii) lack of public services and assistance at relocation site, and (iv) deterioration of livelihood due to less fertile land and lack of other income generating opportunities at relocation site.                                                              106 Villagization Program Action Plan: 1. 107 Ombudsman (2013): 4. 108 PBS III PAD: 37f. 109 Management Response: xi. 18 103. As mentioned earlier, the Panel noted at the eligibility stage that the “investigation will not seek to verify allegations of specific human rights abuses linked to VP, nor will it examine the underlying purposes of VP”110 as the Panel does not see this to be within its mandate. As a result, the investigation report does not include findings of facts and compliance on the first two issues of harm noted above. Nevertheless, in the course of the Panel’s review of documents and interviews in the field, the Panel came across information regarding those allegations. Given that these issues were raised in the Request for Inspection, the Panel records this information below, without attempting to verify them or otherwise. a) Involuntary Taking of Land and Use of Force and Intimidation 104. The Requesters allege that under CDP, they did not have the right to refuse leaving their traditional lands. They claim that a prime political driving force behind CDP is to free land for long-term lease to private domestic and foreign investors, and accordingly, people did not have the option to object to relocation. They state that “through this program, the Anuak Indigenous People are being forcibly transferred from their fertile ancestral land, which is then being leased to investors.”111 The Request also alleges that “those farmers who opposed the relocation, and government workers who refused to implement the program, including the Requesters and/or their relatives, have been targeted with arrest, beating, torture and killing.”112 105. Management states in its Response that it “is deeply troubled by the Requesters’ reports they have suffered abuse in connection with villagization in Gambella. Management takes these reports very seriously, and has taken advantage of various opportunities outside of the context of any specific project agreements to raise these issues with the GoE. However, to date neither the Bank nor other Development Partners have been able to identify any evidence to substantiate possible links between reported abusive behavior and villagization of the kind described in the Request.”113 Management is of the view that there is lack of “evidence to substantiate possible links between abusive behavior and villagization of the kind described in the Request”.114 Furthermore, Management notes that “the harm described by the Requesters does not stem from the Bank-supported PBS 3, or from a failure by the Bank to apply its operational policies and procedures.”115 106. The Government, in meetings with the Panel in January 2013, stated that the country has abundant unused land that can be operated by agri-businesses without posing a threat to the livelihood of smallholders. There are plans to allocate 3.3 million hectares by 2015 to investors, in addition to the 3.5 million already allotted by 2010.116 They also categorically stated that there is no relationship between these plans and villagization. It was explained to the Panel that villagers who had opted to relocate would retain user rights to their original land for two years, when they would have to decide whether to move back or relinquish claims to the original land. In this two-                                                              110 Report and Recommendation: 21. VP stands for Villagization Programme (CDP). 111 Request for Inspection: 1. 112 Request for Inspection: 1. 113 Management Response: x. 114 Management Response: 19. 115 Management Response: viii. 116 Rahmato, D. (2011): Land to Investors: Large-Scale Land Transfer in Ethiopia. Forum for Social Studies. Addis Ababa: 11. Available online: http://www.landgovernance.org/system/files/Ethiopia_Rahmato_FSS_0.pdf (accessed November 12, 2014) 19 year period, people could use both the land allocated in the new resettlement site and their customarily held land. Government representatives also informed the Panel that coerced resettlement does not take place. They underscored that the Ethiopian Constitution protects people from being coerced, that “abuse will be investigated,” and that Regional governments have to get the consent of people; “voluntarism is without any caveat.” 107. During the Panel’s eligibility visit in January 2013, the Ministry of Federal Affairs (MOFA) organized an opportunity for the Panel mission to meet with Gambella regional government representatives and 15 villagers from five different ethnic groups from Gambella State to talk about the CDP. During the meeting, the Government officials noted the importance of CDP for their five-year plan to double economic growth and added that in the five developing regions where the CDP is being implemented there is vast land, water sources are available, and the population is scattered. They stated that CDP has a cost-effective and people-centered approach and no person can coerce another by force. The commune program can only happen if it is voluntary. They added that the objective of CDP is to improve the livelihoods of people participating in the program. 108. The 15 villagers at the meeting explained to the Panel team that people demand to move to new villages, and there needs to be more villagization centers. The villagers confirmed that no one can force the people to participate in the program and that CDP is implemented following meetings with the communities in the kebeles. They noted that they have been marginalized before and were now very happy with this program as it provided them with services, health, education, water, and even access to markets. 109. DP representatives whom the Panel met reiterated during the two visits that they do not have evidence of systemic human rights abuses related to CDP. They reported, however, that half of the people interviewed said they did not want to move to new places, and there were reports of some pressure and unmet promises. 110. These statements contrast with the interviews with the Requesters and others, whom the Panel met in January 2013 and who are refugees living in camps in South Sudan and Kenya. Several testified that their original land had been given to investors after people had moved to the relocation site. These testimonies referred to land in the Abobo and Gok woredas. The Requesters interviewed by the Panel described concrete incidents of people who had been beaten, detained and even, in some cases, killed by soldiers or police engaged in overseeing the relocation process when they were showing some resistance. Several of the Requesters stated that they fled because they feared for their lives. They were very concerned about the situation of relatives back home and of the prospects of being able to return to their ancestral lands. Some former civil servants stated that they were targeted as “trouble makers” when the villagers refused to move, and had to leave the country due to safety concerns after being beaten or put in jail. 111. While some Requesters specifically cited CDP as the reason for fleeing the country, others recounted instances of being or feeling threatened for reasons other than villagization. The Panel heard multiple reasons why people perceived to be vulnerable and insecure. 112. During its investigation visit, the Panel came across some villages designated as part of CDP where not all villagers had moved, and those not wanting to leave could remain. The Panel’s expert also heard allegations of intimidation and abuses in this context. 20 113. As stated earlier, the Panel does not consider it within its mandate to verify any of the above information, or otherwise. b) Harms Related to Lack of Services at Relocation Sites and Deterioration of Livelihoods 114. The Panel’s investigation and analysis focuses on the two sets of claims, which have, in the Panel’s view, operational links to PBS: (i) lack of access to basic services in new villages and (ii) a resulting deterioration of people’s livelihoods. 115. The Request states that promised basic services and facilities at the new sites were either not provided or were not operational, and there was little access to food or land suitable for farming. Management recognizes that the DAG-sponsored missions found that planning had often been inadequate, relocation was too rapid, the sequencing of site improvements was poor and poorly financed, and implementation capacity was weak. When the Panel team visited Gambella in early 2014, villagers complained that their situation had not changed and that livelihood opportunities, including availability of land to farm, were scarce, which they saw as the cause of an overall deterioration of their livelihoods. 116. Overall, the Panel found a mixed picture: there was some deterioration (e.g. in agriculture), some improvements (e.g. in education), but in general terms, the level of basic services in the villages the Panel visited was still inadequate. Below is an account of information gathered by the Panel during the field visit and provided by DPs following visits to Gambella and other states where CDP was implemented. The information is divided by sector and presents facts on access and quality of services in the five sectors covered by PBS. Agriculture 117. This section presents information relating to land and agricultural production, and also includes information about food security and housing. These topics were raised together in most interviews, and are closely interrelated. 118. Panel Observations. During its investigation visit, the Panel found that many households had received land, but in some cases less than promised. In most of the villages that the Panel visited, people explained that their new land was not cleared and that they were not provided with the necessary tools to make the land useable for agriculture. Promised government assistance for clearing the land was not received. Many communities informed the Panel that they consequently had difficulty in securing their livelihoods. According to several people interviewed by the Panel, the government had also promised support for the construction of houses and provision of food aid but both services were lacking in many cases. Grinding mills, which had been provided in some villages, were allegedly not functional at most sites. 21 Box 1: Panel Interviews in the Field - Land and Agriculture, Food And Housing   Clearing the land was described as a major challenge in most communities: “[i]t was only machetes (panga) that were given to us […] We could only cut branches and grasses with them;” other communities were not given any tools. “We started building our own homes by ourselves. Because we were not given tools to work with, some of us who could not afford to buy tools were using whatever tools they could find to build their homes. Some of us who had small money bought the house materials and hired people to build the hut for them. There are some single women who have built their own huts by themselves.” Many people explained that the government had informed them that it would bring tractors to clear the farmland, would help building huts and provide food, but none of this happened. In one village, the Panel team was told that “[…] some of us who were not able to do this hard work had to look for the money to hire people to do the work for them. When it came to the weeding time most people ran out of money and gave up. They totally stopped farming.” In several villages, the Panel learned that community members continued to cultivate the land in their previous sites, which was, in some cases, up to three hours walking distance away. Some people explained to the Panel that they had not been hindered from going back. In one case, the Panel team was told that the government had prohibited villagers to go back to their original lands, but nothing happened to those who decided to return nevertheless. Locals in one village explained to the Panel that the food produced by those who went back to work on their old land was shared with the people who stayed. Others described to the Panel team that seeds of various crops were distributed to them; however, this happened at a time when it was too late for sowing, so the locals ate them. Several people reported that they did not receive any food rations, cooking material or cans for water. In one village, hunger was cited as the biggest problem of living there. One villager explained to the Panel that his community had been told to build big huts that would provide enough room for the materials that would be provided. “So, I built a big hut like that [pointing at his own hut] and moved here. Before that I used to live on the river bank. This big house is empty. Nothing has been given…” The villager claimed that “[m]any things were promised, but they were not delivered. […] Food aid was given only to the poor and disabled, not to all of us […].” One site, which can be considered a success case according to the Panel’s expert consultant, had more than 10 agricultural extension workers managing a demonstration farm, distributing crop seeds, and demonstrating how to do better bee keeping. Also, 16 women were trained to use cooking stoves in a better way, and 15 households were provided with goats and sheep, and accorded training. The villagers were supplied with axes and hoes. The Panel team, however, was also told that there was no good land available and therefore the villagers would look for more land in the vicinity. The villagers were provided with tools to clear the land, but food security remained a concern. A promised grinding mill had not been provided and women had to do the pounding manually. 119. Other Sources. DPs who undertook missions to Gambella encountered similar issues. In 2011, they noted that some land had been allocated prior to moving, but none of the communities had received agricultural inputs or assistance for clearing the land; the scale and speed of relocation disrupted livelihoods and threatened food security; in most cases, the new land requires mechanical clearance and communities need farming inputs. In 2012, DPs found improvement in conditions of villagized communities, though land clearance remained an issue and people emphasized continued problems with food security. In 2013, DPs reported that the communities visited said that they had received the services the government had promised them in the new villages, but grain mills in almost all villages were not functional; most communities explained that they were still able to access their previous lands and continued to use them for farming. The Gambella office of the Ethiopian Ombudsman in its Villagization Investigation Report notes that most grinding mills had broken down; better farming equipment supported by better technologies has not been 22 made available; the supply of seeds was not timely, and there was an insufficient level of awareness about modern agricultural technologies in some areas.117 The issue of agricultural services linked to livelihoods and resettlement is examined in greater detail in Chapter 3. Picture 2:  Panel Team in Anuak Village in Gambella                                                              117 Ombudsman (2013): 11f, 18. 23 Education 120. Panel Observations. In most Box 2: Panel Interviews in the Field - Education villages visited by the Panel team, schools existed before villagization and were The Panel team learned in one village that the school had opened a seventh grade that year, but before villagization it upgraded as a result of it. In many cases, only provided up to the fourth grade. The government had schools now include a higher grade level. constructed four classrooms and the population had built Several communities reported that the two classrooms themselves using local materials; these two number of students and teachers had classrooms have since collapsed. The school employed 14 increased. People in a few villages, teachers and enrolled 136 pupils. Several other village schools had also opened new grade levels; some went up to however, expressed that not much had the eighth grade. In one village, a new school had been changed since the start of the CDP: promised, but was never built. villagers informed the Panel that village schools still did not have enough space and In one area, the Panel team learned that two small villages that they struggled with too few teachers, from the riverbank were merged together. A school had already existed but only went up to the fourth grade before deteriorating quality of teaching or the villages moved. Now, the school goes up to the seventh decreasing numbers of students. grade. There are 15 teachers and 136 students; the number of teachers increased after the number of grades also 121. Other Sources. In February 2011 increased. In a different case, the number of students had DPs visited a number of communities, half increased from 400 to 600, and the number of teachers from 16 to 29. of which reported that the access to education had improved. Education was In a village that rejected villagization, the Panel was told found to be generally more accessible in the that basic services remained at a minimum: the school only relocation sites, and more children went up to the fifth grade, and there were five teachers for appeared likely to attend school but they 109 students. People were told to move to a different also reported a shortage of equipment in village with a school up to the eighth grade. Children above the sixth grade needed to walk four hours each-way every schools and, in some places, not enough day. school buildings or classroom space. In June 2012, DPs found that in all villages, primary education was reported as better, more accessible and the journey to school safer than in previous locations. In October 2013, DPs found that all communities visited noted that children were now receiving a basic education in schools, generally at grades one to eight. The Ethiopian Ombudsman found that schools up to the secondary level had been built; however, problems with quality and insufficient space were found. The report also states that “[i]n some areas, students are compelled to attend classes in shades of trees.”118                                                              118 Ombudsman (2013): 12. 24 Health 122. Panel Observations. In the health sector, Box 3: Panel Interviews in the Field - Health the information collected in the villages suggests inadequacies in the delivery of services: in most In many communities that had a health center, the villages, local people explained to the Panel Panel team was told that the health workers could only team that they lacked sufficient personnel and provide limited help due to a lack of medicines. In one village, the health center was built before villagization that medicines were largely unavailable. The and employed three female health workers. According Panel was not able to verify whether services to local people, there is very little medicine available, had improved or deteriorated. Several of the except for anti-malaria tablets and medicine for communities visited by the Panel team had been dressing wounds; the health center does not have any promised a new health center, but none of them antibiotic medicine in stock. In another village, the Panel team was told that there was no health center and had been provided with one. the only nurse had left because no medicines were available. 123. Other Sources. In February 2011, DPs found that healthcare was particularly In several sites, the Panel team was told that patients problematic. Most communities reported that with a serious sickness needed to walk between two they relied on traditional medicines for minor and four hours to clinics in other towns. According to community members in one village, four children had illnesses, and took people to Gambella town in died on the way to clinics in the past year. more serious cases. Health centers do not have sufficient medicine and health workers. In June In one location, community members told the Panel 2012, DPs found that people generally did not that they had mosquito nets: health workers gave the find that health care had improved since moving. first nets to them, but they were not replaced when they broke, so the locals bought new ones. There were The main reasons cited were a lack of medicines two health extension workers in the village but they and personnel. In October 2013, DPs noted that did not have access to medicines. new health posts and health extension workers had been provided, and most communities The Panel team was told that since the community had moved near the road, people could call the woreda reported that the situation had improved when somebody is sick, and the woreda would send a compared to that in their old villages. However, car to pick up the patient and drive them to the next some health posts were in bad condition and the clinic. In another village, the Panel heard a similar drug stock was limited or out of date. The story of patients being brought to other clinics by the Ombudsman found that “[f]ew health posts and woreda. veterinary clinics have been constructed but have not begun services; those that are already functional sustain shortages in the supply of pharmaceuticals.”119 One of the advantages of the new locations, people said, is better access to roads that can take them to town in case of emergencies.                                                              119 Ombudsman (2013): 12. 25 Water Supply and Sanitation 125. Panel Observations. During the Panel’s Box 4: Panel Interviews in the Field – interviews in the field, different issues with regards to Water Supply and Sanitation water and sanitation were raised, but concerns with other services dominated. Most people in the People interviewed in several villages told the Panel that a few new boreholes were drilled: in communities visited by the Panel team did not one village people originally had two mention major improvements or a deterioration of the boreholes; after complaining with the situation. government, two more were installed. In another village, after political changes in the 126. Other Sources. In 2011, DPs found that 3 out region, a new borehole was drilled. Residents at a different site reported having four of 12 communities reported improved access to boreholes, of which one was new. drinking water. Nine stated that the number of pumps was too little for the population size. None of the People in one location informed the Panel communities visited had functioning latrines, and only team that they were told to dig holes for pit one had any under construction. Sanitation, latrines, which they did. The promised corrugated galvanized iron sheets, however, particularly waste disposal, is an issue that is almost were not provided so the community ended up completely unaddressed, and there is an increased risk not using the pit latrines. The village had five of water-borne disease. Many communities that were boreholes: the first two were drilled during the using river or rainwater for drinking in the past now Derg era, one by the Catholic Church and two have access to safer water through hand pumps. In by the current government, but one of them was broken. most cases, however, not enough pumps existed for the growing number of people. In 2012 DPs found that People in one community explained to the with the exception of one, all villages visited remarked Panel team that water from their only borehole that access to water was better than in their original was very salty and therefore not used by the villages. The community had pumps next to schools, community; the villagers continued to fetch water from the river. Some people told the health centers and in the village center, most of which Panel team that they had to pay 10 Ethiopian were functional, although some of them produced Birr per month for water.  