Document of The World Bank Report No: ICR2963 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-73440) ON A LOAN IN THE AMOUNT OF US$ 49.296 MILLION TO THE STATE OF BAHIA WITH THE GUARANTEE OF THE FEDERATIVE REPUBLIC OF BRAZIL FOR THE BAHIA POOR URBAN AREAS INTEGRATED DEVELOPMENT PROJECT – VIVER MELHOR II April 30, 2014 Sustainable Development Department Brazil Country Management Unit Latin America and the Caribbean Region CURRENCY EQUIVALENTS (Exchange Rate Effective March 28, 2014) Currency Unit = Real 1.00 = USD 0.44 USD 1.00 = 2.26 ABBREVIATIONS AND ACRONYMS AVSI Associazione dei Volontari per il Servizio Internazionale (Association of Volunteers in International Service) CAS Country Assistance Strategy CBO Community-Based Organization CONDER Companhia de Desenvolvimento Urbano (State Company for Urban Development) CPS Country Partnership Strategy FM Financial Management IBGE Instituto Brasileiro de Geografia e Estatística (Brazilian Institute of Geography and Statistics) IFR Interim Financial Report MDG Millennium Development Goal PAC Programa de Aceleração do Crescimento (Growth Acceleration Program) PDO Project Development Objective PEB Plano Estratégico da Bahia (Bahia Strategic Plan) PFL Partido da Frente Liberal (Liberal Front Party) PHRD Japanese Policy and Human Resources Development Fund PIF Poligonal de Intervenção Física (Physical Intervention Area) PIS Poligonal de Intervenção Social (Social Intervention Area) PMCMV Programa Minha Casa Minha Vida (My House My Life Program) PNAD Pesquisa Nacional por Amostra de Domicílios (National Survey of Sample Households) PT Partido dos Trabalhadores (Workers Party) SEDUR Secretaria de Desenvolvimento Urbano (State Secretariat for Urban Development) SINAPI Sistema Nacional de Pesquisa de Custos e Índices da Construção Civil (National System for Research of Civil Construction Costs and Indexes) UGP Unidade de Gerenciamento do Projeto (Project Management Unit) UTP Unidade Técnica do Projeto (Project Technical Unit) Vice President: Jorge Familiar Calderon Country Director: Deborah L. Wetzel Sector Manager: Anna Wellenstein Project Team Leader: Alessandra Campanaro ICR Team Leader: Alessandra Campanaro BRAZIL Bahia Poor Urban Areas Integrated Development Project – Viver Melhor II CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design ........................................................... 1 2. Key Factors Affecting Implementation and Outcomes ........................................................... 5 3. Assessment of Outcomes ...................................................................................................... 13 4. Assessment of Risk to Development Outcome ..................................................................... 17 5. Assessment of Bank and Borrower Performance.................................................................. 17 6. Lessons Learned.................................................................................................................... 18 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners ....................... 21 Annex 1. Project Costs and Financing ...................................................................................... 22 Annex 2. Outputs by Component.............................................................................................. 24 Annex 3. Economic and Financial Analysis ............................................................................. 26 Annex 4. Bank Lending and Implementation Support/Supervision Processes ......................... 33 Annex 5. Beneficiary Survey Results ....................................................................................... 35 Annex 6. Stakeholder Workshop Report and Results ............................................................... 36 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ................................. 37 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ................................... 38 Annex 9. List of Supporting Documents................................................................................... 39 MAP A. Basic Information Bahia Poor Urban Areas Country: Brazil Project Name: Integrated Development Project ID: P081436 L/C/TF Number(s): IBRD-73440 ICR Date: 03/05/2014 ICR Type: Core ICR Lending Instrument: SIL Borrower: STATE OF BAHIA Original Total USD 49.30M Disbursed Amount: USD 49.30M Commitment: Revised Amount: USD 49.30M Environmental Category: A Implementing Agencies: CONDER - Companhia de Desenvolvimento Urbano do Estado da Bahia SEDUR - Secretaria de Desenvolvimento Urbano do Estado da Bahia Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 10/27/2004 Effectiveness: 06/20/2006 07/06/2006 Appraisal: 04/26/2005 Restructuring(s): 01/04/2011 Approval: 12/06/2005 Mid-term Review: 06/11/2010 12/19/2008 Closing: 01/31/2011 10/31/2013 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Unsatisfactory Risk to Development Outcome: Substantial Bank Performance: Moderately Unsatisfactory Borrower Performance: Moderately Unsatisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Unsatisfactory Government: Moderately Unsatisfactory Implementing Quality of Supervision: Moderately Satisfactory Moderately Satisfactory Agency/Agencies: Overall Bank Overall Borrower Moderately Unsatisfactory Moderately Unsatisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Project at No Quality at Entry (QEA): None any time (Yes/No): Problem Project at any time Quality of Supervision Yes None (Yes/No): (QSA): DO rating before Moderately Closing/Inactive status: Unsatisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General water, sanitation and flood protection