salty water. Water for livestock was lacking. In 2013, people told DPs mission that access to clean water had improved substantially compared with their original locations. According to the DPs, some communities expressed that they would like to have more water pumps, but there was a general relief that the community no longer had to use dirty river water. Also, all villages had working latrines, but some communities used them more than others, who preferred open defecation. According to the Ethiopian Ombudsman, in some villages people face a shortage of potable water; in others, residents need to travel long distances to fetch water as water supply facilities are far away.120 127. In conclusion, the information gathered above from the Panel’s limited visit, and from other reports, is not intended to be a systematic assessment of the impact of CDP on basic services delivery and outcomes, but simply to identify some trends. Available information indicates that the CDP was having an impact on PBS PDOs, and in some limited instances such as education this impact was positive.                                                              120 Ombudsman (2013): 12. 26 2.2. Risk Assessment Analysis in PBS III Appraisal and Implementation 2.2.1. Requesters’ Claims and Management Response a) Requesters’ Claims 128. The Request notes that the project risks were identified as substantial with respect to governance and commitment to social accountability in Project documents, the mitigation measures are inadequate to address these risks, which have eventually materialized.121 The request argues that the lack of reference in the PAD to the CDP is “striking”122 given the “troubled history of villagization in Ethiopia,” and the Bank’s diligence in carrying out social and environmental risks analysis “has to be called into question.” 129. The Request adds that the Bank did not fully consider the PBS Program’s effects on the well-being of people, including marginalized peoples like the Anuak, and that the Bank did not ensure consistency with applicable international agreements, including human rights treaties to which Ethiopia is a Party. Furthermore, the Bank in its risk assessment did not consider institutional issues at regional and woreda levels as well as the decision-making processes that could affect the achievement of project objectives, and did not consider the sociological aspects of PBS Program activities.123 b) Management Response 130. Management notes that the Bank first became aware of villagization in October 2010. Management states that, based on discussion with the GoE during the JRIS mission in November 2011 and on information gained from other project teams, “the Bank determined that villagization was not linked to the PBS block grants, and that it would be more productive and appropriate to address this issue through its policy dialogue.”124 Together with other DPs working in Ethiopia and collaborating as the DAG, “the Bank has engaged with the GoE since late 2010 in a policy dialogue on villagization, including: (a) undertaking numerous meetings with both federal and regional officials to gather information about the program’s objectives, plans, funding and status of implementation; (b) providing advice to the GoE on good practice guidelines and principles for resettlement; (c) undertaking field visits to learn how the program is being implemented on the ground, and based on these observations; (d) raising concerns about inadequate planning, rapid pace of relocations, poor sequencing of site improvements and weak implementation capacity of the villagization program.”125 131. Management explains that visits by DAG-sponsored missions have, to date, not encountered evidence of forced relocations or systematic human rights abuses connected to the CDP implementation. The multi-agency missions to Gambella, first in February 2011, and later in June 2012, confirmed this finding, but reported that some communities stated “that they objected to relocating and were allowed to stay in their original location.”126 DAG missions, however, “did                                                              121 Request for Inspection: 12. 122 Request for Inspection: 13. 123 Request for Inspection: 13. 124 Management Response: 31. 125 Management Response: 26. 126 Management Response: 10. 27 find that planning had often been inadequate, relocation was too rapid, the sequencing of site improvements was poor and had inadequate finance, and implementation capacity was weak.”127 According to Management, “[t]he GoE accepted the feedback provided by the Bank and other DAG members on problems of implementation of the villagization program and acknowledged that implementation could be improved.”128 132. Management also notes that during the appraisal of PBS III, “the Bank took into account lessons from successful implementation of PBS 2 social accountability and financial transparency and accountability activities, as well as extensive political economy analysis […] to design the citizens’ engagement component and identify potential risks to its implementation.”129 2.2.2. Applicable Bank Policies 133. Bank Policy OMS 2.20, applied during the appraisal stage of PBS III, states that “[i]rrespective of the specific form of analysis adopted for a project, the appraisal normally includes assessing the sensitivity of the project viability to changes in the key parameters of the project, together with a judgment on the likely variation from the basic assumptions. For projects with marginal returns or large risks, further quantification of the risks through formal risk analysis is also desirable. Where necessary, the appraisal also includes recommendations regarding precautionary measures which should be undertaken to reduce the risks […]”130 OMS 2.20 also states that “[f]or large and complex projects, consideration should be given to having a comprehensive risk analysis made by specialists in this field to determine whether risks could be reduced by design changes, and what types and amounts of insurance are most appropriate.”131 Also, with regards to social impacts, OMS 2.20 states “[i]f appraisal determines that the project is likely to be highly risky in social terms, but inadequate information is available to support a firm conclusion, consideration should be given to either a pilot project or postponement of the project until sufficient information is available.”132 134. The Guidance Note on the Operational Risk and Assessment Framework (ORAF) dated July 2011, states that the main purpose of the ORAF is to help managers, project staff and the country team to “look systematically, holistically, in an integrated manner and in real time, at risks to achieving project development objectives (PDOs)”133… […] using the risk assessment during implementation to regularly scan the full spectrum of possible risks helps teams move quickly to identify and address emerging issues, including unanticipated risks.”134 Far from being a one-off event, this Guidance makes clear that risk assessment and management is a dynamic process that takes place throughout the project cycle. It also states that “[i]f a risk management measure is not working, the team may consult with the client and agree to adjust the measure or even adjust the project […].”135                                                              127 Management Response: 27. 128 Management Response: 27. 129 Management Response: 31. 130 OMS 2.20, para. 17. 131 OMS 2.20, para. 23. 132 OMS 2.20, para. 61. 133 Guidance Note on the Operational Risk Assessment Framework (ORAF) (2011): 2. 134 Guidance Note on the ORAF (2011): 2. 135 Guidance Note on the ORAF (2011): 6. 28 135. The Guidance proposes four levels of risk that should be considered as part of the ORAF: - Stakeholder Risk: which includes the risk of opposition to a project that could in turn affect the achievement of project development objectives; - Operating Environment Risk: relating to the environment in which the project operates and not directly to the achievement of PDOs. Awareness of risk is key here as mitigation is not always possible. However, the potential impact of such risks on PDOs should be taken into account in other relevant risk sections of the ORAF i.e. Implementing Agency or Project Level Risk. - Implementing Agency Risk: risk related to specific agency (ies) implementing the project and influenced over the course of the project through mitigation measures and project design, and - Project Level Risk: Related directly to the preparation and implementation of the project and including: design risks, social and environmental risks, program and donor risks, delivery, monitoring and sustainability risk. This is where most scope for mitigation and management of risk lies, primarily through design and implementation. 136. According to the recently approved Investment Project Financing (IPF) Policy (OP/BP 10.00), which is applicable to the implementation phase of PBS III, during implementation the Bank monitors compliance by the Borrower with the obligations included in the legal agreements with the Bank, and provides implementation support “by reviewing information on implementation progress, progress towards achievement of the Project’s development objectives and related results, and updates the risks and related management measures.”136 The Bank Procedure (BP 10.00) specifies that the Bank’s role is to assess the Project periodically, updating Project information and to identify “follow up actions needed as appropriate.”137                                                              136 OP 10.00, para. 21. 137 BP 10.00, para. 40. 29   Figure 1 Timeline of Events 2.2.3. PBS III Risk Assessment a) During Project Design and Appraisal 137. In interviews, Bank staff indicated to the Panel that the Bank first became aware of villagization in October 2010 during unrelated field visits and from press reports. They added that the issue of villagization was discussed among the DPs, and that two separate multi-agency fact- finding missions went to Gambella and Benishangul-Gumuz Regions in February 2011. DPs reported that their missions sought to better understand the CDP, its implementation, and any implications for their supported projects and programs. 138. DPs noted that while there may be practical and reputational risks and implications for international donors and UN-supported programmes with respect to possible aligning with the villagization program, villagization in Gambella was happening and was likely to continue. The need was thus emerging for the international community to identify the best way to engage to mitigate risks to communities and continue to provide support where needed. 139. DPs acknowledged that it was unlikely that any development financing would be provided in support of villagization and that a review of the program against international standards, such as the World Bank Resettlement guidelines, would be required. Following their visits, they recommended additional analyses of any implications for donor-funded developmental programs in order to guard against practical and reputational risks and the need to avoid promoting further relocations or perceptions that there was development support for villagization. DPs further 30 indicated that there could be opportunities to engage with issues regarding villagization without being directly involved with it, and to work at policy level so as to maximize benefits for communities and mitigate harm. This would have to be balanced against the reputational risk for the organizations and the mandates of each of them. 140. During the same period, Management took PBS II Additional Financing to the Board for approval. In response to queries from several Executive Directors on the issue, Management noted that they do not have sufficient information about this program and that DAG-sponsored fact- finding missions were underway to collect information and to assess potential implications for the Bank’s Ethiopia portfolio and PBS. The Management Response notes that “[g]iven the separation between the PBS program and the villagization program, PBS 2 Additional Financing documents do not mention the latter.”138 141. The Management Response notes that the Bank, along with other Development Partners, subsequently followed up with the GoE to learn more about the objectives and modalities of the program. It adds that since the Bank is not financing the villagization program, “there is no scope to deploy the safeguards and other policies that are applicable to Bank financing instruments.”139 The Response continues that “[i]nstead the Bank and the other partners have sought to achieve influence through policy dialogue on the environmental and social development impacts of the program.”140 142. In January 2012, Human Rights Watch published a report titled Waiting here for Death, which examines the first year of the villagization program in Gambella. The report presents information gathered through interviews with over 100 residents affected by the program and finds “widespread human rights violations at all stages of the program.”141 143. In mid 2012, DPs reiterated their dilemma about how to best engage with the GoE with respect to the villagization program. In this context, Gambella’s significant development needs were assessed against the donors’ reluctance to provide direct support for villagization. DPs presented recommendations, which included follow-up assessments to compare progress and challenges over time and use of donor programs’ regular monitoring visits to assess overlap between these programs and CDP, if any. DPs also concluded that any of them who might engage in direct support to villagization should adopt a ‘do no harm’ and ‘conflict-sensitive’ approach. 144. The Bank’s Regional Operations Committee (ROC)142 meeting held on June 20, 2012, cleared the appraisal of PBS III but also discussed risks and safeguards related to the operation, among other issues. The meeting concluded that overall risks were understated and recommended                                                              138 Management Response: 10. 139 Management Response: 10. 140 Management Response: 10. 141 Human Rights Watch (2012): “Waiting here for Death” Forced Displacement and “Villagization” in Ethiopia’s Gambella Region: 2. Available online: http://www.hrw.org/sites/default/files/reports/ethiopia0112webwcover_0.pdf (accessed November 12, 2014) 142 The Regional Operations Committee (ROC) conducts corporate reviews of certain types of World Bank operations. Investment Lending operations that require corporate reviews are those with overall high risk as determined by the Operational Risk Assessment Framework (ORAF). The ROC is chaired by the Regional Vice President. Corporate reviews are tools to manage risks effectively, ensure common standards and high quality across the institution and to support learning and innovations in the institution. 31 to the project team to reevaluate the risks, possibly increase some of the risk ratings and strengthen mechanisms for monitoring and managing these risks. 145. The ROC meeting discussed the reputational risks of villagization and noted that “there is a reputational risk linked to the Government’s villagization or commune program, which contributes to the government’s approach to basic service delivery in the regions of Somali, Afar, Gambella and Beneshanghul-Gumuz.”143 The meeting concluded that this risk could not be eliminated but needs to be managed carefully throughout implementation. The meeting also noted that “the program objectives of promoting improved access to basic services overlap with the Government’s objectives for its commune program.”144 According to the meeting notes, it was “concluded that the program had triggered the appropriate safeguard policies. The risk management strategy should involve including in the oral statement to the Board that the villagization program does pose reputational risks and should describe some of the primary efforts taken to manage that risk. That will involve careful collaboration with other development partners in the broader context of the country program.”145 146. The CDP is not mentioned in the main text of the PAD for PBS III. The only reference to CDP is in the ORAF in Annex 6 of the PAD, which describes the project stakeholder risks and notes the need for the DPs to “continuously demonstrate that their funds are delivering basic service results and not leading to a reduction in political space.”146 The risk description adds that without credible, survey-based accountability mechanisms, the DPs might decide to withdraw support to maintain their institution's reputation. Villagization is mentioned as one of two issues that have emerged over the past year and a half and described as: “[a] ‘villagization’ program in the developing regions of Benishangul Gumuz, Gambella, Somali, and Afar have relocated more than 100,000 households to village clusters at the professed purpose of ensuring their access to basic services. However, the manner in which the relocation has taken place has been alleged by human rights groups to be coercive; […].”147 The reference to CDP in the ORAF points to alleged concerns about the coercive nature of the relocation program and the possible reputational risk for the Bank and other DPs.148 There is no further mention or analysis of risks relating to the concurrent implementation of the PBS and CDP. 147. Regarding the management of these risks, the ORAF notes that PBS III will “continue to strengthen monitoring and evaluation systems that monitor progress towards the MDGs and                                                              143 PBS 3, Decision Note, ROC meeting: 3. 144 ROC meeting: 3. 145 ROC meeting: 3. 146 PBS III PAD: 96. (ORAF) 147 PBS III PAD: 96. (ORAF) The adoption of the CSO law is the other issue dealt with in this section. 148 PBS III PAD: 96. (ORAF): Project Stakeholder Risks – Description: “Development Partners need to continuously demonstrate that their funds are delivering basic service results and not leading to a reduction in political space. Without credible, survey-based evidence of results and strengthened transparency and accountability mechanisms, they might decide to withdraw support to maintain their institution’s reputation. There is also a risk to the Bank’s reputation if it is seen to be supporting or being associated with mechanisms or initiatives that do not include sufficient effort by the Government to open space for citizen engagement. Two issues that have emerged over the last 1.5 year make this particularly important: (i) A “villagization” program in the developing regions of Benishangul Gumuz, Gambella, Somali, and Afar have relocated more than 100,000 households to village clusters at the professed purpose of ensuring their access to basic services. However, the manner in which the relocation has taken place has been alleged by human rights groups to be coercive; […]”  32 improved basic services.”149 The ORAF adds that the DPs “will continue to monitor their effects and ensure coordinated responses, and engage in available policy forums such as those under the Development Assistance Group (DAG). In addition, the High Level Forum has already provided a venue to discuss villagization and the CSO law. As a result of that discussion, the development partners were informed about the process of villagization […].”150 The PBS III project document does not include any further detailed discussion on villagization as Management considered villagization a country and portfolio-level risk. 148. The Board of Executive Directors approved PBS III on September 25, 2012 (one day after the Inspection Panel received the Request for Inspection), and following an extensive discussion on the issues raised in the Request. b) During Project Implementation 149. During implementation, the Bank and the DPs noted the importance of monitoring and mitigating the issues and risks that affect the PBS program performance, including CDP. The May 2013 JRIS mission stated that “[a]lthough the PBS Project pays only for recurrent costs and is not linked to the Commune Development Program (CDP), all parties agree to the importance of continuing to monitor separate programs that present perceived risks to PBS’ results.”151 The mission added that an independent assessment could be arranged if necessary, and DPs committed to accelerate the proposed PSIA agreed as part of the PBS III Equity Review.152 150. The JRIS report also referred to the Ethiopian Institution of the Ombudsman (EIO) and its key role in implementing the PBS III Grievance Redress Mechanism (GRM) sub-component. The EIO undertook an investigation related to the relocation of people in the Gambella Region. According to the JRIS report, the EIO “will accelerate the opening of a branch office in Gambella and hasten the creation of grievance desks in regional bureaus involved with the CDP. DPs will examine the modalities for ensuring appropriate oversight of the GRM component.”153 151. Finally, DPs recommended: (a) accelerating the implementation of the GRM in developing regional states; (b) summarize in one overarching document the Government’s policy on relocation and land tenure, including, resettlement, CDP and economically induced displacement; and (c) improved monitoring of the CDP, by Government and donors, including an independent assessment. 152. PBS III implementation monitoring focused on social accountability issues. The PSIA undertaken by Management explains that PBS aims at helping to strengthen existing GRM offices, including contributing to information and public awareness, delivering technical assistance to develop a common standard for GRM procedures, capacity development and training for grievance officers. PBS provides support for the opening of GRM offices in all regions and Ombudsman branch offices through dialogue and technical and financial support. The PSIA states that “[a]lthough quantitative evidence is unavailable at present, descriptive evidence from the first                                                              149 PBS III PAD: 96. (ORAF) 150 PBS III PAD: 96. (ORAF) 151 Joint Review and Implementation Support (JRIS) Mission Report (May 8-10, 2013): 1. 152 JRIS Mission Report (May 8-10, 2013): 2. 153 JRIS Mission Report (May 8-10, 2013): 1. 33 phase of the Ethiopia Social Accountability Program implies that structured feedback sessions involving citizens and service providers are strengthening citizens’ participation in pilot areas. That evidence, together with the strong guidance emerging from governance and accountability theory, would appear to favor the continued application of social accountability tools and the development of policies to sustain their use in the Ethiopian context.”154 153. Accordingly, PBS III fast-tracked the application of social accountability tools in Gambella, including the establishment of an Ombudsman’s office as part of risk management and mitigation. Picture 3:  Panel meets with Ombudsman in Gambella  2.2.4. Panel’s Analysis of Risk Assessment and Compliance a) Adequacy of the Risk Analysis undertaken during PBS III Preparation 154. The Panel notes Management’s position, expressed in the Management Response, that there is no link between PBS and CDP. Moreover, and although the Bank during the ROC meeting as well as the DPs acknowledged in some instances that there may be an overlap between PBS and CDP, they confined the resulting potential risks to the “reputational” category only. The Bank therefore determined that concerns related to villagization would be more appropriately dealt at the country and policy levels. 155. The Panel appreciates that the Bank took prompt action after being informed of the Government’s CDP by conducting fact-finding missions, assisting the Government on good practice guidelines and principles for resettlement, and raising concerns at the highest levels of Government. The Panel also understands the dilemma faced by the Bank and other DPs on how best to address development issues posed by the Government’s CDP without being linked to the allegations of human right abuses. Nevertheless, as DPs and the Bank acknowledged, there is a clear overlap between the PBS and CDP in providing basic services delivery in four developing                                                              154 PSIA: 44. 34 regions. The Panel is of the view that the risks posed by the concurrent implementation of CDP and PBS III were not limited to “reputational,” but also covered “project operating environment,” “implementing agency,” and “stakeholders” risks. 