sector 23 23 Housing construction 20 20 Other social services 25 25 Sub-national government administration 9 9 Urban Transport 23 23 Theme Code (as % of total Bank financing) City-wide Infrastructure and Service Delivery 17 17 Improving labor markets 17 17 Land administration and management 17 17 Other social protection and risk management 16 16 Urban services and housing for the poor 33 33 E. Bank Staff Positions At ICR At Approval Vice President: Jorge Familiar Calderon Pamela Cox Country Director: Deborah L. Wetzel John Briscoe Sector Manager: Anna Wellenstein John Henry Stein Project Team Leader: Alessandra Campanaro Ivo Imparato ICR Team Leader: Alessandra Campanaro ICR Primary Author: Klas B. Ringskog F. Results Framework Analysis Project Development Objectives (from Loan Agreement) The objective of the Project is to reduce urban poverty in a sustainable manner in the poorest and most vulnerable sections of Salvador and certain other cities in the Borrower’s territory by providing access to basic services and improved housing and social support services. Revised Project Development Objectives (as approved by original approving authority) (a) PDO Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years Increase in the index of satisfaction with basic sanitation conditions in the intervention Indicator 1 : areas (baseline 2010) Value Index of 2.18 on a scale of Index of 2.13, or 25% (increase above quantitative or 0-5, or 31.26% of people 12.31% of people the baseline) Qualitative) satisfied. satisfied. Date achieved 06/01/2012 10/31/2013 12/27/2013 Comments Not Achieved. Sample survey carried out in Mangueira and Encosta de Pirajá revealed (incl. % overall dissatisfaction with basic sanitation. Possibly linked to the fact that works remain achievement) uncompleted. Increase in the index of satisfaction with the provision of new social equipments in the Indicator 2 : intervention areas (baseline 2010) Value Index of 0.33 on a scale of Index of 2.04, or 25% (increase above quantitative or 0-5, or 2.28% of people 27.12% of people the baseline) Qualitative) satisfied satisfied Date achieved 06/01/2012 10/31/2013 12/27/2013 Comments Achieved. Value achieved reflects an increase of over 25%. Reflects a potential increase (incl. % in expectations for social facilities which have been built or under construction but not achievement) yet open for public use. Improved infrastructure and social services: Increase in the percentage of households Indicator 3 : within 250m of paved roads Value 8% increase, or 100% 60% (increase above quantitative or 92.5% territory has paved the baseline) Qualitative) road within 250m Date achieved 01/01/2006 10/31/2013 12/27/2013 Comments Achieved. Indicator effectively reaches 100% of population. Given high baseline, 60% (incl. % increase is impossible. achievement) Improved infrastructure and social services: Increase in the percentage of households Indicator 4 : with access to sanitation networks (availability of sanitation services) 20% increase, or total Value 35% (increase above of 89.7% of total quantitative or 74.78% the baseline) households Qualitative) benefitting Date achieved 01/01/2006 10/31/2013 12/27/2013 Comments Substantially Achieved. Given high baseline, 35% increase is not mathematically (incl. % meaningful. However, indicator effectively shows that nearly 90% of the households achievement) benefited from the sanitation networks. Improved infrastructure and social services: Increase in the percentage of households Indicator 5 : within a 1km radius of social equipments (improved park, public plaza, soccer fields, community center) Value 30% (increase above quantitative or 0% 33% the baseline) Qualitative) Date achieved 01/01/2006 10/31/2013 12/27/2013 Comments Achieved. (incl. % achievement) (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised Target approval Completion or Values documents) Target Years A1- Number of households with improved housing conditions (new, improved, Indicator 1 : indemnities) Value (quantitative 0 2000 1009 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Not Achieved. Reflects a conservative assessment of the indicator by the UGP, which Comments did not include 164 cash compensations. Also, it should be noted that the majority of new (incl. % housing units to be constructed by the Project (535 over 661) are expected for delivery by achievement) August 2014. Indicator 2 : A2- Percentage of households in high risk areas or unhealthy conditions resettled Value 135% , or 92% (if (quantitative 0% 60% actual target or Qualitative) considered) Date achieved 01/01/2006 10/31/2013 10/31/2013 Achieved. From a baseline of 983 households originally living in risk areas, a total of Comments 1332 households were resettled, representing 135 % achievement. However, it should be (incl. % noted that over the course of the Project the overall number of households in risk areas achievement) subject to resettlement has increased to 1,448. Using this adjusted target, the achieved value for this indicator would be 92%. Indicator 3 : A3- Number of households removed from high risk areas or unhealthy conditions Value (quantitative 0 983 1332 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Achieved. The original target of 983 removed from at risk areas was surpassed by 349 Comments households, for a total of 1332. An additional 293 households remained in situ, but risks (incl. % resolved through other measures than resettlement, for an effective total of 1,625 achievement) removed from risk. A4- Number of households with secure ownership of their home (able to be benefitted Indicator 4 : with land regularization - habilitado) Value (quantitative 0 2500 419 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Not Achieved. The Borrower expects an increase in the number of households with secure ownership. Comments As of 01/2014, 661 housing units are still pending delivery. From this date on, and as the (incl. % houses are gradually delivered, titles are expected to be issued. Additional titling (1637, achievement) of which 30 in PIF Alagados III; 296 in PIF Sao Bartolomeu; and 1311 in PIF Encosta de Piraja) is underway through efforts to provide formal identification documents to households. Indicator 5 : A5- Number of households benefitting from urban infrastructure interventions Value 0 6332 1093 (quantitative or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Not Achieved. The low achievement is in part due to the delay in the completion of (incl. % works in Mangueira III, Sao Bartolomeu, Encosta de Piraja, and Sussuarana areas. achievement) Indicator 6 : B1- Installation of Local Technical Offices Value (quantitative 0 4 4 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments (incl. % Achieved. Offices installed in Ribeira, Cobre, Pau de Lima and Managabeira. achievement) Indicator 7 : B2- Number of direct beneficiaries of social activities of the Project Value (quantitative 0 12812 14080 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. Extensive social activities carried out for households removed from risk areas, (incl. % resettled, or receiving indemnifications. achievement) Indicator 8 : B3- Number of beneficiaries attending socio-professional training activities Value (quantitative 0 600 904 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. Vocational training offered through courses at official job training agencies (incl. % SENAI and SENAC. achievement) B4- Number of young people (ages11-29) directly benefitting from social activities of the Indicator 9 : Project Value (quantitative 0 420 467 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. Young people benefitted from activities at social services centers, internship (incl. % training, and training to become "communication agents" within their communities. achievement) Indicator 10 : B5- Number of social organizations strengthened by the Project Value (quantitative 0 72 119 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. 86 social organizations benefitted from training activities and 33 from (incl. % technical assistance activities. achievement) Indicator 11 : C1- Number of O&M plans prepared by target municipalities Value (quantitative 0 4 0 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Not Achieved. Feira de Santana completed its O&M plan after Project closing. (incl. % Agreements with municipality of Salvador still pending. Altogether, only these 2 O&M achievement) plans would be relevant to Project activities. Indicator 12 : C2- Percent implementation of tools for Project management Value (quantitative 0 100 100 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. Satisfactory completion of UGP reports, including implementation and (incl. % procurement plans, FMS reports, and GIS systems of various kinds. achievement) Indicator 13 : C3- Number of community meetings held Value (quantitative 0 558 720 or Qualitative) Date achieved 01/01/2006 10/31/2013 10/31/2013 Comments Achieved. About half of community organized by local organizations and another 25% (incl. % by neighborhood groups, with additional meetings held by the UGP to discuss detailed achievement) designs of Project activities and the Local Community Development Plan. G. Ratings of Project Performance in ISRs Date ISR Actual Disbursements No. DO IP Archived (USD millions) 1 06/05/2006 Satisfactory Satisfactory 0.00 2 12/12/2006 Satisfactory Satisfactory 3.96 3 06/04/2007 Satisfactory Satisfactory 3.96 4 12/17/2007 Moderately Satisfactory Moderately Unsatisfactory 3.96 5 06/26/2008 Moderately Unsatisfactory Unsatisfactory 6.63 6 12/16/2008 Moderately Unsatisfactory Moderately Unsatisfactory 6.63 7 06/24/2009 Moderately Satisfactory Moderately Satisfactory 6.63 8 12/08/2009 Moderately Satisfactory Moderately Satisfactory 6.63 9 05/24/2010 Moderately Satisfactory Satisfactory 10.33 10 02/23/2011 Moderately Satisfactory Moderately Satisfactory 16.33 11 08/14/2011 Moderately Satisfactory Moderately Satisfactory 20.13 12 01/09/2012 Moderately Satisfactory Moderately Satisfactory 27.74 13 10/20/2012 Moderately Satisfactory Moderately Satisfactory 37.44 14 06/22/2013 Moderately Unsatisfactory Moderately Unsatisfactory 48.64 15 11/11/2013 Moderately Unsatisfactory Unsatisfactory 49.30 H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Restructuring Disbursed at Reason for Restructuring & Key Approved PDO Date(s) Restructuring in Changes Made Change DO IP USD millions The first Restructuring was carried out in response to the request of the State of Bahia to amend the Loan 01/04/2011 N MS MS 6.63 Agreement for the Bahia Poor Urban Areas Integrated Development Project (7344- BR) to encompass the following ISR Ratings at Amount Board Restructuring Restructuring Disbursed at Reason for Restructuring & Key Approved PDO Date(s) Restructuring in Changes Made Change DO IP USD millions changes: (i) focusing of geographic coverage within two of the eight eligible municipalities; (ii) change in the scope of coverage to scale up overall beneficiary impact; (iii) reallocation of funds between categories and change in the percentage of expenditures to be financed; (iv) an initial extension of the closing date by 12 months to January 31, 2012, with a contingent second 10-month extension of the closing date until November 30, 2012; and (v) revision of the results framework to reflect