156. The Stakeholder Risk includes the concern expressed by DPs that the “space for citizen’s formal political engagement in Ethiopia is narrowing”155 and that “[w]ithout credible, survey- based evidence of results and strengthened transparency and accountability mechanisms, they might decide to withdraw support to maintain their institution’s reputation.”156 The ORAF also notes the added risk to the Bank’s reputation “if it is seen to be supporting or being associated with mechanisms or initiatives that do not include sufficient effort by the Government to open space for citizen engagement.”157 157. The description of Stakeholder Risk mentions two additional concerns: (i) the ‘villagization’ program and concerns that the “manner in which the relocation has taken place has been alleged by human rights groups to be coercive,”158 and (ii) the recent policy decision by GoE requiring CSOs to ensure no more than 30% of their cost for administration and at least 70% for operations, which DP’s think could “seriously limit CSO operations.”159 158. The overall risk rating for Stakeholder Risk is ‘High’ (before mitigation). The proposed risk management approach is three fold: - To continue strengthening PBS accountability mechanisms, in particular promoting local transparency and accountability through the Citizen’s Engagement sub-component and specifically providing support to improve grievance redress in service delivery, - To strengthen monitoring and evaluation systems including independent survey-based quality checks on results, and - To monitor the effects of PBS and related country policies on citizen space, to ensure coordinated responses by DPs and to engage in high-level dialogue with the GoE regarding matters such as villagization and the CSO law. 159. There is no mention in the PBS III ORAF of any specific Stakeholder Risks arising from the implementation of CDP in the four regions in which PBS III operates. The measures for managing Stakeholder Risk relate mainly to the design of the Citizen Engagement sub-component and the introduction of independent survey-based quality assurance on results. The Panel notes that the effectiveness of these risk management measures depends almost entirely on GRMs being available and utilized. If this is not the case, as claimed by the Requesters and confirmed by several DAG reports, these measures should have been reviewed and if necessary, adjusted (as per the Guidance) to ensure that grievances were received and appropriately addressed. 160. The Panel also notes that the PBS III ORAF contains no assessment of Operating Environment Risk. Operating Environment Risks are the types of risk that shape the context in which PDOs are being pursued and could eventually impact their achievement. The changes to the                                                              155 PBS III PAD: 96. (ORAF) 156 PBS III PAD: 96. (ORAF) 157 PBS III PAD: 96. (ORAF) 158 PBS III PAD: 96. (ORAF) 159 PBS III PAD: 97. (ORAF) 35 CSO law and the roll-out of the villagization program should be considered key Operating Environment Risks deserving assessment and monitoring. Given the sheer scope and complexity of PBS, the Panel considers this to be a major oversight. 161. Bank Management did not consider the possibility that the concurrent roll-out of the CDP would affect the operating context for PBS III, nor that CDP could fundamentally change the critical path for the PDO’s in the relevant regions. Instead, Bank Management seems to have assured itself that because the Project is focused on improving service delivery through the flow and management of block grant funds, anything outside of this does not qualify as a realistic or tangible risk to the Project. 162. Implementing Agency Risks focus on the problems of capacity, including high staff turnover at the local level, and the need to strengthen local accountability focusing on the weak capacity in procurement and financial management, especially at the woreda level. Given that the woreda administration is central to the design of PBS delivery, the mitigation measures lack concrete steps and time frames. Given also that the implementing agency for PBS and the CDP are one and the same at the woreda level, a clearer articulation of the possible risk of inadequate management capacity for concurrent implementation, and weak oversight arising from the dual mandates posed by PBS and the CDP would have been appropriate. 163. In the PBS III PAD, Project Risks are focused mainly on risks arising from complexity and weak citizen engagement at the local level, together with the risk to sustainability arising from the continued dependence on donors to fully fund the block grant mechanism. Social and environmental risks were not adequately assessed because of PBS III’s focus on recurrent costs and capacity building of the local government staff and system.160 Management’s decision not to apply the safeguard policies to the operation is discussed in Chapter 3. 164. The Panel notes that some form of ex-ante social assessment would have significantly improved the quality and effectiveness of the risk assessment and may have resulted in the identification of a series of key risks to consider during implementation. The sensitivities that were clearly building around the Project, even if unverified at the point of appraisal, demanded the Project to include measures to address any potential increase in risk or negative social impact. The Panel recognizes that additional measures, such as the fast-tracking of social accountability mechanisms to Gambella were identified following the receipt of the Request by the Inspection Panel, following the Board meeting approving PBS III, and during Project implementation. 165. The Panel’s analysis in Chapter 3 highlights the potential impact of CDP through its resettlement activities on the PDOs of PBS in the agricultural sector in Gambella, pointing to an operational interface that might have been taken into account in relevant risk sections of the ORAF i.e. Implementing Agency or Project Level Risk. 166. The Panel notes that during preparation of PBS III, a thorough assessment of Operating Environment Risk could have systematically reviewed:                                                              160 PBS III PAD: 96. (ORAF) 36 - Country-level risks, not just macroeconomic risks, but also political, institutional, and social risks linked to shifts in the authorizing environment for PBS, i.e., changes in the policy space for citizen engagement and the priority assigned by the GoE to the CDP. - Specific sector and geographic risks arising from the concurrent roll-out of PBS III and the CDP. The fact that the Bank was made aware of the planned villagization program around the time of the PBS III appraisal should have triggered revised Risk Assessments for those locations where the overlap between PBS and CDP was deemed greatest, and a consequent adjustment of the Risk Assessment at the Implementing Agency level. These should have included risks associated with weak capacity for effective oversight of PBS vs. CDP expenditures; inadequate staffing increasing the risk that staff fully funded under PBS would divert time to the CDP, and weak governance and accountability for the outputs of the two programs. 167. As noted above, a social assessment would have been justified both to verify the robustness of the results chain linking the block grant mechanism to PBS results, and to assess any changes in the local governance context (which would, it could be safely assumed, have identified the potential overlap with the CDP) that might put future results achievements at risk. Such an exercise might have identified a series of risks to monitor during implementation, or an alternative set of risk management measures designed to avoid potential negative impacts from the CDP. 168. The Panel notes that the PBS III ORAF does not meet the expectations set out in the Bank’s Guidance Note to “look systematically, holistically, in an integrated manner and in real time, at risks to achieving project development objectives (PDOs)”161… “[…] using the risk assessment during implementation to regularly scan the full spectrum of possible risks helps teams move quickly to identify and address emerging issues, including unanticipated risks.”162 Considering the magnitude of the operation, the nature of block grant financing and the overlapping implementation between PBS III and CDP, the Panel finds that Management did not carry out the required full risk analysis, nor were its mitigation measures adequate to manage the concurrent roll-out of the villagization program in four PBS III regions. The Panel finds that Management’s approach did not meet the standards of a systematic or holistic assessment of risks, as called for in the ORAF Guidance, which is aimed among other objectives at identifying adequate risk management measures for affected communities. The Panel finds these omissions in non-compliance with OMS 2.20 on Project Appraisal. b) Adequacy of the Risk Analysis undertaken during PBS III Implementation 169. During the Executive Board consideration of PBS III (September 25, 2012), there were extensive discussions among the Executive Directors and Management related to the issues raised by the Requesters. In its Response, Management states that reacting to concerns raised by Executive Directors during their discussion of PBS III, Management committed to undertake additional analysis “through a Poverty and Social Impact Analysis, which would, amongst other things, aim to verify further the observed experience under PBS that increasing funds for the delivery of services across Ethiopia and increasing transparency and accountability through                                                              161 Guidance Note on the ORAF (2011): 2. 162 Guidance Note on the ORAF (2011): 2. 37 citizens’ engagement provides proportionately greater returns to marginal areas and vulnerable groups.”163 170. PBS III implementation paid particular attention to the establishment of GRMs and the implementation of the social accountability component of the program. Management fast-tracked the implementation of the Project’s social accountability component and GRM in Gambella. A Regional Ombudsman’s Office was created by the EIO in the Gambella Region. During its investigation visit, the Panel team met with the Ombudsman of the EIO’s Gambella branch. The Ombudsman noted to the Panel team that the office was established officially in December 2013, but that he had started as the Ombudsman in Gambella in September 2013. 171. During its investigation visit, the Panel team also met with the representative of VECOD, the NGO responsible for the implementation of the social accountability component of PBS III in Gambella. The representative noted that the activities started in January 2013, and that he joined in August 2013. The representative explained VECOD’s objective as providing grass-root level accountability for PBS sectors and to teach people their rights. The Panel understands that VECOD works in 5 woredas, including Anuak, Nuer, Mojang and mixed woredas, targeting 3 kebeles per woreda. VECOD representative noted to the Panel team that the communities mostly raise health and water issues. The VECOD representative noted that the NGO contract to provide social accountability services is ending in December 2014, and it is not clear what will happen after this date. 172. The PSIA states that “[t]his study attempts to determine the extent to which spending at the woreda level on basic services is associated with key policy outputs and human outcomes. […] A parallel objective of the study is to assess the incidence of these expenditures by wealth quintile, in line with the World Bank’s objective of achieving shared growth by reaching the bottom 40 percent. A final objective is to investigate whether the allocation of woreda-level block grants reflects the constitutional objective of providing additional resources to historically marginalized populations.”164 Accordingly, the Panel appreciates that the PSIA analysis has shown that “[s]ome historically disadvantaged areas are significantly favored under the current spending framework […] [and] [f]our majority-Anyiwak woredas are noteworthy for receiving more public resources than all other woredas in the nation.”165 The Panel notes, however, that while higher allocation of resources is positive, it does not automatically translate into better quality and coverage of services, nor does it address the issue of interface between PBS and CDP and their mutual impact. 173. In its Eligibility Report, the Panel focused specifically on the PSIA, stating in its Recommendation that the “Panel's investigation will report on steps and actions taken by Management during the course of the investigation to address issues of compliance and the concerns raised by the Requesters, including, if feasible, the outcomes of the proposed Poverty and Social Impact Analysis.”166 The Panel thus notes that while the PSIA showed favorable allocation of resources to Gambella, it did not address the issues of concerns raised by the Requesters.                                                              163 Management Response: 21. 164 PSIA: xvii. 165 PSIA: 43. 166 Report and Recommendation: 21f. 38 174. The Panel recognizes Management’s efforts to strengthen social accountability during PBS III implementation. As noted in the Implementation Completion and Results Report (ICR) of PBS II, there were delays in the implementation of the social accountability activities in PBS II.167 The Panel welcomes the establishment of the Ombudsman office in Gambella and appreciates that the social accountability and grievance mechanism components have been fast-tracked in Gambella. It is hoped that moving forward, efforts for citizen engagement and social accountability and their effective implementation, which are essential for PBS service delivery, will be closely coordinated and monitored in a mutually reinforcing way. 175. The Panel understands and appreciates that, more recently, Management is currently supporting a study on the interface between World Bank-funded and nationally-funded programs to ensure coherence in their implementation.168 The Panel welcomes this development and expects that the study will lead to necessary adjustments in project implementation, when called for. 176. Notwithstanding these developments, it is the view of the Panel that the lack of recognition and analysis during appraisal of the operational interface between PBS III and CDP, as required by the ORAF and described above, meant that the resulting risks were not adequately taken into account or properly managed and mitigated during PBS III implementation. Picture 4:  Anuak Village in Gambella                                                               167 For more information see: PBS II Implementation Completion and Results Report (2013): 18, 34, 46. 168 Ethiopia – Country Partnership Strategy Progress Report (2014): 36. 39 Conclusion 177. It is the view of the Panel that while Management was aware of CDP being implemented in four regions, the full spectrum of risks associated with its concurrent roll-out with PBS was not appraised. This inadequacy meant that mitigation measures were focused largely on country level dialogue through the DAG (with no specificity on timeframe or success indicators), and putting in place a social accountability program, most notably through GRMs. These measures, though effective to strengthen transparency and accountability in service delivery, did not address risks from CDP implementation. Furthermore, the risk assessment was not adequately adjusted during implementation to take account of the concurrent implementation of CDP in the relevant regions. The Panel recognizes that Management is now taking measures to study the interface between Bank-funded and nationally funded programs, with a view to ensuring coherence among the programs. 40 Chapter 3: Application of Safeguard Policies and Distribution of Services in Gambella 3.1. Application of the Safeguard Polices to PBS III 178. This Chapter analyzes the Requesters’ allegations that safeguard policies should have been applied to PBS III, in particular the policies on Indigenous Peoples and on Involuntary Resettlement, given that its implementation overlapped with the CDP program. The Chapter includes a discussion on the harms alleged by the Requesters regarding basic service delivery and livelihood conditions in resettled villages, and presents the Panel’s findings and observations on whether these alleged harms resulted from Bank’s non-compliance with its policies and procedures. 3.1.1. Application of Safeguards to Projects Focused on Recurrent Expenditures   179. The Requesters believe that the social safeguard policies on Indigenous Peoples (OP 4.10) and Involuntary Resettlement (OP 4.12) should have been applied to the PBS Program. They refer to the Project Appraisal Document (PAD) of PBS II169 which claims that “the Bank decided that given that Subprogram A, the main component of PBS, only supports recurrent expenditures, social and environmental safeguard policies, including OP 4.10 and OP 4.12, are not triggered.”170 The Requesters argue, however, that “neither the instrument utilized for PBS nor the decision to fund only recurrent expenditures negate the Bank’s obligation to trigger and comply with relevant safeguard policies;”171 rather, the high level of discretion allowed by the financing instrument requires strong Bank oversight.172 180. In its Response, “Management disagrees with the Requesters’ assertion that the Bank’s safeguard policies should have been triggered in PBS 3 to address potential environmental and social risks related to villagization because […] the PBS project is not linked to the villagization program.”173 Management argues that therefore “there is no scope or justification to apply the Bank’s safeguards policies to that program as demanded by the Requesters.”174 The Response adds that the decision not to apply safeguard policies in PBS III is consistent with the overall approach for the three phases of the PBS program “as well as for several other Bank investment operations involving recurrent costs.”175 Management notes that “the application of safeguards is calibrated in view of the potential environmental and social impacts”176 of the proposed operation and, in this context, PBS III’s impacts were evaluated as justifying the environmental category rating of ‘C.’177                                                              169 Request for Inspection: 15. (reference to PBS II PAD: 9.) 170 Request for Inspection: 15. 171 Request for Inspection: 15. 172 Request for Inspection: 15. 173 Management Response: ix. 174 Management Response: 15. 175 Management Response: 16. 176 Management Response: 16. 177 Management Response: 16. 41 181. The Panel addresses the general issue on whether there is scope for the application of safeguard policies in projects providing for recurrent expenditures. The Panel notes that Bank policies do not exclude the application of safeguard policies in Investment Lending (IL) operations providing only for recurrent expenditures. The Panel finds that to the extent that one or more safeguard policies are found to be relevant to the areas of operation of the proposed project under the World Bank Operational Policies and Procedures, Management should trigger and apply them. The Panel does not concur with Management’s assertion that recurrent expenditures, a priori, do not have any environmental and social impacts, as further elaborated below. 3.1.2. The Involuntary Resettlement Policy 182. The Requesters maintain that the Policy on Involuntary Resettlement (OP 4.12) should have been applied to the PBS Program. In their view, had OP 4.12 been triggered, the PBS Program would have required adequate safeguards for the resettlement of people under the CDP and would have ensured processes in respect of human rights. They argue that it was the responsibility of the Bank to take measures that would make sure that people were relocating voluntarily and were adequately informed, consulted, compensated and received resettlement assistance in accordance with the Bank policy. As noted in other parts of this Report, they believe that the villagization program is “directly and significantly related to PBS, and indeed is the means by which the regional government officially aimed to achieve PBS objectives, using PBS funds.178 183. Management states that there is “no scope” for application of the safeguard policies to the CDP, as the Bank does not finance it and PBS III does not involve any involuntary taking of land, thus arguing that the decision not to trigger the policy on involuntary resettlement (OP 4.12) was correct. Management further notes that OP 4.12 is not applicable to PBS III because the project supports only recurrent expenditures and does not involve taking of land or restriction of access. In addition, PBS III does not depend on villagization to achieve its objectives and therefore, even if the resettlement under the program was involuntary, such Government activities would fall outside of the scope of the policies applicable to the Bank supported project.179 Management adds that its position with respect to the applicability of OP 4.12 has been consistent through the three PBS phases and other Bank operations involving recurrent costs. 184. OP 4.12 applies to Involuntary Resettlement in development projects as it is recognized that “if unmitigated, [involuntary resettlement] often gives rise to severe economic, social and environmental risks”180 and general impoverishment of the people affected by the relocation. The policy covers “direct economic and social impacts”181 that are caused by an involuntary taking of land as a result of a Bank-financed project. The policy applies to all components of the project that cause involuntary resettlement, regardless of the source of financing. It also applies to other activities that result in involuntary resettlement that, “in the judgment of the Bank are a) directly and significantly related to the Bank-assisted project, b) necessary to achieve its objectives as set forth in project documents; and c) carried out, or planned to be carried out, contemporaneously with the project.”182 In other words, OP 4.12 applies to involuntary resettlement that is the direct                                                              178 Request for Inspection: 18. 179 Management Response: ix. 180 OP 4.12, para. 1. 181 OP 4.12, para. 3. 182 OP 4.12, para. 4. 42 result of activities under the Bank-financed project, or activities that are directly and significantly related to the design and the performance of the Bank project, and is necessary to achieve the objectives of the Bank Project. 