the changes in geographic coverage and scope. The second Restructuring extended the closing date by 11 months until October 31, 2013 for a cumulative total Closing date extension of 33 months. I. Disbursement Profile 1. Project Context, Development Objectives and Design 1. The Bahia Poor Urban Areas Integrated Development Project (Viver Melhor II, as also referred to by the Borrower as Bahia Dias Melhores) was an emblematic multi-sectorial slum upgrading project for the World Bank urban portfolio in Brazil, following a number of previous initiatives in the State of Bahia. The Project was prepared from 2002 through 2006, approved by the World Bank Board on December 6, 2005, and declared effective on June 20, 2006. The original closing date of the Project was January 31, 2011, but it was extended for a cumulative 33 months through two restructurings, with a final closing date of October 31, 2013. Activities remained incomplete at the time of closing. 1.1 Context at Appraisal 2. Country and sector issues and rationale for Bank involvement. At the time of Project preparation and appraisal, the urban poor in Brazil lived predominantly in unplanned communities characterized by insecure land tenure, lack of adequate physical infrastructure, and lack of access to educational, health and other social services. Their inhabitants were also subject to crime victimization, poor health, unemployment, and other self-perpetuating conditions of poverty, which has been proven to have a strong correlation with social exclusion and informal housing and slums. 1 The State of Bahia was no exception, with roughly a third of its urban dwellers – around three million people – living below the poverty line. 2 3. The Viver Melhor II Project was thus designed to alleviate urban poverty in Bahia’s capital, Salvador, and in another seven strategic municipalities throughout the state. The Project satisfied two of the four vectors in the Bank’s 2003-2007 Country Assistance Strategy (CAS) for Brazil: (i) scaling up programs to provide services to the urban poor; and (ii) expanding assistance for capacity building at the municipal level. 3 4. The Project followed a number of multi-sector investment operations in Bahia. These included the Novos Alagados Urban Upgrading and Social Promotion Project, begun in 1995 and financed under Bahia Municipal Infrastructure Development and Management (PRODUR, P006562), as well as the Technical and Social Assistance Project of the Urban Poverty Reduction Program in the Ribeira Azul Area of Salvador, Bahia (Ribeira Azul, P074058), launched in 1999, and also supported by grants from the Cities Alliance. These operations incorporated small pilot components aimed at broadening the urban upgrading concept through combined physical infrastructure and social services investments. 5. Viver Melhor II intended to build on the previous experience and increased implementation capacity gained by the State through these earlier projects. By incorporating an integrated approach toward slum upgrading through combined provision of basic physical infrastructure with social services, the Project was designed to meet the lowest-income population’s basic needs in a more inclusive and sustainable fashion. 1 This was highlighted by a study carried out by the State of Bahia. It was based on the social exclusion mapping and targeting methodology developed by Professor Aldaisa Sposati in Sao Paulo, and was run for Bahia using data from the 2000 Census. 2 Verner, 2004, (as cited in the Project PAD, October 31, 2005). 3 Country Assistance Strategy (CAS) for Brazil covering 2003-2007 (Report No. 27043-BR), discussed by the Executive Directors on December 9, 2003. 1 1.2 Original Project Development Objectives (PDO) and Key Indicators 6. Project Development Objective (PDO). The objective of the Project was to reduce urban poverty in a sustainable manner in the poorest and most vulnerable sections of Salvador and certain other cities in the Borrower’s territory by providing access to basic services and improved housing and social support services. 4 7. Specific objectives. The specific objectives of the Project were: (i) improved quality of life and asset base for the urban poor through the provision of basic infrastructure, improvement of housing conditions and access to urban land, and regularization of land tenure in slums and unplanned settlements; (ii) improved access to social programs in the areas of income and employment generation, direct support to the families’ survival strategies, health care, education and job training, and initiatives in community security at the local settlement level; and (iii) enhanced state and local government capacity to plan, implement, monitor and evaluate integrated urban poverty reduction strategies involving several different departments, undertake urban upgrading and land delivery, coordinate social delivery on the ground through participatory planning, and monitor and evaluate government programs through poverty and quality of life indicators. 8. Key indicators. The original results framework contains one PDO indicator, “Improvement in the quality of living conditions for target communities,” as measured by the Índice de Vulnerabilidade Social e Urbana (Social and Urban Vulnerability Index), and an additional 19 intermediate indicators corresponding to the three components. Additional key indicators of attainment of the specific objectives included in the PAD were: (i) Improvement in access to basic infrastructure, services and housing conditions for 30,000 households located in informal urban settlements in 8 municipalities; (ii) Increased access to social programs and income generating opportunities and enhanced safety for 78,000 households located in target informal urban settlement areas in 8 municipalities; and (iii) Improved capacity of 6 State Secretariats and 8 municipalities to plan, implement, monitor and evaluate integrated urban poverty reductions strategies. 