185. Taking into account that activities financed under PBS III do not require involuntary taking of land, the question at hand is whether the Government-financed CDP, which is a resettlement program, is directly and significantly related to the Bank-assisted project, necessary to achieve its objectives and carried out contemporaneously with the project as argued by the Requesters. Picture 5:  Panel meets with Anuak Village in Gambella  186. The Panel notes that Management analyzed whether OP 4.12 would be applicable in the context of World Bank-funded programs. A joint World Bank-Finland mission to Benishangul- Gumuz (BG) was carried out in 2011 to understand the CDP program, its design and implementation, and to determine whether there was a linkage with other donors’ programs in BG, but also in the other states where CDP was implemented. Bank staff also assessed the applicability of OP 4.12. According to internal Bank documents reviewed by the Panel, the Bank mission concluded that Bank-supported projects in the region may be carried out contemporaneously with the CDP and could provide indirect support to it, albeit “coincidentally,” in Bank staff’s words. The mission also determined that the criteria of OP 4.12 discussed above (direct, significantly related and necessary to achieve objectives) were not met. 187. The Panel notes that the PBS program is a nationwide program which was initiated in 2006, and is expected to close in 2018. The CDP program began in Gambella in 2010 and terminated in 2013. Thus, while CDP in Gambella was contemporaneous with PBS, the latter was being implemented four years prior to the commencement of CDP, and is continuing after the conclusion of CDP. Furthermore, PBS was designed to cover the entire population of Gambella, irrespective of whether they relocated under CDP. Thus, the Panel does not consider CDP a necessary activity to achieve the objectives of PBS III. 188. The Panel finds that paragraph 4 of OP 4.12 is not applicable and that Management acted consistently with the provisions of OP/BP 4.12 by not triggering it. 43 3.1.3. The Indigenous Peoples Policy a) Requester’s Claims 189. With regard to the safeguard policy on Indigenous Peoples (OP 4.10), the Requesters claim that “the Anuak fall squarely into any definition of ‘indigenous peoples’ and possess the precise characteristics described in OP 4.10”183 and thus the policy should have been applied. The Requesters go on to argue that “it was impossible for PBS not to affect Indigenous Peoples. Yet not a single reference is made to Indigenous Peoples in project appraisal documentation.”184 Furthermore, they argue that “[i]t appears from publically available information that the Bank did not conduct a robust screening to identify whether Indigenous Peoples are present in, or have collective attachment to, the project area, which in the case of PBS is every region in the country.”185 Also, the Requesters state that no social assessment evaluating potential adverse effects on Indigenous Peoples was conducted and no measures taken to address potential effects and to ensure culturally appropriate project benefits.186 The Requesters claim that if the Policy had been applied “PBS funds could not have been used to implement any aspects of villagization as a means to improve access to basic services that required Indigenous People, including the Anuak, to move away from their ancestral lands”187. b) Management Response 190. With regard to the safeguard policy on Indigenous Peoples, Management explains that the GoE “has had concerns that if applied without due care the policy would be inconsistent with the Ethiopian Constitution and might also create tensions between ethnic groups rather than reduce them.”188 According to Management, the GoE has been concerned about the application of OP 4.10 to IDA-lending to Ethiopia because it worries about “singling out ethnic groups for distinct treatment, which is inconsistent with certain principles of the Ethiopian Constitution, in particular Article 39(5) which defines the concept of ‘Nation, Nationality or People’”189 and grants each group equal protection. The concept of ‘Nation, Nationality or People’ is described in similar terms to those of the Bank’s policy on Indigenous Peoples. Thus, the GoE argues that, based on the county’s Constitution, all people in Ethiopia are indigenous according to the Policy.190 c) Indigenous Peoples in Ethiopia’s Constitution 191. Article 8 of Ethiopia’s constitution reads that “1. [a]ll sovereign power resides in the Nations, Nationalities and Peoples of Ethiopia. 2. This Constitution is an expression of their sovereignty.”191 A Nation, Nationality or People is defined in the Constitution as “a group of people who have or share a large measure of a common culture or similar customs, mutual                                                              183 Request for Inspection: 15f. 184 Request for Inspection: 16. 185 Request for Inspection: 16. 186 Request for Inspection: 16. 187 Request for Inspection: 16. 188 Management Response: 37. 189 Management Response: 37. 190 Management Response: 16f. 191 Constitution of the Federal Democratic Republic of Ethiopia (1995). Available online: http://www.refworld.org/docid/3ae6b5a84.html (accessed July 5, 2014) 44 intelligibility of language, belief in a common or related identities, a common psychological make- up, and who inhabit an identifiable predominantly contiguous territory.”192 192. The Constitution guarantees every Nation, Nationality and People an unconditional right to self-determination, including the right to secession, the right to speak, write, and develop an own language, to develop and promote its culture and to preserve its history.193 Article 41(3) states that “[e]very Ethiopian national has the right to equal access to publicly funded social services”194 d) Background and Chronology relating to the Application of OP 4.10 in Ethiopia 193. In its Response, Management indicates that OP 4.10 was not applied to Bank operations in Ethiopia before 2013, including the PBS project, because of the GoE’s concern that the policy was not compatible with the Constitution and country context.195 The Response notes that discussions with the GoE were ongoing since mid-2009 about how to apply the Policy for future Bank operations in Ethiopia, and a joint World Bank-GoE workshop, aimed at enhancing the government’s familiarity with OP 4.10, was held in 2011. The issue was further discussed with the Ethiopian delegation at the WB-IMF Annual Meetings in 2011, and a proposal for the application of the policy was sent to the GoE. This delegation reiterated the difficulties of applying the policy but noted that a joint resolution between the government and the Bank was needed. The issue was then raised again during the Annual Meetings in 2012, this time in the specific context of the present Request (which had been registered by that time), and the GoE reconfirmed its commitment to respond to the concerns. In mid-2012, the government and the Bank agreed on the Terms of Reference for a screening of the ethnic groups in five regions of Ethiopia. 196 194. In early 2013, a World Bank-commissioned field-based screening process was carried out to assess both the relevance and appropriateness of applying OP 4.10 in the Ethiopian context. Sixty four Nations, Nationalities and Peoples in five regions (Afar, Oromia, Somali, SNNPR and Gambella) were screened against the Bank policy’s reference to “distinct, vulnerable, social, and cultural groups” and the four elements of paragraph four of OP 4.10 which define ‘indigenous peoples’ to which the policy is applied. However, most groups fulfilled all four elements. Thus, the research team decided to additionally use the government’s four categories (regional and intra- regional disparity in human development indicators, historical marginality of pastoralist areas and national minorities) for their screening process. They then identified 34 groups as vulnerable, for which they recommended the application of OP 4.10. The Anuak are among those 34 groups and meet all screening criteria to a large degree.197 195. Staff informed the Panel that during the preparation of PBS III, there was an understanding that OP 4.10 would not be applied, given the position of the GoE, but, “where necessary for certain groups, due diligence would be conducted.” The Panel was also informed that Bank’s social protection experts were of the opinion that the Project, as designed, protected indigenous peoples because “everybody was treated the same.” Staff also noted that a vulnerability assessment could have been carried out as part of due diligence.                                                              192 Constitution, Article 39(5). 193 Constitution, Article 39. 194 Constitution, Article 41(3). 195 Management Response: 16. 196 Management Response: 16f. 197 This study was available to the Panel for review but it is at the moment confidential and not publicly available. 45 196. The Panel notes that neither Project preparation documents nor Management’s Response mention specific due diligence actions carried out during Project preparation with respect to certain groups meeting the criteria of OP 4.10. The Response describes the concerns of the Government in relation to the application of OP 4.10 and the “uncertainty as to its compatibility with the country context”198, but also indicates that, in general, the application of safeguards is calibrated with the Project’s social and environmental impacts, and PBS III, as designed, was considered a category C project, likely to have minimal or no environmental and social impacts.199 197. The Panel also notes that Management’s Response refers to the concept of “functional equivalence with the policy” on Indigenous Peoples as applicable to projects that would normally trigger OP 4.10. e) The Concept of Functional Equivalence 198. In early 2012, prior to the start of the negotiations for the Productive Safety Net Project (PSNP APL III) Additional Financing, discussions about the application of OP 4.10 in Ethiopia had taken place, based on which senior management approved a course of action referred to as functional equivalence.200 The Memorandum of the President (MOP) on a proposed additional credit to Ethiopia for the PSNP APL III, dated March 1, 2012, outlines the discussion with the GoE up until the point when the Bank sent its proposal for the application of OP 4.10 in 2011, stating: “if agreement is reached on appropriate application of this policy in the Ethiopia portfolio, but in any event starting with operations approved after December 2012 […], it will be applied to the extent that it is found to be relevant to the areas of operation of the proposed projects. Relevant operations presented to the Board in the meantime will endeavor to contain features that approach functional equivalence with the policy even when it is not formally triggered.”201 199. The Management Response mentions the concept of functional equivalence by referring to the Board discussion of the Country Partnership Strategy (CPS) for Ethiopia.202 The CPS was adopted in August 2012, a few months after the issuance of the PSNP MOP. The CPS includes a footnote explaining that a note, which appears in the MOP of every Board package for Ethiopia since February 2012, has summarized the situation as follows: “(a) dialogue between GoE and the Bank on OP 4.10 is ongoing, (b) when agreement is reached, but in any event starting with operations considered by the Board after December 2012, the policy would be applied to the extent that it is found to be relevant to the areas of operation of the proposed projects; and (c) relevant operations presented to the Board in the meantime will endeavor to contain features that approach functional equivalence with the policy even when it is not formally triggered. In lieu of agreement with GoE on application of OP 4.10, in some projects, task teams have been able to achieve much                                                              198 Management Response: 16. 199 OP 4.01, para 8. 200 Confidential Internal Memorandum. The Panel was informed during interviews with staff that an earlier memo dated 2009 had mentioned functional equivalence, but had not been issued because of the difficult relationship with GoE with regards to indigenous peoples at that time. 201 Memorandum and Recommendation of the President of the International Development Association to the Executive Directors on a Proposed Additional Credit to Democratic Republic of Ethiopia for the Productive Safety Net Project (PSNP APL III): 2. 202 Management Response: 17. 46 of the intent of OP 4.10 without triggering the policy (through the Environmental Assessment or Involuntary Resettlement policies and procedures).”203  200. According to Bank staff, the core of OP 4.10 is reflected in the concept of functional equivalence, which is based on five principles: free, prior and informed consultation leading to broad community support, mitigation of adverse impacts on people who would trigger the policy, culturally appropriate benefit sharing, grievance redress mechanisms, and monitoring and evaluation of outcomes for indigenous peoples. 201. The Panel notes that from January 2013 and when relevant, OP 4.10 was triggered for proposed projects in Ethiopia to be approved by the Board. The Panel also draws attention to the PAD of the Productive Safety Net Phase 4 which states: “OP 4.10 Indigenous Peoples is triggered as it is determined that the physical and sociocultural characteristics of the sites where subproject activities could be implemented, and the people living in these sites, meet the policy requirements. The decision to trigger the policy is also based on the Ethiopian Constitution, which recognizes the presence of different sociocultural groups, including historically disadvantaged or underserved communities, as well as their rights to their identity, culture, language, customary livelihoods and socio-economic equity.”204 (Emphasis added) f) The Panel’s Analysis 202. As noted above, the Bank commissioned-screening study determined that the Anuak people meet the criteria set forth in OP 4.10 and can be considered indigenous peoples under the Bank policy. While acknowledging the difficult context in which PBS III was prepared, the concerns expressed by the GoE regarding the definition of Indigenous Peoples, and the application of the related policy in the Ethiopian context, the Panel also notes Management’s indication that based on an agreement with the GoE, OP 4.10 would be applied to operations considered by the Board after December 2012. As noted above, this is consistent with the Ethiopian Constitution. Prior to December 2012, Bank operations would follow the approach known as functional equivalence with the policy. 203. Bank staff informed the Panel that the concept of functional equivalence was discussed during Project preparation and that the Bank applied the “spirit of functional equivalence” to PBS III. It was explained to the Panel that at that time, the screening study had not been completed, and staff felt that the GoE would not have accepted a report on Indigenous Peoples; furthermore, waiting to apply the policy would have delayed the Project for one year. Staff explained to the Panel, however, that the PBS III social accountability mechanisms at the community level had the                                                              203 Country Partnership Strategy for Ethiopia (2012): 47 (footnote 87). Bank staff informed the Panel that the GoE changed its policy on indigenous peoples after the Management Response had been submitted: in January 2013, Management had announced the application of the policy to MoFED and had held back 1.3 million USD of lending until the government agreed to the policy application in June 2013. Since then, the Bank applies OP 4.10 in Ethiopia. Staff informed the Panel that the GoE agreed that Social Assessments and Action Plans in PADs or ESMPs could be public; however, the GoE did not want the term Indigenous Peoples to appear in PADs, but rather favors calling them ‘people who trigger OP 4.10’. According to Bank staff, there are two options for applying the policy: one, to apply it according to the letter of the policy; two, to apply it to groups who also meet a vulnerability criterion, which is an ad- hoc criterion not explicitly mentioned in the policy, which creates difficulties and could lead to conflict with excluded groups. The Panel was told that Management now reviews the policy application with the government on a project- by-project basis. The General Education Quality Improvement Project II and the Sustainable Land Management Project II applied OP 4.10 in Ethiopia. 204 PSNP 4 PAD: 26. 47 five core principles of OP 4.10 embedded in their procedures. In this regard, staff noted that the social accountability component, community consultations and grievance redress mechanisms are among the measures the Project employs to “cover the basis of OP 4.10”. Management notes in the Response that the preparation of PBS III “has benefited from a number of studies and evaluations of the impact of PBS, including of opportunities under the program to promote improved engagement, voice and services for vulnerable groups […] [and] includes numerous elements to ensure that the operation promotes fairness, equity and transparency in service delivery results and strengthens citizens’ engagement and social accountability.205” 204. The Panel notes that the PAD for PBS III, the PID, and other preparation documents the Panel has reviewed, do not mention the concept of functional equivalence with the policy, as discussed or applied to PBS III. The PAD includes a general statement that “as Project financing is limited to recurrent expenditures for basic services and does not include financing for capital investment or civil works, World Bank safeguard policies are not triggered.”206 The PAD does not include a discussion of Project impacts on vulnerable groups, as Management Response suggests. 205. In this regard, the Management Response indicates the intention to carry out a PSIA aimed at verifying that under PBS, “increasing funds for the delivery of services across Ethiopia provides proportionately greater returns to marginal areas and vulnerable groups.” 207 The Terms of Reference (TORs) for the PSIA specify that while the Project is designed to provide fair and efficient service delivery through a transparent and equitable distribution of resources to local administrations, PBS does not target directly the needs of vulnerable groups. The PSIA would then ascertain whether under PBS, there is inequitable access to services between different ethnicities.208 206. The Panel notes that the Bank policy on Indigenous People states that: “The Bank recognizes that the identities and cultures of Indigenous Peoples are inextricably linked to the lands on which they live and the natural resources on which they depend. These distinct circumstances expose Indigenous Peoples to different types of risks and levels of impacts from development projects, including loss of identity, culture, and customary livelihoods, as well as exposure to disease. As social groups with identities that are often distinct from dominant groups in their national societies, Indigenous Peoples are frequently among the most marginalized and vulnerable segments of the population. As a result, their economic, social, and legal status often limits their capacity to defend their interests in and rights to lands, territories, and other productive resources, and/or restricts their ability to participate in and benefit from development.”209 207. The PBS III preparation took place when the agreement with the GoE to adopt an approach of functional equivalence with OP 4.10 in project areas where indigenous peoples are present was being applied. The Panel reviewed Project documents, Management’s Response and discussed with various stakeholders, but it did not find indications, prior to the interviews with staff that Management applied the functional equivalence approach to PBS III, although many groups affected by the Project qualify as indigenous peoples under the Bank’s policy. In contrast, Project                                                              205 Management Response: 17 (footnote 8). 206 PBS III PAD: 34. 207 Management Response: 18. 208 PSIA TORs (PSIA TRUST FUND - Application Form for Africa Region – FY13) 209 OP 4.10, para. 2. 48 preparation documents illustrate Management’s position that safeguard policies in general do not apply to projects providing exclusively for recurrent expenditures, such as PBS III. 208. The Panel finds that, barring the triggering of OP 4.10, Management should have adopted the “functional equivalence” approach in the design of PBS III. The relevance and need for adopting such approach to the Anuak is strictly related to their distinct characteristics that the Bank policy OP 4.10 well lays out. While indigenous peoples, in general, may be characterized as vulnerable groups, the reasons for their vulnerability differ from that of other groups, e.g. street children or the disabled, and are historically rooted in their attachment to ancestral land and territories. The Panel notes that livelihoods, well-being and access to basic services, which are closely tied to the Anuak’s access to land and natural resources was not taken into account in the design of PBS III, in non-compliance with OP 4.10. 3.2. Delivery of Services and Livelihoods 3.2.1. Summary of the Requesters’ Claims and Management Response a) Requesters’ Claims 209. The Requesters claim that, through the GoE’s CDP Program, “[t]he Anuak have been relocated to infertile land, which is unsuitable for farming, and forced to build new villages there”210, which was carried out “under the pretext of providing better services and improving the livelihoods of the communities”211. The Requesters argue that at the new locations “they found not only unfertile land, but also no schools, clinics, wells or other basic services.” 212 They explain that the Anuak had been forced to leave their crops right before the harvest season and did not receive any food assistance from the government. Consequently, many families who had moved faced hunger and some vulnerable people, including children, died from starvation.213 210. The Annex to the Request for Inspection elaborates that the “promised basic services and facilities at the new sites were either not provided or were not operational and there was little access to food or land suitable for farming at the time that they were forced to move.”214 b) Management Response 211. Management states in its Response that the Requesters “are not able to demonstrate that the Promotion of Basic Services program Phase III (PBS 3) has harmed them or is likely to harm them”215 and that the Bank is not financing CDP. 212. With regard to the alleged inadequacy of service delivery and food insecurity, Management states that “DAG-sponsored missions did find that planning had often been inadequate, relocation was too rapid, the sequencing of site improvements was poor and had inadequate finance, and                                                              210 Request for Inspection: 1. 211 Request for Inspection: 1. 212 Request for Inspection: 1. 213 Request for Inspection: 1. 214 Request for Inspection: 8. 215 Management Response: 22. 49 implementation capacity was weak.”216 According to the Management Response, the GoE accepted feedback by the Bank and other members of the DAG about implementation issues and acknowledged that the implementation of CDP could be improved. A June 2012 visit (at the end of the second year of CDP implementation) suggested improvements of the situation on six out of eight key measures compared to February 2011. Management explains in its Response that “[f]ood insecurity and malnutrition are major concerns in certain, localized areas across Ethiopia, sometimes as an acute problem requiring humanitarian relief, and sometimes as a chronic development challenge”217; therefore, the government and DPs have implemented a variety of policies and programs for rural economic development and food security.218 3.2.2. The Panel’s Analysis on Agriculture and Livelihoods 213. In Chapter 2, the Panel provided information on basic services in the villages visited, supplemented by additional information from existing reports. This section deepens the analysis on services under agriculture. Attention to this sector is important because the Government looked upon the CDP in Gambella as a type of agriculture project assigned to the Gambella agriculture department. The Villagization Program Action Plan for Gambella indicates that that “although the Gambella Region is endowed with natural resources suitable to expand agricultural production it is one of food unsecured areas of the country. Traditional/cut & burn farming practices; scattered settlement; and river side settlement are the causes for the vulnerability among others. To alleviate the problem the regional government has developed strategy on villagization program on voluntarily, participatory and upon thorough investigation.”219 214. It is the Panel’s view that the agriculture sector and livelihood issues have been particularly affected by the CDP’s resettlement operations, and as a result, the quality and effectiveness of service delivery in the agriculture sector under PBS were adversely affected. The Panel finds that, by not considering such risks during appraisal, Management was not attuned to developments and specific results during implementation for this sector in Gambella. The Panel also believes that delivery of agriculture services for the Anuak as indigenous peoples calls for particular attention to their special circumstances and livelihood needs. The paragraphs below illustrate these points. 215. The Request referred to the “inhumane conditions at the new villages including a lack of access to food and livelihood opportunities, in some cases leading to starvation”220. Management, in its Response, stated that it “continues to have a close regard for environmental and social issues as they relate to the achievement of the project’s development objective”221 and argues that PBS III supports strengthening of environmental and social assessments at the local level through assessing capacity, targeting capacity building and encouraging best practices in planning and implementation. 216. The PBS program supports the agriculture sector by paying the salary of trained development agents, intended to provide agricultural extension services for crops, livestock and natural resource management. According to the PAD for PBS III, one of the underlying rationales                                                              216 Management Response: 27. 217 Management Response: 27. 218 Management Response: 26f. 219 Villagization Program Action Plan: 1. 220 Request for Inspection: 7f. 221 Management Response: 18. 50 of this is that a “large, trained group of nationally deployed public sector agriculture agents is a potentially valuable asset to help deliver results in agricultural productivity.”222 According to the PAD, the progress towards attaining the PBS Project Development Objectives in the agriculture sector are measured by the number of “agriculture sector beneficiaries (direct beneficiaries) of the project;”223 the number of specialized development agents in villages is an intermediate results indicator, while the achievement of higher level objectives of PBS III is an increased agricultural productivity.224 217. Based on interviews with villagers in the field, the Panel found that availability of land and land-related issues (agricultural production and food) are of the greatest concern to people affected by the CDP program. The Panel observed a general trend in many villages where land was provided to resettled people, but they could not work it because the land had not been cleared and little or no support was provided for clearing it. Thus, plots of land at the resettlement site stay unused while a large number of people keep going back to the fields they used to work before moving, in many cases located at several hours walking distance away. Some have moved back to their original villages, as already described in Chapter 2. Some Requesters complained that the new land was dry and the soil not suited for their traditional staple crops. In another village, people told the Panel team that only some are able to work on their old fields while for others, the distance makes this impossible; the former then share their produce with the other families, who would otherwise have nothing to eat.225 While in two villages the Panel heard from local level officials that food has become more plentiful after relocation, in another village the Panel was told that “they need to go very far to get fire woods, poles and grasses for construction, which has made their livelihood difficult.” 218. The considerations above point to the grave impact from resettlement on villagers’ livelihood and their food security. As DPs noted, livelihood options were limited and in some cases, the lack of access to fishing and riverside mango trees worsened the situation. In 2012, DPs noted that the scale and speed of relocations were significantly disrupting livelihoods and that there was a need to address promptly the concerns over land and agricultural inputs to prevent worsening of the food security situation and to guarantee sustainable livelihoods to resettled people. 219. Government representatives that the Panel met in January 2013 are of the view that the four hectares of land allocated per household is adequate to maintain people’s livelihood. It is acknowledged that the relocation will imply changes in farming practices, and hence the Government’s emphasis on strengthening agricultural extension services, including with support from PBS. The Panel met several newly hired agricultural extension workers, also known as Development Agents (DAs) in each village visited. In one of the larger villages, the team was told that there are 18 DAs.226 In some villages, people complained that DAs cannot help them because the plots of lands they were given after moving are not cleared and thus not workable. In one village, the DAs told the Panel that they are encouraging people to go back to their old fields so as                                                              222 PBS III PAD: 5. 223 PBS III PAD: 39. 224 PBS III PAD: Annex 1. 225 The interpreter who was assisting the Panel was carrying a bag of cereals for his relatives in his native village, because there was food scarcity in that village. 226 At a meeting with a regional official in Gambella, he informed that Panel that there are now too many Development Agents employed, with the job is being given to large numbers of jobless 10th grade graduates. 51 to be able to farm and produce some food for their families, but not all villagers have the option to do that. 220. There seems to be a broad recognition that the speed of resettlement under CDP has not been met with commensurate resources and capacity, thus affecting basic services provision in Gambella, especially in the early years. As noted above, the Government representatives the Panel met during its eligibility visit referred to three challenges: lack of human resources in Gambella (i.e., people with required skills); very few contractors with the required capacity (“it may take a year to build a health post”); and not enough services to meet demands. Picture 6:  Panel Team meets with Anuak Village in Gambella  221. The effectiveness and quality of services in the agriculture sector is closely connected and influences the quality of livelihoods of villagers largely dependent on agriculture. Access to appropriate agricultural services is particularly important for indigenous groups such as the Anuak. As described in Chapter 1, the Anuak have a close attachment to their land and territories and depend predominantly on a subsistence economy. Livelihood for the Anuak is essentially determined by a combination of agriculture, fishing, gathering and hunting. Impacts of CDP in Gambella included the inability for villagers to farm the land in the new sites or the necessity to walk long distances to be able to farm, as well as difficulty in accessing rivers for fishing and forests for accessing building materials, fuel and fodder, with detrimental impacts for their livelihood. 222. The Panel notes that the pattern of land utilization and residence is based on the strong sense of Anuak identification with land and village. The Panel understands that in addition to other factors, this strong attachment to land and village is relevant to understanding, at least in part, Anuak’s resistance towards relocation, even if within the same woreda. 3.2.3. Project Appraisal 223. According to OP 4.10 (paragraph 2), the Bank recognizes that the identity, the culture and the livelihood of Indigenous Peoples are inextricably linked to the lands on which they live and 52 the natural resources on which they depend. Moreover, the policy acknowledges that their economic, social, and legal status often limits their capacity to defend their interests in and rights to lands, territories, and other productive resources, and restricts their ability to participate in and benefit from development. 224. The Panel finds that the application of OP 4.10 to PBS III, albeit under the form of functional equivalence, would have highlighted the need to prepare a social assessment of PBS beneficiaries meeting the criteria of the IP policy, and may have highlighted the impact that CDP was having on PBS beneficiaries, known and documented in a number of reports at the time of Project preparation, which could possibly affect PBS III’s results achievement. This further underscores the need for a risk analysis in the appraisal of PBS III regarding the impacts of a program aimed at resettling 70% of household, which are PBS beneficiaries in Gambella, and for identifying adequate mitigation measures that could warrant delivery of quality services for IPs. 225. Moreover, the Panel notes that the CDP program started its implementation in 2010 and PBS III preparation began concomitantly with the implementation of PBS II in 2011/2012. The Panel notes that JRIS mission reports under PBS II and PBS III do not mention the kind of issues noted above in their reporting on progress in PBS implementation. Furthermore, the Panel has noted that the PSIA study, which includes disaggregated agriculture sector results for regions, has omitted all data related to the agriculture sector’s results in Gambella, and is therefore unable to observe the specific results being achieved in agriculture in Gambella under PBS.227 226. It is the conclusion of the Panel that access to services in the agriculture sector in Gambella, along with the possibility of achieving the higher level objectives of the Project, was adversely affected by the CDP program as it was implemented. The lack of availability of workable land at the new resettlement sites is an aspect of CDP program which has had an impact on the delivery of results under PBS. The Panel notes that while the PBS results, indicated by the number of DAs and the number of direct beneficiaries, were achieved, the Panel was unable to find information on the effectiveness and quality of the service delivered in the agriculture sector, let alone the higher level objective of increased agricultural productivity in Gambella. As noted by the villagers the Panel met, access to the services provided by DAs had minimal effect when beneficiaries could not make efficient use of such services for lack of workable land nearby. The Panel’s position is that the above is also the result of a “thin” results framework for PBS, which “put little emphasis on the quality of or impact of basic service delivery”228 as already noted in an IEG evaluation of PBS I. Following a review of the results indicators for PBS I, II and III, it is the Panel’s view that such a conclusion on the results framework is relevant to PBS III as well, especially with respect to indicators related to the agriculture sector. 227. The Panel finds that, in accordance with Bank Policies, the operational interface between CDP and PBS should have been taken into account at the PBS project level, both during the appraisal and implementation phases, especially in a region such as Gambella where 60% of households, which are also PBS beneficiaries, were resettled as part of the Government’s CDP. The Panel finds that Management’s approach has not enabled PBS to mitigate or manage the harms described in the Request for Inspection with respect to access                                                              227 PSIA: Appendix D. 228 IEG (2013): x. 53 and quality of basic services in the agricultural sector and livelihoods of affected people in Gambella. 3.2.4. Monitoring during PBS III Implementation 228. As noted in Chapter 1, PBS is monitored through interrelated "Core PBS Principles” for tracking progress towards common objectives. One of these principles is Effectiveness, which focuses on how to carry out effective service delivery while identifying ways to improve them. Service delivery effectiveness is measured by looking at adequate sectoral resource allocation, balanced intra-sectoral allocation, and results achieved.229 In turn, results achieved are also measured with an Effectiveness Review, which examines results in the various sectors in terms of access, quality and inclusiveness.230 The PAD indicates that each JRIS mission will report on effectiveness (as on the other Core Principles) and will also review action plans for effectiveness of service delivery. 229. With respect to results in the agriculture sector, the May 2014 JRIS report presented the results of the recently completed PSIA study stating that the study showed that “woreda-level spending in education, health and agriculture has been effective in achieving important results”231 and “it is increasing productivity in agriculture.”232 The November 2013 JRIS, in reviewing the effectiveness principle, indicated that the donors had agreed to focus on health and agriculture results during this mission and thus participants received summaries of the results: although crop productivity was seen to be slightly declining, overall progress was encouraging and such developments were attributed, by the Government, to its efforts to scale up technologies and share best practices across farming communities.233 230. The Panel notes that PBS III also includes a “Managing for Results” component (M4R) aimed at enhancing the project effectiveness by ensuring that data, systems and analytic capacity are strengthened to deliver results throughout implementation. Under this component, PBS results will be monitored and targeted surveys and studies will be undertaken, to better understand how PBS results can be sustainably achieved. An Effectiveness Review on Agriculture, linking financing with results, is to be carried out under the M4R component. 231. It is the Panel’s position that, while monitoring was expected to follow the principle of effectiveness and results, the quality of services in Gambella was not in alignment with actual conditions on the ground. In particular, there is no indication that the agricultural extension services that are of paramount importance for the livelihood of rural populations was being monitored and aligned to deliver appropriate results for beneficiaries. 232. The Panel notes the results of the PSIA: “The evidence assembled here implies that decentralized spending at the woreda level is both effective and pro-poor”;” […] support for decentralized services in Ethiopia appears to be an effective use of development partners’ resources from both an efficiency and equity perspective. The only exception to these findings is                                                              229 PBS III PAD: 12. 230 PBS III PAD: 45. 231 JRIS Mission Report (May 12-15, 2014): 4. 232 JRIS Mission Report (May 12-15, 2014): 5. 233 JRIS Mission Report (November 11-14, 2013): 26. 54 agriculture, for which the impact of PBS-financed IGFT expenditure was smaller for the bottom quintile. In this instance, a wide array of factors is likely to be at work, especially poor farmers’ inability to buy inputs or the poor quality of their land.” 234 The Panel finds that results in agriculture seem to be particularly valid for Gambella and realities on the ground observed by the Panel and others during multiple visits to Gambella to review the implementation of the PBS and the CDP program. The Panel reiterates that there was an operational interface between the implementation of the CDP program and the PBS III project, by virtue of which the results of one were mutually impacted by the performance of the other. 233. Since PBS III began implementation, three JRIS missions were undertaken, and the related reports are silent on the issues noted above. The Panel finds that this is not consistent with the supervision provisions of the Investment Lending Policy, OP/BP 10.00.                                                              234 PSIA: 43f. 55 Chapter 4 Financial and Human Resources Link Analysis 234. The Requesters claim that government workers in the woredas, who receive their salaries ot of the PBS budget, have been forced to implement the CDP program. Bank Management maintains that PBS funds were not diverted and can be fully tracked. Management adds that it is “neither feasible nor desirable”235 to require all officials, whose functions are being partly financed by the Bank, to work exclusively on those functions. This Chapter addresses these claims by analyzing the financial risk assessment and management issues raised in the Request, followed by the claims related to human resource issues. 4.1. Financial Risk Assessment and Management 4.1.1. Requesters’ Claims and Management Response a) Requesters’ Claims 235. The Request claims that funds are transferred through country financial systems, pooled with other funds and transferred to the regions and woredas through existing government systems; it states that “[t]here are no separate bank accounts beyond the initial entry point into the Treasury and no separate disbursement or accounting procedures.”236 The Requesters also claim that there is “no publically available information that shows the precise source of financing for the villagization budget.”237 The Requesters cite the 2011 Study on Strenghtening Grievance Redress Mechanisms for PBS, which was commissoned by the GoE and its DPs and states that “[i]n effect, one can argue that PBS pays a portion of the compensation of all regional government and local government employees (not just salaries and benefits in the five sectors enumerated above) because PBS funds are commingled with funds from other sources that regional state and local governments use to pay employee compensation.”238 236. The Request also refers to a meeting between Bank Management and IDI, where the Bank “insisted that PBS funds do not contribute directly to the Villagization Program in Gambella or elsewhere. Bank representatives stated that they were able to track how PBS funds are spent down to the woreda level, and emphasized that PBS funds were used to pay the salaries of public servants such as schoolteachers, health professionals and agricultural extension workers”239. 237. The Requesters also argue that PBS funds were diverted through deductions from their salaries or non-payment of their salaries to finance the implementation of the CDP Program. IDI was informed by a teacher that “he was told that the budget from the federal government for implementation of the Villagization Program ran out, so woreda-level civil servants […] had 5% of their salaries deducted from their payroll in order to cover the shortfall”240.                                                              235 Management Response: 15. 236 PBS II PAD: 11. 237 Request for Inspection: 6. 238 Randolph, R., Edjeta, B. (2011): Study on Strengthening Grievance Redress Mechanisms for the Protection of the Basic Services (PBS) Program in Ethiopia: 9. 239 Request for Inspection: 6. 240 Request for Inspection: 7. 56 238. The Requesters conclude by stating that “[i]t is therefore apparent that Bank funds through PBS are substantially contributing to the implementation of the Villagization Action Plan.”241 The Requesters add that from publicly available information, or the explanation provided by the Bank it is not clear how the Bank financial tracking and accountability systems for PBS would detect the diversion of funds towards the implementation of villagization. b) Management Response 239. In its Response, Management notes that since 2005, there were allegation of misuse of funds related to programs in Ethiopia supported by DPs. Management adds that PBS has always been one of the programs that have been subject to such allegations. Management states that given the general nature of the allegations, there was no basis to re-visit the implementation support arrangements for PBS, which were carefully thought through, strengthened over time, and were considered robust.242 240. Management explains that “Villagization is administered by the GoE, along with decentralized levels of government, but it has a separate budgetary authority, and separate implementation arrangements.”243 The Response argues that the Bank does not finance the CDP Program and that “the regular implementation support arrangements for PBS have never found any evidence of funds diversions of the kind alleged”244. Management then explains that it “considers this finding credible in view of the careful arrangements in place within the program to track the use of funds”245. 241. The Management Response states that “[o]nly recurrent expenditures (salaries, operations and maintenance) in the relevant service sectors are eligible for financing from pooled PBS donor and Government sources. In depth reviews have been undertaken by the Bank for the last four years to track the resources it makes available by category and usage at the woreda and regional levels”246. In the Annex, the Response then outlines the financial tracking and management procedures of the Bank in detail.247 242. The Management Response explains the choice of lending instrument and notes that an investment lending instrument enables the Bank to finance recurrent expenditures at woreda-level for defined basic services. Management adds that through extensive project implementation support, “the Bank tracks eligible expenditures to woreda level.”248 243. Management states that the Bank conducts extensive financial management review of program resources. The Management Response elaborates on the financial tracking and states that the Bank ensures that when block grant funds for the program are disbursed, they are separately deposited in a Designated Account at the federal level. Separate books and documentation are                                                              241 Request for Inspection: 7. 242 Management Response: 11. 243 Management Response: 22. 244 Management Response: 23. 245 Management Response: 23. 246 Management Response: 25. 247 For more information, please see the Management Response: 25. 248 Management Response: 7. 57 maintained at the federal level. Management adds that before the funds are disbursed to lower levels, they are put through the GoE treasury systems and pooled with Government resources. Management states that block grant allocations from the federal level to regions and from regions to woredas are determined based on objective criteria and publicly available formulae. Management clarifies that these formulae provide more per capita resources to regions with lower development outcomes to address equity concerns. Management states that, on a per capita basis, Gambella receives more than three times the national average block grant allocation.249 Management notes that at the woreda level, the funds are spent according to dedicated budget lines. Records are kept, sectors keep an appropriate record of staff payments and the attendance register. Also, fund lines for salaries from the program are complied with. Management notes that the GoE requires woredas to submit monthly reports to regional governments, where the reports are consolidated and reported to the federal level. These reports inform quarterly expenditure and Interim Unaudited Financial Reports, which the Bank uses to track fund flows and expenditures. Management notes that, “based on these financial controls, the Bank will only disburse Basic Service Block Grant resources if PBS 3 financial reports are found acceptable.”250 Management adds that no diversion of funds has been indicated to date.251 4.1.2. Panel’s Analysis 244. The PBS III PAD notes that, since the inception of the PBS program in 2006, the World Bank used investment lending instruments to support the PBS program. According to the PAD, PBS III continues to use this category of instrument and was structured as a five year Specific Investment Loan (SIL). The Basic Service Block Grants component, which is the sub-program A of PBS III, with a total amount of US$6.2 billion, including IDA financing of US$555 million, to finance recurrent (salaries, operations and maintenance) expenditures in five basic service sectors (education, health, agriculture, water and sanitation, and rural roads) at local levels, constitutes the largest component of PBS III. 245. Management in its Response explains that it carefully considered alternative lending instruments, such as Development Policy Operations (DPOs) and the new Program for Results (PforR), but did not find these instruments superior to the investment lending approach of PBS: the chosen investment lending instrument allows fast disbursements through government systems combined with significant efforts to strengthen the systems’ capacities. It allows the Bank to support recurrent expenditures based on joint monitoring of principles and results. Through incorporating reviews about the progress of systems strengthening, PBS “allows a frank and productive dialogue on progress and constraints […]”252. Management further explains that the chosen investment lending instrument lets the Bank define eligible expenditures and, through extensive implementation support, track the eligible expenditures to the woreda level.253 246. According to the Bank’s Independent Evaluation Group (IEG), the PBS program “presents an unusual case in which the Bank provides large scale financial support for expanded service                                                              249 Management Response: 25. 