1.3 Revised PDO and Key Indicators 9. The PDO was not revised as part of the Project restructurings. 5 1.4 Main Beneficiaries 10. The Project’s main targeted beneficiaries were the poorest and most vulnerable segments of populations living in informal settlements in Salvador and in strategic cities in the interior of Bahia. The original number of beneficiaries targeted was 108,000. 11. As a result of Level II Restructuring completed in 2011, the Project’s geographic coverage was reduced to the two municipalities of Salvador and Feira de Santana, which would receive more concentrated infrastructure and social services investments in more densely 4 The wording of the PDO in the Loan Agreement is used, which is almost identical to the PAD wording. 5 For a detailed description of the changes made to the key indicators of the Project, refer to section 1.6, Revised Components. 2 populated territories. While the Project scope was reduced, the total number of intended beneficiaries roughly doubled to 201,000. 6 12. A study highlighting key characteristics of the population further confirmed the difficult economic conditions under which the target population lived: (i) the level of unemployment in the State of Bahia was calculated as 14 percent; (ii) women constituted 74 percent of those self-employed, most dealing with the preparation and distribution of food in the locality; (iii) 54 percent of households earned less than one minimum salary; (iv) 86 percent had no form of Social Security; (v) 86 percent had received no professional training for their area of employment; and (vi) there was no micro-credit available to enable small entrepreneurs to grow their businesses. 7 1.5 Original Components 13. The Project’s original components are were follows: 14. Component A – Urban Infrastructure Delivery, comprising: (i) urban infrastructure investments (drainage, roads, street paving, water supply and sewerage, solid waste management, home improvement, replacement housing); (ii) macro-level infrastructure (such as road links, collectors, and parks); (iii) low cost housing alternatives; and (iv) engineering design and supervision consultancies. 15. Component B – Social Services Delivery, comprising: (i) strengthening of community-based organizations (CBOs), government institutions, NGOs, and CBOs providing local social services; income and employment generation; health and environmental education; direct support to families; sports and cultural education; and crime and violence prevention; (ii) refurbishment or construction of communal facilities such as day-care centers, kindergartens, facilities for primary education and health care, sports and professional training facilities; and (iii) architectural designs and operations and maintenance plans for specific facilities. 16. Component C – Institutional Strengthening and Project Management, comprising: (i) studies of various kinds; (ii) strengthening of institutions active with slum upgrading and urban poverty reduction, both at the local and state government levels; and (iii) project management, including monitoring and evaluation. 1.6 Revised Components 17. The Project’s scope and focus were adjusted through its first restructuring, conducted in January 2011. 8 The first restructuring introduced the following changes (refer to Table 1 for a summary of the changes made): 18. Geographic coverage / scope. The original geographic scope of the Project was reduced from eight to two municipalities (Salvador and Feira de Santana) to focus and intensify the impacts of the physical and social interventions in light of the limited resources available. 6 The revised number of total beneficiaries was calculated along with an in depth re-assessment of the limits of the PISs and PIFs, in alignment with demographic information from the National Census. Refer to section 1.6 and Table 1 for further detail. 7 Kraychete G. (2012), Fomento à Economia dos Setores Populares, Urban Development Secretariat and Urban Development Company, State of Bahia, 2012 8 Detailed reasons leading to the necessity of a restructuring, including a significant increase in the original cost estimations, are described in section 2.2, Implementation. 3 The two cities were chosen for their status as the largest municipalities in the state, their disproportionate share of urban poverty, and greater capacity to implement Project activities. 19. Targeted beneficiaries. Although the number of municipalities was reduced from eight to two, the intensification of activities in Salvador and Feira de Santana, combined with the higher population density of these two cities, resulted in an overall increase in the total number of intended beneficiaries, from 108,000 to 201,000. Specifically, the restructuring significantly increased the number of beneficiaries targeted by Component A’s urban infrastructure works and reduced the number targeted by Component B’s social services delivery. 20. Results framework. The results framework was modified to reflect the change to the Project’s geographic coverage and scope. Modifications were also introduced as a means to clarify previous imprecisions. For instance, the original rather vague PDO indicator of “Improvement in the quality of living conditions for targeted communities” was replaced by five more specific PDO indicators. The original set of 19 intermediate indicators was also reduced to 13 to better reflect the new Project priorities, and particularly to account for the lack of financial or legal commitments provided for education and health activities described in the PAD, resulting in an inability to ensure their implementation.9 1.7 Other significant changes 21. Reallocation of funds among categories and change in percentages of Bank financed activities. Project funds were reallocated to account for several factors, including increase in Project costs, the need to increase Project management funding, and the shift in Project focus away from Component B social services activities. The increase in Project costs was driven by a number of economic factors further analyzed in section 2.2. Project management funding, formerly supported by a grant for EUR 5 million from the Italian government’s Association of Volunteers in International Service (AVSI), ran out. Lastly, Component A and C activities were prioritized and scaled up while Component B was scaled down (refer to Table 1). The amount and percentage of Bank financing under the different disbursement categories was also modified, with an increase from the original 54 to 100 percent of works financed, so as to accelerate disbursement while reducing pressure on the counterpart funding. 22. Extension of closing date. The closing date of the Project was extended three times, through two restructurings. The first restructuring included an initial extension of 12 months, and a second conditional extension of 10 months. The second restructuring included the third and final closing date extension of 11 months to October 31, 2013, for a cumulative total of 33 months. 9 For further analysis of the Project’s results framework, refer to section 2.3, M&E Design, Implementation and Utilization. 4 Table 1: Main Project Changes through Level II Restructuring PROJECT ASPECT ORIGINAL REVISED (January 2011) 8 municipalities: Salvador, 2 municipalities: Salvador and Feira de Santana, Ilhéus, Feira de Santana Geographic scope Itabuna, Juazeiro, Barreiras, Brumado and Teixeira de Freitas 10 Targeted Beneficiaries TOTAL 108,000 201,000 11 Component A 30,000 158,000 Component B 78,000 43,000 12 Total Allocations 82.20 82.65 (IBRD+ CPF, USD M) Component A 51.40 60.66 Component B 19.45 8.01 Component C 7.75 12.20 Results Framework 13 5 PDO indicators (more PDO Indicators 1 PDO indicator 14 specific) 15 16 Intermediate Indicators 19 indicators 13 indicators st 1 Restructuring: Extension 1: January 31, 2012 (12 months); Extension 2: November 30, 2012 Closing Date January 31, 2011 (10 months). nd 2 Restructuring: Extension 3: October 31, 2013 (11 months). 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 23. Ambitious scope and PDO. The Project’s quality at entry was affected by the ambitious PDO, which placed high expectations on what Project activities should achieve in the target communities. The original inclusion of eight municipalities in the geographic scope led to an excessive number of sub-projects which had not been assessed in detail at the time of preparation. Furthermore, the Project was intended to coordinate with other Bank investments in sectors such as education and health, including the Bahia Education Project – Second Phase (P070827) and the Bahia Health System Reform Project (P054119), in order to achieve synergies. However, the process for achieving these synergies remained unclear, and no institutional or legal provisions were made to ensure that they be established. In consideration of these challenges, the likelihood of achieving a sustainable reduction of urban poverty resulting from deep-seated socio-cultural inequities was over-estimated. 10 Component C – Institutional Strengthening, not included in assessment of targeted beneficiaries in the PAD or Restructuring Paper. 11 Of the total 158,000 targeted beneficiaries, 24,000 were targeted to benefit from improved housing, and 134,000 targeted to benefit from community-level infrastructure and facilities. 12 Refer to Annex 1, Table A1-1 for a more detailed breakdown of original and revised allocations by component 13 Original PDO indicator: Improvement in the quality of living conditions for targeted communities 14 Refer to Datasheet 15 Dropped intermediate indicators: Reduction in water-borne or sanitation related illness; Increased trunk infrastructure for W&S and urban drainage for target communities (% of demand); Increased number of children vaccinated; Increased number of children attending preschool; Reduction in child malnutrition; Reduction in incidence of reported crime and violence; Number of municipality staff trained in key policy issues (urban planning and occupation control, SGI, cost recovery, etc.) 16 New indicator added: Number of social organizations strengthened by the Project 5 24. Underestimated Project risks. Although the risks of this approach were analyzed at appraisal, inter-institutional coordination and cooperation across the political spectrum were both underestimated through their “Moderate” ratings. Vertical and horizontal coordination were assumed to exist at the outset among secretariats and municipalities, but without the appropriate assessment of the necessary financial resources, allocations and legal agreements to make such coordination possible. Expectations for the Project’s implementing agency, the Secretariat of Urban Development (SEDUR) were also set very high, especially considering that the Secretariat had only been established as recently as 2003. Lastly, the high social risks linked to resettlement activities, one of the major factors resulting in implementation delays, were not taken into consideration in the PAD Risk Framework. Nor was the amount of time required to carry out such activities. 