250 Management Response: 8. 251 Management Response: 24f. 252 Management Response: 7. 253 Management Response: 7. 58 delivery”254, which, unlike conventional DPOs, “is multi-year, does not rely on annual programs built around ‘prior actions’ and ‘triggers’, and is in keeping with Paris Declaration principles placing program ownership firmly in the hands of Government”255. This program “effectively takes on key features of the Bank’s new Program for Results that links disbursements to defined results.”256 247. The IEG report notes that, responding to concerns that the PBS program could release funds for the government to exercise control over the population, or that the government could steer funding away from opposition regions towards those it considered loyal, the Bank needed to build tests of additionality and transparent and fair allocation into its operation. Partnerships with CSOs were also regarded as crucial to support the good governance agenda.257 248. According to the IEG Report, PBS is technically an investment program; however, most of the program consists of budget support for block grants.258 The PBS program was intended as a short-term response to a crisis situation, supplementing the Bank’s DPOs, which provided large amounts of budget support. Therefore, some important aspects, such as the end-result of the project, were not sufficiently considered, if at all. The IEG evaluation rates the project objectives of PBS as substantial, as a continuation of general budget support would not have been backed by other donors. The PBS approach was designed to strengthen the government’s own system to manage decentralization, and it increased the Bank’s and other donors’ access to regional and local governments.259 However, the IEG evaluation concludes that the most important PBS instrument, the basic services block grant, was set up as a Sector Investment Loan, but it was apparent from the start that it was “simply an alternative mechanism for continuing to derive the perceived benefits of budget support”260. 249. Against this overall context in the understanding of the financing mechanism, the Panel’s analysis deals with three issues raised by the Requesters: i. Whether there was diversion of financial resources from PBS to CDP, ii. What lessons did the Bank take from PBS II and incorporate in the design of PBS III to mitigate fiduciary risk, especially since it became aware of CDP in the late stages of PBS II, and iii. Whether the Bank’s appraisal and supervision of government arrangements for the financial management of PBS III comply with Bank policy. c) Diversion of Financial Resources 250. The Inspection Panel reviewed three issues with regards to the claim of diversion: potential, detection and evidence. The analysis below follows this structure.                                                              254 IEG (2013): xii. 255 IEG (2013): xii. 256 IEG (2013): xiii.   258 IEG (2013): 27. 259 IEG (2013): ix, x, 5f. 260 IEG (2013): 6. 59 Potential for Diversion 251. The potential for diversion from PBS funds to CDP activities has to be assessed by reviewing both: (a) the fiscal transfer system, and (b) the regional expenditure system. Regarding the fiscal transfer system, PBS uses a block grant, which comingles Development Partners (DP) funds with domestic funds and does not earmark them. Regarding the regional expenditure system, PBS III, similar to all phases of PBS, uses the government’s financial system. 252. The Fiscal Transfer System. The principal source of funding for regions (approximately 80% or more) comes from the federal-to-region fiscal transfer (FRFT). Once the block grant reaches the region, it is combined with the region’s revenue collected from three tiers – region, zone and woreda. This pool of funds is then allocated to regional sector bureaus (e.g. Bureau of Health), zone sector departments (e.g. Department of Health) and then to woredas through a region-to-woreda fiscal transfer (RWFT). In accordance with GoE’s policy to decentralize service delivery to woredas, regions transfer the bulk of regional resources, without specification as to the amount, including what has been received through the FRFT ,to woredas by a block grant. The RWFT provides nearly all of the financial resources available to a woreda though there are other sources of funds such as project support from foreign aid agencies that do not flow through the treasury system. 253. The guiding principle for both the FRFT and the RWFT – and a cornerstone of Ethiopia’s fiscal devolution – is that the recipient has discretion as to the use of these funds. While there are multi-year plans at the national and regional level and in some woredas, the plans are for guidance. They do not determine how the transfer is translated into the budget of regions, zones and woredas. Restated, there is no document issued by the federal government to a region that accompanies the FRFT that specifies the use of funds. The same is true for the RWFT—the region cannot specify to the woreda how it is to allocate the funds.261 254. The pool of funds available to the federal government for allocation to the regions through block grants has increased through all phases of PBS.262 Comingled with domestic funds, PBS funds increase the size of the FRFT and thus in turn the RWFT. 255. The following Table 1 presents the budgeted and actual expenditures in Gambella for the period 2007/08 to 2011/12. PBS is meant to sustain the level of recurrent expenditures for basic services. The region has more than maintained the level and has increased it over time at both the region and woreda levels (Table 1, rows 5 and 7)263. Since the World Bank became aware of CDP in 2010, the budgeted recurrent expenditures for basic services at the regional level increased                                                              261 By examining the annual budget law of the federal and regional governments it can be inferred that both the FRFT and RWFT are not specified. At the federal level the law appropriates to each region a lump sum without specification of sectors for it to be used. The same practice occurs at the regional level for their woredas - a lump sum is appropriated for each woreda. 262 PSIA: 6. Table I.1. 263 The table presents the budgeted and actual expenditures in Gambella for the period 2007/08 to 2011/12. The budget listed in Table 1 includes the FRFT and all revenue collected in Gambella region. The FRFT which includes PBS funds accounts for approximately 80% or more of Gambella’s resources. The percentage of the FRFT in recent years is as follows: 87% for 2009/10, 84.6% for 2010/11, and 79.4% for 2011/12. Calculated from Table 1 and Figure 8 in PBS Secretariat (2013a): Briefing Book: PBS Supervision Mission Cross-Cutting Issues Gambella National Regional State. (March 18-22, 2013): 13. 60 206.5% and the actual expenditure from these budgets increased 28.5%. For the same period, the budgeted recurrent expenditures for basic services at the woreda level increased 174.1% and the actual expenditure from these budgets increased 40.2%. What this means is that, since the World Bank has become aware of CDP, funds flow for basic services to both the regional and woreda have not only been maintained, but have increased substantially.   Figure 2 Gambella - Budget Performance (in million birr) 256. The PBS III PAD recognizes that woredas have access to “unearmarked resources” through block grants.264 The use of FRFT and RWFT has two implications with respect to the issues raised in the Request. The Panel notes that the unearmarked and discretionary nature of block grants raises the potential for diversion. The Panel also notes that, the fact that FRFT and RWFT are not specified and give discretion as to their use means that Management’s claim that they can track PBS funds at the woreda level is not supported 257. The Regional Expenditure Systems. A key principle of the PBS Program, including PBS III, is to use the financial systems of the Government of Ethiopia. Therefore, the potential for diversion of funds from PBS depends on the quality of these systems. PBS II and III funds are disbursed through the Government’s Treasury System (TS), the principal conduit for the disbursement of domestically raised funds.265 The TS has seven components (budget, disbursement, accounts, reporting, internal control, external audit, and legislative scrutiny), which, with some exceptions, operate at four tiers of government (federal, region, zone, woreda). To use the TS means that funds are managed by the full array of the seven components of financial management and that funds flow through three critical processes: they are proclaimed in a budget so expenditure has legal authority; they are disbursed through Government facilities; and they are recorded and reported in consolidated government accounts.266 The Government reports to DPs on                                                              264 PBS III PAD: 2f. 265 Management Response: 25. 266 The Treasury System is often referred to as the “Channel 1” system. There are six other “Lines of Financial Management” used by government and/or donors to disburse funds to regions and woredas. The TS disburses funds through the finance organizations at the respective tiers of government/administration where expenditure are incurred: 61 any funds disbursed through the TS. The TS is supported by a financial information system (IBEX 1.3), which operates selectively at all tiers of government and facilitates the recording and consolidating of financial reports. 258. The TS has been rated as high risk for PBS III.267 The PAD for PBS III considered both overall inherent risk and control risk to be high.268 Of particular concern is the PAD’s finding that “there remain weak internal audit controls at the federal and regional levels and continued inadequate follow up on audit reports.”269 The supervision mission to Gambella found that there has been no significant reduction in fiduciary risk and the country financial system has serious weaknesses. The areas of greatest concern it noted were the weakness in internal control and the absence of internal audit.270                                                              the Ministry of Finance and Economic Development (MoFED) at the federal level; Bureau of Finance and Economic Development (BoFED) at the region level; Department of Finance and Economic Development (DoFED) at the zone level; and Office of Finance and Economic Development (OoFED) at woreda level. 267 The World Bank has defined fiduciary risk as: [T]he risk that Bank funds will not be used for the intended purposes, or that they will be used without due attention to economy and efficiency. In projects using country FM systems, Bank funds are commingled with the country’s own funds: therefore, a fiduciary risk assessment also needs to consider broader country PFM risks that could affect the fiduciary risk. The fiduciary risks are mainly ‘control’ risks, whereas the broader PFM risks are mainly ‘inherent’ risks. (World Bank (2009): Assessment of Fiduciary Risks in the Use of Country FM Systems in Bank Financed Investment Projects: Interim Guidance Note for FM Staff: 7.) The World Bank’s CFAA guidance note defines fiduciary risk as: [A]n assessment of the risk to World Bank funds that are managed through the country’s PFM system. (This risk also applies to the country’s own funds and those of other development partners that are managed through the PFM system)….The main component of financial management risk is the probability that the PFM system will not provide appropriate management of all public funds. Also at issue in financial management risk is whether there is sufficient transparency (quality of information) to determine how funds are spent or managed. (World Bank (2003): Country Financial Accountability Assessment Guidelines to Staff: 6.) 268 PBS III PAD: 70f. 269 PBS III PAD: 30. 270 PBS Secretariat (2013a): 10f. 62 Box 5: Financial Management in Gambella (March 2013) Financial Management in Gambella (March 2013)  Budgeting. A complete approved budget for EFY 2005 is not yet captured in IBEX as some woredas did not complete the budget form properly.  Staffing. Staffing and staff capacity is a key constraint to effective accounting and reporting. Within the Office of the Regional Auditor General (ORAG), only 57 percent of available positions are filled due to budget constraints. Three-quarters of ORAG’s auditors hold only diploma. Within the Bureau of Finance and Economic Development’s (BOFED) IBEX data entry section, only 2 staff are handling the significant responsibility of capturing woredas’, zones and region’s financial reports.  Reporting. The Region’s accounts for EFY 2002, 2003 and 2004 were not published and submitted to ORAG for audit. Although the year-end accounts for EFYs 2000 and 2001 were recently submitted to the ORAG, the accounts are not yet audited due to the Office’s operational budget constraints and staff capacity problem. As such, there is presently an audit backlog of 5 years (from EFY 2000 to EFY 2004) (approximately Fiscal Years 2007 to 2011) (emphasis added).  Internal controls. Cash management control at the visited woredas was weak. Regular cash reconciliation and monthly bank reconciliation is not performed. This is also true at BoFED (the Bureau of Finance and Economic Development at the regional level). The effort being made to control property management is encouraging, this is not likely to be sustained given the staffing issues.  Internal audit. At the visited woredas, the mission noted that the internal auditing is not done. The assigned staff does not have the relevant qualifications or experience. Source:  PBS Secretariat (2013b): Report. PBS Supervision Mission. Gambella National Regional State. (March 18- 22, 2013): 11f. 259. The high financial risk of the TS, which is used to implement PBS, has two implications. The weakness of internal controls supports the possibility that funds could have been diverted. It also means that diversion cannot be verified, and that the World Bank’s assertion that it can fully track expenditures cannot be supported. Evidence of Diversion 260. The Panel did not have access to Government financial records, which would have been necessary, to find evidence of diversion. Therefore, this analysis cannot present findings on this issue. Detection of Diversion 261. Regarding the detection of diversion, the relevant issue is the robustness of the PBS III measures to report and monitor PBS-eligible expenditures to detect distortions in allocations of 63 resources. The issue of whether the FRFT and RWFT promote detection is particularly relevant in assessing the capacity of PBS III to detect diversion of funds. 262. The use of the block transfer system (the FRFT and RWFT) means that detection can be done only in terms of aggregate flows. Moreover, the use of the TS, which is weak, compromises the Bank’s ability to detect diversion. One key means of detection that the Bank uses is timely and good-quality external audit reports. For Gambella, the 2013 supervision mission found a five-year backlog of audits and an absence of internal audit.271 Given the backlog of external audit and absence of internal audit in Gambella, the Bank’s principal means of detecting diversion lacks timeliness and quality. In addition, the detection system itself is not improving; the 2013 Supervision Mission found that no woreda fully implemented the prior period’s audit recommendations. 272 The PBS III PAD acknowledged these weaknesses: “Some challenges to the timely submission of quality report and audits remain: there remain weak internal audit controls at the federal and regional levels and continued inadequate follow up on audit reports.”273 d) Fiduciary Risk Management 263. Given that the World Bank became aware of the CDP in late 2010, the Panel’s analysis includes what lessons from the previous phases of the program were taken into account in the design of PBS III to mitigate fiduciary risk. 264. One of the important lessons to improve the fiduciary management includes financial capacity building at the woreda level.274 The PBS III PAD also outlines perceived financial management risks to PBS III and suggested mitigation measures. The fiduciary risks mainly cover: (a) weak capacity at the federal, regional and woreda levels, (b) high staff turnover, (c) quality and timeliness of external and internal audits. 265. The PAD notes that overall staffing issues are being addressed by the Civil Service Reform Program. Strenghtening of monitoring and supervision activities at the federal, regional and woreda levels is also mentioned as a risk mitigation measure. The PAD states that “[l]ate reporting is being dealt with by taking appropriate measures including withholding funds to woredas and Regions that report late.”275 The PAD also refers to a separate PFM component to address woreda level weaknesses as part of the overall program. 266. As outlined in Box 5 above, many of the fiduciary risks outlined in the PAD continue to exist for Gambella. Although several improvements took place, as noted in the latest JRIS reports, there are still outstanding issues related to the quality and timeliness of audit reports in Gambella.276 Furthermore, despite the five-year backlog of audits in Gambella, the Panel did not                                                              271 See Box above. 272 PBS Secretariat (2013b): Report. PBS Supervision Mission. Gambella National Regional State. (March 18-22, 2013): 13. 273 PBS III PAD: 30. The recent JRIS Aide Memoire from the May 2014 mission shows a low rate of some indicators, including the number of woredas producing IBEX and IFMIS-based financial management reports. (JRIS Mission Report (May 12-15, 2014): 38.) 274 PBS III PAD: 21ff. 275 PBS III PAD: 70. 276 JRIS Mission Report (November 11-14, 2013): 18. 64 find any information regarding any funds being withheld due to late reporting, as stated as a mitigation measure in the PAD. 267. Thus, in conclusion, despite efforts through PBS as well as other projects, the government’s financial system remains weak and its key elements have persistently been rated as high risk.277 e) Compliance with Bank Operational Policies and Procedures Relevant Bank Policies 268. The World Bank policy on Financial Management applicable during the design and approval of PBS III (OP/BP 10.02 on Financial Management) states that “[f]or each operation supported by a Bank loan, the Bank requires the borrower to maintain financial management arrangements that are acceptable to the Bank and that, as part of the overall arrangements that the borrower has in place for implementing the operation, provide assurance that the proceeds of the loan are used for the purposes for which the loan was granted.”278 269. This policy adds that “[t]he Bank assesses the adequacy of the borrower’s financial management arrangements during the preparation and implementation of each operation and requires the borrower to undertake appropriate measures, including institutional capacity strengthening, to mitigate risks posed by weaknesses that are identified.”279 According to the policy, “[i]f the borrower fails to maintain acceptable financial management arrangements, or to submit the required financial reports by their due dates, the Bank takes action to rectify the situation.”280 270. The Bank policy related to financing of recurrent expenditures, applicable during the appraisal and approval of PBS III (OP/BP 6.00 on Bank Financing), states as one of the guiding principles for the Bank to finance recurrent expenditures is to have “acceptable oversight arrangements, including fiduciary oversight arrangements, are in place to ensure that such loan proceeds are used only for the purposes for which the loan is granted, with due attention to considerations of economy and efficiency.”281 271. The Bank policy on Supervision (OP/BP 13.05) and the Bank policy on Investment Project Financing (OP/BP 10.00), which replaced OP/BP 13.05 in April 2013, both indicate the Bank’s responsibility to monitor “the Borrower’s or the Project Participants’ compliance, with its (or their) obligations as set out in the legal agreements with the Bank.” The policy adds that “[t]he Bank also provides implementation support to the Borrower or the member country by reviewing information on implementation progress, progress towards achievement of the Project’s development objectives and related results, and updates the risks and related management measures.”282                                                              277 PBS III PAD: 71. 278 OP 10.02, para. 1. 279 OP 10.02, para. 2. 280 OP 10.02, para. 5. 281 OP 6.00, para.1. 282 OP 10.00, para. 21. 65 272. In determining the Bank’s compliance with its policies, the Panel analyzed whether: (a) the Bank adequately assessed the GoE’s financial management arrangements during project appraisal in accordance with World Bank policy; and (b) the Bank adequately supervised and provided adequate implementation support with regards to the financial arrangements in accordance with Bank policy. Project Appraisal 273. The PBS III PAD recognized the persistent high levels of fiduciary risk facing the project.283 As in PBS I and II, PBS III uses the country system and the country’s institutions to manage the Project’s finances, to ensure acceptable financial performance and mitigate fiduciary risk. The PBS III PAD states that “[w]hile Ethiopia's federal system is complex, allowing strategies, resources, and information to flow between ten regional/urban governments and then to more than 1000 woreda administrations, its fiduciary and economic governance system is robust. […] While strong in design, execution of that woreda-level economic governance system needs to be made more resilient, particularly in some jurisdictions”.284 274. Accordingly, a key focus of PBS III was to build financial capacity at the woreda level.285 However, a problem that was highlighted repeatedly was staffing. The PBS III PAD noted, at the woreda level, that there is ”high staff turnover, estimated to be 25 percent per annum, where PFM (and other) woreda staff frequently move into and out of their positions. Often this staff turnover involves those without fiduciary training taking on financial management or procurement tasks.”286 The March 2013 supervision mission found staffing to be a key constraint at both the regional and woreda levels, the result of “inadequate civil service management.”287 275. According to the Bank Policy on Financial Management, OP/BP 10.02, “the Bank requires the borrower to maintain financial management arrangements that are acceptable to the Bank and that, as part of the overall arrangements that the borrower has in place for implementing the operation, provide assurance that the proceeds of the loan are used for the purposes for which the loan was granted.”288                                                              283 PBS III PAD: 70f. Control risk was disaggregated into the following components and their scores: budgeting (satisfactory), accounting (high), internal control (high), funds flow (satisfactory), financial reporting (high) and auditing (high). Of particular concern is the finding of the PBS III PAD that “there remain weak internal audit controls at the federal and regional levels and continued inadequate follow up on audit reports.”(PBS III PAD: 30.) 284 PBS III PAD: 3. 285 “The Woreda PFM and Procurement Strengthening Sub-Component will provide support to woreda-level PFM systems. […], the sub-component will provide technical support to woredas, help strengthening the existing PFM committees at zones and woredas to provide hands-on support to woredas on PFM issues, support strengthened internal control and procurement systems at woredas (that PFM studies have repeatedly pointed out as the weakest parts of the PFM system) […]” (PBS III PAD: 58.) 286 PBS III PAD: 16. 287 PBS Secretariat (2013b): 11f. While the Government’s Zone and Woreda PFM committees are to provide technical support, training is to be delivered through the Regional Management Institutes (RMIs). The very institutions that were to manage PBS III funds and mitigate fiduciary risk are themselves at risk. 288 OP 10.02, para. 1. 66 276. The Bank policy on Appraisal, OMS 2.20 also requires the Bank to give due consideration to the capacity of the borrower’s financial management institutions and systems and introduce adequate mitigation measures. Considering the specific features of the block grant mechanism to finance basic services, which uses the Government systems and, as confirmed by the IEG, is in essence a budget support, this was particularly important. 277. As noted above, the PBS III PAD acknowledged the overall inherent risk and control risk related to fiduciary arrangements to be high. The PBS III PAD also noted weaknesses in PFM systems at the federal, regional and woreda levels, capacity issues, high staff turnover and issues related to the timeliness and quality of internal and external audit. Weak internal audit controls at the federal and regional levels and continued inadequate follow up on audit reports are also described in the PAD. Furthermore, although these fiduciary risks have been acknowledged in PBS III documents, the mitigation measures were not adequate. 278. The PBS III PAD specifies mainly three PDO results indicators for financial management: i. Woreda-level Offices of Finance and Economic Development (WoFEDs) that have effectively rolled out IBEX. ii. Audit coverage of the annual budget (in terms of woreda offices and federal agencies and ministries that are audited annually, and of woredas that have implemented prior audit recommendations). iii. Regions that submit quarterly Interim Financial Reports without major deficiencies.289 279. The institutional means for mitigating fiduciary risk, principally at the woreda level, is the establishment and functioning of Zone and Woreda PFM committees, yet there is no PDO results indicator for these institutions. None of the PDO results indicators addresses the fiduciary risk created by the lack of internal control. It is the view of the Panel that the PBS III PDO results indicators that are relevant to the sources of fiduciary risk were not adequate. 280. This inadequacy was critical, in that the central tenet of the risk framework as per ORAF Guidelines is that the risks being assessed are risks to achieving PDOs. Thus, according to the ORAF Guildelines, the initial risk analysis can only be completed after PDOs and selected PDO results indicators have been established. In using ORAF to identify and assess risks, and to monitor risk management and mitigation during implementation, the starting point, according to the Guildelines, is always the PDOs and progress towards their achievement. Accordingly, in the Panel’s view, inadequacy in establishing appropriate results indicators for financial management meant that the framework for risk identification, assessment, as well as for monitoring progress, remained incomplete. 281. The Panel notes that while the “Financial Management and Disbursement Arrangements,” in Annex 4 of PBS III PAD scores both “Overall Inherent Risk” and “Overall Control Risk” as                                                              289 PBS III PAD: 40f. 67 high290, the ORAF risk assessment in Annex 6 of the PAD scores the implementing agency risk and project risk as “moderate,” despite the many known weaknesses.291 282. The Bank policy related to financing of recurrent expenditures, applicable during the appraisal and approval of PBS III (OP/BP 6.00 on Bank Financing), states as one of the guiding principles for the Bank to finance recurrent expenditures is to have “acceptable oversight arrangements, including fiduciary oversight arrangements, are in place to ensure that such loan proceeds are used only for the purposes for which the loan is granted, with due attention to considerations of economy and efficiency.”292 283. Based on the above, the Panel finds that Management did not comply with the requirements of OMS 2.20 and OP/BP 10.02 in the design and appraisal of PBS III. The Panel notes that the Bank’s assertion that the funds can be tracked at the woreda level in Gambella cannot thus be sustained. Supervision   284. The ORAF Guidance Note states that “Risk assessment is a dynamic process starting with preparation and continuing through implementation.(emphasis in the original). The assessment will help to continuously monitor the evolution of risks; to identify the emergence of new risks; to assess progress with, and impact of the implementation of risk management measures; and, as necessary, to devise appropriate adjustments to support the achievement of the project’s results.” 293 Accordingly, missions have assessed performance in financial management as below. 285. The March 2013 PBS III supervision mission found serious deficiencies in the Gambella region’s performance in financial management and on the PFM PDOs results indicators.294 The supervision mission to Gambella found that there has been no significant reduction in fiduciary risk and the country financial system has serious weaknesses. The areas of greatest concern it noted were the weakness in internal control and the absence of internal audit. It noted the absence of results indicators for internal control and proposed the addition of new ones: the number of WoFEDs that (a) have connected to WoredaNET for online connection to IBEX, (b) have strong internal audit units; (c) have sound cash management control; and (d) have sound property and inventory management control.295 The Panel recognizes that noting the absence of PDOs results indicators for internal control is in line with the ORAF Guideline. However, the proposal to add new ones has not been adopted; no new PDOs and indicators have been added to PBS III.                                                              290 PBS III PAD: 70f. 291  Staff turnover; lack of trained staff; little or no internal control; reliance on weak institutions (zone and woreda PFM committees); reliance untested regional institutions (Regional Management Institutes to deliver in-service training in financial management); and, the creation of new institutions (woreda support units). (PBS III PAD, pp. 40f.)   292 OP 6.00, para 1. 293 Guidance Note on the ORAF (2011): 1. 294 PBS Secretariat (2013b): 13. 295 Kalua, S., Mkandawire, A., Kalu, U. M., Wole, Bonsu, V. (2010): Ethiopia: Draft Consolidated Report on In- Depth Financial Management of Productive Safety Net (I, II, & III): Food Security: Protection of Basic Services (PBS I & II): WASH: Rural Capacity Building: RSDP (I & II): ULGDP: 21. 68 286. The financial management review mission, which was conducted from October 2013 to December 2013 as part of the JRIS, concluded that “there is reasonable adequacy that the FM system in place provides the necessary reasonable assurance that Bank loan proceeds are being used for the intended purposes and that the reports being produced by the system can be relied upon to monitor project. However, certain areas were identified to need management attention, action and follow up”296, including regular analysis of budget and explanation of variances between actual and budgeted expenditures, more clarity as to the roles and responsibilities of accountants, further regular trainings on the PFM system and Bank procedures to curb turnover issues, and others.297 287. The Panel was also informed during its visit to Gambella that due to a combination of frequent staff turnover, capacity issues, as well as unreliable electricity and inaccessibility, unless the project accountants go regularly to assist some of the weakest woredas, they are not able to send their accounts as required. 288. The Panel notes that Management paid attention to problems in financial management through these missions. However, the starting point for both appraisal and monitoring of fiduciary risk is the adequacy of PDO results indicators, as discussed already. The Panel finds that, since PDO results indicators that directly address fiduciary risks were inadequate in the initial planning, and subsequently have not been adjusted, the supervision of those risks is not in compliance with Bank policy OP/BP 10.00. 289. Nevertheless, the Panel notes that the issue of variance among woredas is recognized in the design of PBS III and is being addressed through sub-component B, which acknowledges that the country's decentralized, integrated system of fiduciary, administrative and information governance should continue to be strengthened, particularly in the large number of very diverse woreda administrations. The Sub-program B, Strengthening Local Accountability and Transparency Systems, aims at maintaining and strengthening these systems by focusing on the woreda as the “frontline of administration with responsibility for PBS objectives.”298 The sub- program includes three interrelated components: Citizen's Engagement, Local Public Financial Management, and Managing for Results. Given the Panel's findings on the situation in Gambella, going forward, it is hoped that particular attention may be paid to the woredas in developing regional states such as Gambella, to strengthen their capacity to ensure adequate implementation and monitoring of Project’s financial management arrangements.                                                                296 Financial Management Review Mission, part of Joint Review and Implementation Support (JRIS) Mission (October 2013 – December 2013): 2. 297 For more information, see Financial Management Review Mission, part of JRIS Mission (October 2013 – December 2013): 2ff. 298 PBS III PAD: 14. 69 4.2. Human Resources Link Analysis 4.2.1. Requesters’ Claims and Management Response a) Requesters’ Claims 290. The Requesters claim that government workers in the woredas, who receive their salaries out of the PBS budget, have been forced to implement the CDP program; those who refused to take part in the implementation of the program, have been targeted with arrest, beating, torture and killing.299 The Annex to the Request states that “former teachers and agricultural workers from Gambella region, who are among the Requesters, have testified to IDI and Human Rights Watch that they and other civil servants were ordered to implement the Villagization Program in Gambella”300. During the eligibility visit, affected people who met with the Panel in South Sudan also told the team that their salaries were decreased by 5% to fund activities under the CDP program. b) Management Response 291. Management claims that there may be some situations at the woreda level, where local government workers paid under PBS may have responsibilities relating to the CDP Program; however, “[t]his reflects the reality of governments everywhere, in which officials, particularly at local level, have multiple roles in discharging government policy”301. According to Management, the institutional arrangements for the CDP implementation include steering and technical committees at the regional level and a similar structure at the zone, woreda and kebele level. PBS III, which contributes to cover the cost of all staff working in the five basic service sectors, also contributes to the salaries of those working in the committees. The Management Response states that “it would be neither feasible nor desirable to pursue an arrangement under which all officials, whose functions are being partly supported by Bank finance, are required to work exclusively on those functions. While such an arrangement might protect the Bank from allegations such as those raised in this Request, it would be highly dysfunctional and inefficient and burden the borrower with fragmented administration and additional costs. Moreover, it would run counter to the ongoing international efforts and agreements for Aid Effectiveness”302. Management continues to argue that the key concern for PBS III must be whether each government worker has discharged her/his duties under PBS III as the basis for the payment of salaries.303 4.2.2. Panel’s Analysis 292. The Panel heard a range of different testimonies on this issue. Affected people, whom the Panel met during the eligibility visit in South Sudan, told the Panel that civil servants in Gambella were forced into helping in the implementation of the CDP and suffered retaliation and reprisals if they refused or complained, or were labeled as agitators and arrested. During the visit to Gambella, the Panel team was informed of civil servants involved in the resettlement of people in one village. Villagers told the Panel of teachers or Development Agents (DAs from other kebeles) who were                                                              299 Request for Inspection: 1. 300 Request for Inspection: 6. 301 Management Response: 24. 302 Management Response: 15. 303 Management Response: 15. 70 called to their own village to assist with clearing the land, building houses, cutting grass etc. They added that all woreda workers in the village contributed 4% of one month salary to pay those who came to help. They described the 4% amount as a “voluntary contribution” they gave after their salaries were paid, rather than a deduction from the salary before it is paid. 293. The Panel was unable to find conclusive information on this claim by the Requesters. The information gathered by the Panel is not sufficient to determine whether woreda workers, financed by PBS, were forced to participate in the implementation of the CDP program against their will, or what kind of activities they carried out in this respect; whether their salaries were decreased to finance activities under the CDP or whether the CDP was partially financed with so-called voluntary contributions of civil servants. As Management notes in its Response, woreda workers did participate in the implementation of the CDP. 294. Importantly, the Panel was also not able to find in JRIS or other reports any information from Gambella or from the other regions where CDP was being implemented, regarding the extent to which workers financed under PBS were undertaking additional or alternative work in relation to CDP, which may, or may not have impacted on achieving PBS PDOs. 295. The Panel notes that it does not have sufficient information to draw a firm conclusion on this issue but it again notes how the operational interface between the CDP and PBS, and mutual impacts affecting the achievement of results under PBS, were not adequately considered at appraisal nor during implementation and therefore that this human resources link cannot be overruled. 71 Chapter 5 Conclusions 296. This investigation has proven to be among the most challenging for the Inspection Panel for several reasons. First, although the Requesters’ claim of harm is focused on the CDP, they allege that PBS enabled this harm caused by CDP. This presented a methodological challenge in terms of determining whether there was a link between the alleged harms and the Bank-financed PBS project.   297. Second, Bank funding is intermingled with financing from GoE and other development partners to pay for recurrent expenditures at the woreda level based on joint monitoring of core principles and program results. This lending instrument therefore does not allow for attribution of specific project outputs or outcomes at the woreda level exclusively to Bank financing. 298. Third, the Request concerned the Anuak of Gambella, an indigenous group that have suffered a long history, in their narration, of abuse, injustice and violence, which in turn has led to a heightened sense of vulnerability and marginalization, leading them, in some cases, to flee from the country. In this context, their sense of being victimized, acutely felt by many, stem from a combination of causes, some overlapping, which are difficult to disentangle. It has not been the intent of the Report to do so. 299. Fourth, the extent of external attention focused on this request heightened tension around the Panel process, making it more complicated than usual, and resulting in a rather lengthy investigation. This combination of methodological and contextual challenges resulted in a report that is less orthodox and unequivocal than is generally the norm. 300. Furthermore, regarding the issue of Indigenous Peoples, Management’s Response states that the GoE “remains uncomfortable with the [World Bank] policy’s potential application in Ethiopia,304” and “the GoE contends that based on its Constitution, all its people are indigenous.”305 Therefore, progress has been slow in reaching a joint understanding with the GoE on the application of the Indigenous Peoples policy. The Panel recognizes and appreciates that Management has been diligently pursuing action on this sensitive issue and, as planned, started to apply OP 4.10 for projects being presented to the Board since early 2013. The Panel also notes that Management indicates in the PAD of a project recently approved by the Executive Board that the decision to trigger OP 4.10 is stated as being “based on the Ethiopian Constitution.” 306                                                              304 Management Response: 17. 305 Management Response: 17. 306 For example, OP 4.10 was recently triggered for the project, PSNP4 that was approved by the Board in September 30, 2014. The PAD states that “OP 4.10 Indigenous Peoples is triggered as it is determined that the physical and sociocultural characteristics of the sites where subproject activities could be implemented, and the people living in these sites, meet the policy requirements. The decision to trigger the policy is also based on the Ethiopian constitution, which recognizes the presence of different sociocultural groups, including historically disadvantaged or underserved communities, as well as their rights to their identity, culture, language, customary livelihoods and socio-economic equity.” (PSNP4 PAD: 25f.) 72 301. The Panel recognizes that the investigation was equally challenging for Management, as well as for the Board of Executive Directors, and thus appreciates the full collaboration and support that it has received from all quarters in pursuing its work.   302. Importantly, the Panel acknowledges the important contribution that the PBS projects (three phases), together with other related basic services projects, have made towards supporting Ethiopia’s achievement of the MDGs, as widely hailed by the international community. This is also considered a positive example of development partnership, collaboration and support, using an innovative mechanism of block grants, in line with the “Paris – Accra – Busan” development effectiveness principles and based on mutual accountability. 303. As stated in the ICR of PBS I, there was a particular set of circumstances in early 2006, including “a political crisis that had resulted in key donors withdrawing direct budget support at a critical time when the Government of Ethiopia (GOE) had committed to significantly expand financing for decentralized delivery of basic services and for improvement of the Millennium Development Goals (MDGs).”307 It was against this background that PBS was developed as an alternative instrument that allowed “the Bank, other donors and the GOE to: (i) follow-through on the commitment to scale-up financing and delivery of basic services, thus protecting the poor from suffering unduly as a result of the political crisis; (ii) support continued strengthening of local government service delivery systems” 308 as well as to sustain aid inflows and avoid a macroeconomic crisis. Accordingly, the IEG Report rates the relevance of the PBS project objectives as substantial, and characterizes the project as “a creative response to a difficult situation.”309 304. To support basic services delivery, the GoE developed a decentralized system of economic governance, focused on the woreda level. PBS III is thus characterized by the use of the financing mechanism of Block Grants, with Bank’s resources co-mingled with those of other development partners and the GoE. Disbursement of the Block Grants depends on the Bank being assured, through semi-annual JRIS missions, that GoE has followed the Core PBS Principles of Effectiveness, Sustainability, Additionality, Equity, Transparency, Fiduciary Probity, and Predictability. 305. While fully appreciating the context of the Project, the Report has brought out the ramifications of applying this mechanism, in terms of its limitations to monitor the flow of resources, both financial and human, in the context of the decentralized system of governance. Additionally, the limitations in monitoring and assessing results in ways that ensure that “the PBS is on the right course and to take corrective actions as soon as possible when results are found off track.”310 306. Reports of villagization, followed by publication of reports linking it to PBS, emerged at a time when the Bank was engaged in the implementation of PBS II, and preparation and appraisal of PBS III. Instead of investigating or acknowledging these linkages, the initial reaction of DPs, including the Bank, was to distance PBS from CDP by not acknowledging links between the two.                                                              307 PBS I Implementation Completion and Results Report (ICR) (2010): 1. 308 PBS I ICR: 1. 309 IEG (2013): x. 310 JRIS Mission Report (May 12-15, 2014): 18. 73 Management responded together with other DPs in the context of the DAG by recognizing the serious implications of CDP and pursuing high level policy dialogue with the GoE, expressing concerns, attempting to obtain independent verification, and even providing “best practice” advice on resettlement activities. 307. The Panel notes that Management has been increasingly active in the context of this delicate policy dialogue, and appreciates the proactive leadership role that Management has played in a difficult context, including through its leadership of the DAG. It is the view of the Panel, however, that by distancing itself from acknowledging the operational interface between CDP and PBS, Management failed to recognize the impact that CDP was having and will continue to have on PBS results, given the relocation of over 60% of households in Gambella. This was not taken into account during Project appraisal or monitoring of PBS III. This happened despite the fact that the ROC clearance of the appraisal for PBS3 noted that “the program objectives of promoting improved access to basic services overlap with the Government’s objectives for its commune program.”311 308. The Panel notes that for the Government of Ethiopia, the stated objective of CDP was always the improvement of livelihoods through resettlement and expansion of basic services as confirmed in a recent letter from Ethiopia’s Ministry of Finance and Ministry of Federal Affairs to DAG, that the “CDP is only one among several approaches pursued to improve the livelihoods, expand basic services and build grassroots institutions of communities. This is clearly outlined in Agricultural and Rural Development Policy and Strategy, as well as in the GTP.” 312 309. The Panel has found that there is an operational interface between CDP and PBS as developed in Chapter 2. From a development perspective, the two programs mutually influence the results of the other, have the objective of providing improved basic services to the same populations, operate in the same geographical areas, and overlapped during a span of more than three years (2010-2013) when they were implemented concurrently. 