25. Low level of implementation readiness. Despite the Project’s lengthy preparation period (2002-2006, supported by a grant for USD 1 million from the Japanese Policy and Human Resources Development Fund, PHRD), preparation efforts emphasized analysis and justification of Project activities over the development of practical technical inputs to facilitate implementation. These could have included topographic studies, social cadasters, or detailed technical designs, all necessary for implementation, and some of which could have been carried out in advance. Likewise, there was a limited level of consideration given to ‘readiness’ at entry regarding aspects of the Project that later proved to affect implementation, such as how to deal with crime and violence in the Project targeted areas. This lack of emphasis on concrete technical inputs during the preparation period also contributed to the overall underestimation of Project costs. The technical designs that were eventually carried out needed to be redone in 2008 and 2009 due to the changing conditions of the Project, indicating a general lack of efficiency in the preparation stage. The extended preparation process and lack of implementation readiness also had the effect of increasing beneficiaries’ expectations and drawing out resettlement negotiations, which could not be accomplished in a timely manner. Because resettlement negotiations could not be initiated until the technical designs had been finalized, negotiations began with a significant delay and contributed to the slow speed of implementation over the course of the Project. 26. A model for multi-sectorial urban projects in Brazil. Viver Melhor II followed a series of prior successful Bank-financed urban development operations in Bahia. The Project’s design was therefore driven by the desire to leverage the positive impacts of its predecessors (which focused on integrating physical infrastructure and social services delivery in low-income communities) and thereby create a new emblematic project for the State of Bahia. To implement this comprehensive and integrated approach, the Project’s design focused on two types of geographic territories: the Poligonais de Intervencão Física (PIFs –Physical Intervention Areas), which combined local infrastructure in smaller areas and macro-level infrastructure (such as the Pista de Borda and the São Bartolomeu Park); and the Poligonais de Intervencão Social (PISs – Social Intervention Areas), which focused social activities, institutional strengthening and productive inclusion inputs, the latter intended to benefit a broader spectrum of the population (refer to the map at the end of this report). Through this approach, Viver Melhor II has served as a model for a generation of multi-sectorial operations in Brazil, and is representative of the preparation and implementation challenges of integrated slum upgrading initiatives. 6 2.2 Implementation 27. Shifting political landscape. The October 2006 elections shifted the political landscape in Brazil, both at the national level as well as in the State of Bahia. The landslide victory of the left-wing umbrella party, Partido dos Trabalhores (PT), in the elections effectively ended a long period of Partido da Frente Liberal (PFL) dominance in Bahia. The incoming government that took office on January 1, 2007 sought to reflect the electoral results at the level of the Viver Melhor II Project through: (i) replacing the director of the Unidade de Gerenciamento do Projeto (UGP – Project Management Unit) in SEDUR; and (ii) making adjustments to the Project to meet the changed political priorities, including revising and/or stopping components that were at an advanced stage in the bidding process. 28. UGP and UTP implementation capacity. The abrupt switch of political emphasis negatively affected the capacity of the implementing UGP and Unidade Técnica do Projeto (UTP – Project Implementing Unit) agencies (the later within Companhia de Desenvolvimento Urbano – CONDER), resulting in virtually no disbursements of the Bank loan for two years. Eventually, the political cost of a stalled project grew in light of the high expectations for the Project in target communities. The potential failure of this emblematic and highly visible project led state political authorities to appoint new UGP and UTP directors in SEDUR and in CONDER in late 2008. The new directors had to essentially re-launch the Project, and begin to work toward improving the implementation capacity that had been negatively impacted in the 2007-2008 period. The implementation got started, but ramping up and getting to a steady pace of execution took time. 29. Increase in total Project costs and loss of purchasing power. Over the course of the lengthy preparation and implementation periods, total Project costs increased by 23 percent, from USD 82.2 million in the original design to an estimated (as of Project closing date) USD 100.85 million through completion of the Project’s pending activities (currently scheduled to mid-2015). It is worth registering that the Borrower claims to have invested and budgeted for additional counterpart funding (not officially reported in the World Bank financial management system) to fully complete the Project’s activities 17. 30. Reasons for this significant increase include: (i) high inflation in the construction sector, particularly after the launch of the Programa de Aceleração do Crescimento (PAC) in 2007, which overheated the market and subsequently raised prices; 18 (ii) the lack of detailed design projects at preparation, leading to difficulty in producing reliable cost estimates; and (iii) increased costs of housing construction linked to the use of higher housing standards than originally anticipated, in accordance with national housing requirements. 