310. Regarding the four issues of harm alleged by the Requesters, the first two (involuntary taking of land and use of force and intimidation) are alleged consequences of the resettlement under CDP but not of PBS. Since the Panel found that the Resettlement Policy does not apply to PBS, the Panel concludes that the Bank is not responsible for these alleged harms. Although the last two harms (lack of services at relocation sites and deterioration of livelihoods), are not a direct consequence of PBS either, it is the view of the Panel that given the operational overlaps between CDP and PBS in the context of the provision of these services, this operational interface should have been recognized and addressed throughout appraisal and during monitoring; if this had happened, there may have been opportunities for PBS to mitigate and manage some of the negative consequences of CDP, especially in the earlier years. 311. In reviewing the first two allegations of harm, specifically the claim of human rights violations, the Panel had stated in its eligibility report that the “Investigation will not seek to verify allegations of specific human rights abuses linked to VP, nor will it examine the underlying                                                              311 PBS 3, Decision Note, ROC meeting: 3. 312 Letter from the GoE to the DAG, dated March 18, 2014. Subject: - Reply to DAG Findings and Recommendations on CDP and South Omo (DAG/OU/3/2014A, 18 March 2014): para. 2. 74 purposes of VP”313 as the Panel does not see this to be within its mandate.314 In the course of the Panel’s review of documents and its interviews in the field, however, the Panel came across information regarding the above allegations, and these are presented in the Report without attempting to verify them or otherwise. 312. The issue at hand for the investigation, however, is not in determining the veracity of these allegations, which as stated are alleged consequences of CDP and not of PBS; what is relevant is that these allegations were being widely raised in-country during the time of appraisal of PBS III. The Panel found that the Project ORAF referred to the villagization program as a potential “reputational risk” only, with the consequent risk mitigation measures limited to coordinated responses and policy dialogue. In the Panel’s view, the reliance on dialogue to manage the high level stakeholder relationships is to be expected in a project of this nature, and the Panel recognizes Management’s increasingly proactive role in this regard, but the Panel also found that the approach was imprecise, lacking milestones for success and clear responsibilities. This is in non-compliance with the standards of a systematic or holistic assessment of risk as called for broadly in OMS 2.20 and in the ORAF Guidance. 313. The Panel understands and appreciates that, more recently, Management is currently supporting a study on the interface between World Bank-funded and nationally-funded programs to ensure coherence in their implementation.315 The Panel welcomes this development and expects that the study will lead to necessary adjustments in project implementation, when called for. 314. On the alleged harms resulting from lack of services at the relocation site and deterioration in livelihoods, there is wide agreement, including by DPs and the Ethiopian Ombudsman, that services were not adequate, especially in the first years of the CDP program. The Panel notes that there is broad recognition that the speed of the CDP has not been met with commensurate resources and capacity to provide basic services. During the Panel’s visit, some improvement in services was observed, especially in the education sector. With regards to the alleged harm pertaining to the deterioration of livelihoods resulting from CDP, while the Panel heard some positive comments, many affected villagers informed the Panel how they have lost access to their farms, or had to travel long distances back to their former farms, in order to survive. Reports of the DPs also raised serious concerns around affected livelihoods. 315. The Panel focused specifically on the basic services sector of agriculture, as linked to the livelihoods issue, also recognized as a concern by DPs. The Panel notes that the Regional Government of Gambella considers CDP to be primarily an agricultural program that addresses livelihood security issues. It is the view of the Panel that access to services in the agriculture sector in Gambella under PBS, along with the possibility of achieving the higher level objectives of increasing agricultural productivity, was adversely affected by CDP through inadequate availability of workable land at the new resettlement sites or sometimes, restrictions in accessing old farms. The Panel also found that restricted access to the river and forests had a negative impact on the Anuak, for whom these played a central role in their livelihood strategies. It is the Panel’s                                                              313 Report and Recommendation: 21f. 314 1999 Clarification, para. 13. If the request alleges a material adverse effect and the Panel finds that it is not totally or partially caused by Bank failure, the Panel’s report will so state without entering into analysis of the material adverse effect itself or its causes. 315 Ethiopia – Country Partnership Strategy Progress Report (2014): 13f. 75 position that while the PBS results indicated by the number of DAs and the number of direct beneficiaries may have been achieved at the national level, little can be said about the effectiveness and quality of the service delivered specifically for Gambella.316 As noted by most villagers the Panel met with, access to the services provided by DAs had minimal effect when beneficiaries could not make use of such services for lack of workable land nearby. This was also recognized by DPs in 2012. 316. The lack of recognition of the operational interface at appraisal therefore resulted in inadequate attention during the implementation stage to the consequences of CDP on the achievement of agriculture sector PDOs. It is the view of the Panel that the concurrent rollout of CDP, with overlapping objectives concerning improved delivery of services and agricultural productivity affects significantly the operating context of PBS III in the regions where CDP is being implemented, and in this particular instance, Gambella, where 60% of households are being moved. 317. Regarding the allegation of non-compliance due the non-application of the Indigenous People’s Policy, the Panel concludes that, applying the "functional equivalence" of OP 4.10 was necessary to adequately take into consideration how the customary livelihoods of the Anuak in the implementation of the agricultural services component. 318. The Panel notes that PBS III component 3b (Managing for Results), aims, among other objectives, to undertake targeted surveys and studies that fill crucial gaps in the Government's and DP's understanding of how PBS results can be sustainably achieved. The Panel notes that under this component, an effectiveness review of agriculture, linking financing with results, is to be carried out. While the Panel welcomes this focused attention on agriculture, the review might also address the quality of agricultural services and their impact, especially on livelihoods, and lead to concrete measures that guarantee sustainable livelihoods to resettled people, as stated by DPs already in 2012. 319. The Requesters also raised an issue of non-compliance with the Policy on Involuntary Resettlement. The Panel does not concur with Management that PBS’ focus on provision of resources for recurrent expenditure automatically implies that safeguards are not triggered. The Panel, however, finds that CDP and its resettlement activities were not a necessary activity to achieve the objectives of PBS and thus finds that Management is in compliance with OP 4.12.   320. The Requesters also claim that PBS funding was diverted towards the implementation of CDP. Management states that through extensive project implementation support the Bank tracks eligible expenditure to woreda level. In addressing the question of whether there was diversion of funds from PBS to CDP, the Panel examined three issues: potential for diversion, evidence of it, and capacity and systems built into PBS III to detect any diversion. The Panel finds that there was potential for diversion of funds to the CDP because of the fiscal transfer systems from federal to regional governments and from regional governments to woredas. Once the transfers are made, the weakness of the government’s expenditure system means that it is unable to ensure with full                                                              316 It is noted that the recently published PSIA includes a series of detailed tables on development results in the agricultural sector, e.g., increases in crop production; unfortunately, data for Gambella was consistently missing in the regional breakdowns and thus the Panel was not able to review whether results have been improving for the region. PSIA: 77-94. 76 certainty the uses of funds. While potential for diversion exists, without access to government financial records it was not possible for the Panel to affirm the existence of evidence of such diversion. The Panel experts did not find that PBS III measures to report and monitor PBS-eligible expenditures were sufficiently robust to detect distortions in allocations of resources given the weakness of the financial system. Therefore, neither the Requesters’ allegation that funds were diverted to CDP, nor the assertion by World Bank Management that they can track expenditures to the woreda level, can be supported. 321. Based on information gathered on the situation in Gambella, the Panel finds that the Bank’s appraisal of the Government’s financial management arrangements and their implementation did not comply with Bank policies on project appraisal (OMS 2.20 and OP/BP 10.04). The Bank did not adequately assess the Government’s institutional capacity, include project development objectives (PDOs) to address fiduciary management, or adopt additional PDO results indicators in this area, as recommended by the March 2013 Bank supervision mission to Gambella.317 The Panel also finds that the World Bank did not incorporate lessons from PBS II into the design of PBS III to mitigate fiduciary risk. 322. With regards to the Requesters claim that public servants such as school teachers, health professionals and agricultural extension workers, “were ordered to implement the Villagization program in Gambella,”318 the Panel was unable to find conclusive information on this claim. The Panel emphasizes, however, the operational interface between the CDP and PBS, and potential mutual impacts affecting the achievement of results under PBS were not adequately considered during project appraisal. 323. IEG described PBS as “effectively takes on key features of the Bank’s new Program for Results that links disbursements to defined results.”319 P4R is characterized as supporting government programs and linking the disbursement of funds directly to the delivery of defined results with a special focus on strengthening institutions, which very much fits the description of PBS III. When the P4R instrument was launched in 2012, the Inspection Panel issued observations, appreciating the importance of results-based lending approaches in the context of the principles of aid effectiveness and better outcomes, but raised questions as to whether the instrument fully fostered key dimensions of mutual accountability. 324. In the case of PBS III, the Panel does not question the application of the “quasi-P4R” modality, which is in line with the accepted principles of aid effectiveness. Fundamental to the effective delivery of results and successful outcomes, however, is that the system of delivery is robust and meets the Bank’s environmental and social safeguards principles. 325. In this context, this investigation brought out the challenges of effectively implementing a system in the context of the sub-national level in Gambella that requires capacity strengthening, and in an operating environment where a concurrent program (i.e. CDP) is having a significant impact on PBS operations and results.                                                              317 PBS Secretariat (2013b): 22. 318 Request for Inspection: 6. 319 IEG (2013): xiii. 77 Annex A: Table of Findings   Panel Investigation of the Ethiopia: Promoting Basic Services Phase III Project Issue Findings and Compliance PBS III and CDP: The Panel finds an operational interface between the CDP and PBS programs because Operational of a) commonality of their objectives, b) mutual impacts on each other’s results, c) Interface geographical overlap and d) concurrent implementation. Risk Assessment Considering the magnitude of the operation, the nature of block grant financing, and during Project the overlapping implementation between PBS III and CDP, the Panel finds that Appraisal Management did not carry out the required full risk analysis, nor were its mitigation measures adequate to manage the concurrent roll-out of the villagization program in four PBS III regions. The Panel finds that Management’s approach did not meet the standards of a systematic or holistic assessment of risks, as called for in the Operational Risk Assessment Framework (ORAF) Guidance aimed at identifying adequate risk management measures for affected communities. The Panel finds these omissions in non-compliance with OMS 2.20 on Project Appraisal. The Panel notes that a social assessment would have been justified both to verify the robustness of the results chain that links the block grant mechanism to PBS results, and to assess any changes in the local governance context, which might put future results achievement at risk. Risk Assessment The Panel recognizes Management’s efforts to address social accountability during during Project PBS III implementation. The Panel also appreciates that the establishment of the Implementation grievance redress mechanism and the implementation of the social accountability component of PBS III were fast-tracked in Gambella. The Panel understands and appreciates that, more recently, Management has recognized the issue of the operational interface, and is currently supporting a study on the interface between World Bank-funded and nationally-funded programs to ensure coherence in their implementation. It is the view of the Panel, that the lack of recognition and analysis, at appraisal, of the operational interface between PBS III and CDP, as required by the ORAF and described above, meant that the resulting risks were not adequately taken into account and properly managed and mitigated during PBS III implementation. Application of The Panel finds that the World Bank operational policies and procedures do not Safeguards Policies exclude the application of safeguard policies in Investment Lending (IL) operations providing only for recurrent expenditures. The Panel finds that to the extent that one or more safeguard policies are found to be relevant to the areas of operation of the proposed IL operation, Management should trigger and apply them. Application of OP The Panel does not consider CDP a necessary activity to achieve the objectives of 4.12 on Involuntary PBS III. The Panel finds that paragraph 4 of OP 4.12 is not applicable and that Resettlement Management acted consistently with the provisions of OP/BP 4.12 by not triggering it.     78 Issue Findings and Compliance Application of OP The Panel found no evidence in Project documents that the functional equivalence 4.10 on Indigenous approach was applied to PBS III. In contrast, the PAD clearly noted that World Bank Peoples safeguard policies are not triggered as project financing is limited to recurrent expenditures for basic services. The Panel finds that, barring the triggering of OP 4.10, Management should have adopted the “functional equivalence” approach in the design of PBS III, which is centered on promoting access to basic services. The Panel notes that livelihoods, well-being and access to basic services, which are closely tied to the Anuak’s access to land and natural resources was not taken into account in the design of PBS III, in non-compliance with OP 4.10. Agriculture and The Panel notes that access to services in the agriculture sector in Gambella, along Livelihoods with the possibility of achieving the higher level objectives of the Project (increasing agricultural productivity), was adversely affected by the CDP program as it was implemented. The Panel finds that, in accordance with Bank Policies, the operational interface between CDP and PBS should have been taken into account at the PBS project level, both during the appraisal and implementation phases, especially in a region such as Gambella where 60% of households, which are also PBS beneficiaries, were resettled as part of the Government’s CDP. The Panel finds that Management’s approach has not enabled PBS to mitigate or manage the harms described in the Request for Inspection with respect to access and quality of basic services in the agricultural sector and livelihoods of affected people in Gambella. The Panel finds that, while monitoring was expected to follow the principle of effectiveness and results, the quality of services in Gambella was not in alignment with actual conditions on the ground. In particular, there is no indication that the agricultural extension services that are of paramount importance for the livelihood of rural populations, and especially indigenous groups was being monitored and aligned to deliver appropriate results for beneficiaries. Since PBS III has begun implementation, three JRIS missions were undertaken, but the resulting reports are silent on the issues noted above. The Panel finds that this is not consistent with the supervision provisions of the investment lending policy, OP/BP 10.00. Financial Analysis The Panel finds that Management did not comply with the requirements of OMS 2.20 and OP/BO 10.02 in the design and appraisal of PBS III. The Panel notes that the Bank’s assertion that the funds can be tracked at the woreda level cannot be sustained. The Panel finds that, since PDO results indicators that directly address fiduciary risks were inadequate in the initial planning and subsequently have not been adjusted, the supervision of those risks is not in compliance with Bank policy OP/BP 10.00. Human Resources The Panel notes that it does not have sufficient information to draw a firm conclusion on this issue but it again notes how the operational interface between the CDP and PBS, and mutual impacts affecting the achievement of results under PBS, were not adequately considered at appraisal or during implementation.   79 Annex B: About the Panel   1. The Inspection Panel was created in September 1993 by the Board of Executive Directors of the World Bank to serve as an independent mechanism to ensure accountability in Bank operations with respect to its policies and procedures. The Inspection Panel is an instrument for groups of two or more private citizens who believe that they or their interests have been or could be harmed by Bank-financed activities to present their concerns through a Request for Inspection. In short, the Panel provides a link between the Bank and the people who are likely to be affected by the projects it finances. 2. Members of the Panel are selected “on the basis of their ability to deal thoroughly and fairly with the request brought to them, their integrity and their independence from the Bank’s Management, and their exposure to developmental issues and to living conditions in developing countries.”320 The three-member Panel is empowered, subject to Board approval, to investigate problems that are alleged to have arisen as a result of the Bank having failed to comply with its own operating policies and procedures. Processing Requests After the Panel receives a Request for Inspection it is processed as follows:  The Panel decides whether the Request is prima facie not barred from Panel consideration.  The Panel registers the Request—a purely administrative procedure.  The Panel sends the Request to Bank Management, which has 21 working days to respond to the allegations of the Requesters.  The Panel then conducts a short 21 working-day assessment to determine the eligibility of the Requesters and the Request.  If the Panel recommends an investigation, and the Board approves it, the Panel undertakes a full investigation, which is not time-bound.  If the Panel does not recommend an investigation, the Board of Executive Directors may still instruct the Panel to conduct an investigation if warranted.  Three days after the Board decides on whether or not an investigation should be carried out, the Panel’s Report (including the Request for Inspection and Management’s Response) is publicly available through the Panel’s website and Secretariat, the Bank’s Info Shop and the respective Bank Country Office.  When the Panel completes an investigation, it sends its findings and conclusions on the matters alleged in the Request for Inspection to the Board as well as to Bank Management.  The Bank Management then has six weeks to submit its recommendations to the Board on what actions the Bank would take in response to the Panel’s findings and conclusions.  The Board then takes the final decision on what should be done based on the Panel's findings and the Bank Management's recommendations.  Three days after the Board’s decision, the Panel’s Report and Management’s Recommendation are publicly available through the Panel’s website and Secretariat, the Bank’s Project website, the Bank’s Info Shop and the respective Bank Country Office.                                                              320 IBRD Resolution No. 93-10; IDA Resolution No. 93-6. 80 Annex C: Biographies   Stephen Peterson is a Professor of Public Finance at the Melbourne School of Government. He was a member of the faculty of Harvard University 25 years where he taught graduate courses and the executive program in public financial management, which trained over 1,600 senior government officials. He has worked as a project director, consultant, and technical advisor in nineteen countries in Africa, Asia, Latin America, North America, and the Middle East, plus Australia. His work in PFM encompasses medium-term expenditure planning, fiscal transfers, budgeting, accounting, treasury, and financial information systems. He advised the Palestinian National Authority and with the U. S. Treasury Tax Advisory Service advised several governments in Eastern Europe and the former Soviet Union. He was an advisor to the Government of Kenya for eight years (1986-1994). In 2008, he completed a twelve-year reform in Ethiopia that changed that country’s budgeting, accounting, cash management and financial information systems. He is the author of numerous publications on public sector and financial reform including his forthcoming book on the Ethiopian PFM reform: Public Finance and Economic Growth: Lessons from Ethiopia’s Reforms (Routledge, May 2015). He holds a doctorate from the University of California at Berkeley and an MBA in finance and accounting from the Anderson School of Business, UCLA. Roberto Mosse is presently a retiree-pensioner of World Bank Group (WBG) but continued to work as a consultant to the Bank. While a consultant to TWBG from April 1999 to the present he worked mainly in East Asia and Latin America and provided analytical evaluations of country portfolios; carried out country and project financial management accountability and fiduciary assessments; designed project financial management and institutional arrangements; supervised project implementation; and designed and carried out project performance monitoring and evaluation assessments. During his employment at the WBG from 1972 through 1999 he held various responsibilities, the last one from 1993 through 1999 as Operations Adviser to the Managing Directors for Operations. Previously he was a Senior/Principal Operations Officer, Financial Analyst and Team leader for transportation, industrial and energy privatization projects in Latin America. Prior to joining the WBG, he worked as a Public Accountant for Wallace, McMullin and Partners Chartered Accountants in Melbourne, Australia, and for Mauricio Hochschild/Mantos Blancos in Santiago, Chile. Biographies of Dr. Alison Evans and Dr. Eisei Kurimoto to be provided prior to publication of this Report. 81