19 In addition to the increased costs, the Project lost considerable purchasing power due to macro-economic developments in Brazil. The fluctuation of the Brazilian Real against the US dollar (in which the Bank loan was denominated) was substantially higher than anticipated in the PAD, and 17 Refer to Annex 1, Table A1-1: Original and Revised Allocations of Project Funds (PAD vs. Project Restructuring, 2011), in USD (M) and Table A1-2: Allocation of Project Funds through Project Completion, in USD. 18 Construction costs increased much beyond what the price contingencies of 4 percent of base project costs. The construction price increased 73 percent from 2005-2013, according to the Sistema Nacional de Pesquisa de Custos e Índices da Construção Civil (SINAPI). Investments of the PAC program therefore mirrored but also competed by the types of activities proposed by the Bank. Additionally, the consumer price increased 50% in the same period. 19 While the original intention was to build just an “embryo” home, which could be expanded later by the beneficiaries, the Project had to change its designs to produce homes of 42m2 in accordance with national standards for housing construction programs such as PAC and Programa Minha Casa Minha Vida (PMCMV). In some instances, the finishing details on built houses were upgraded to match the required standards of houses built under the said programs. Thus, while the PAD had assumed a unit cost of BRL 7,000, the housing units built in the years 2011-2013 range from BRL 55,000 to as high as BRL 70,000. 7 the loan lost about 29 percent of its value due to Real appreciation. Additionally, the delayed implementation translated into later disbursements than projected, with the result that higher inflation eroded the purchasing power of the Bank loan in Brazil (since about 99.85 percent of project expenditure was in local currency). The sharp increase in costs required a forced increase in counterpart funding from the original USD 32.9 million to a total of 62.5 million, or an 89 percent increase. 31. Level II Project Restructuring. The Level II Restructuring was carried out as an attempt to redirect activities in a more realistic manner, as well as adjust to the changing circumstances of the Project. The restructuring attempted to account for the loss of purchasing power and adapt to changing political priorities after the 2006 state elections by redirecting financial resources to Component A infrastructure and Component C institutional strengthening activities. The reduced geographic scope was intended to intensify activities and reduce the high risks to implementation involving the original eight municipalities. Lastly, the restructuring addressed the increased burden on the State for counterpart funding by increasing the disbursement percentage of the “Works” loan category from 54 to 100 percent financing by the Bank. The restructuring did not revise the PDO (which could have occurred through a Level I Restructuring) and only addressed shortcomings of the results framework to reflect the Project’s new scope and priorities. As such, the PDO remained ambitious given the investments financed, and the indicators do not adequately reflect the range of activities carried out. 32. Turnover in Project leadership within the Bank. The changes of directors and staff at the Bahia state level were matched at the end of 2008 by the departure of the World Bank task manager who had been associated with the Project since the start of preparation in 2004. A succession of five Bank task managers followed during the remainder of the Project implementation period 2009-2013. The high turnover of Bank task managers implied a loss of institutional memory, and the need for the Borrower to assist each incoming task manager in becoming familiar with the complex project. The frequent change of Bank leadership, however, was partially mitigated by the Brasilia-based co-task manager who remained unchanged for most of the period from 2010 onwards. 33. The nature of resettlement under the Project. Activities related to resettlement became increasingly critical to the implementation process over the course of the Project. First, initiation of resettlement activities was delayed by the lack of technical designs at the beginning of implementation. In order for many of the works to begin construction, resettlement of families living on the sites had to be previously resolved. This was a lengthy process resulting in even further delay. Additionally, the resettlement process required the UTP to adapt to constantly evolving conditions at each site, especially the pressure of new families moving into the intervention areas over the duration of the Project. This also affected the implementation pace, as it added to the complexity of the social work carried out. 34. Crime and Violence. Crime and violence on occasion and to some extent also interfered with Project implementation. As drug gangs were present in the Project areas of intervention, sometimes (especially when there were confrontations between gangs or between gangs and the police) the functioning of the works’ sites had to be temporarily suspended. Site visits of the Bank during supervision missions were arranged in close coordination with the social team working with the targeted communities, which mitigated any potentially sensitive situations. 8 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 35. M&E Design. A comprehensive Monitoring and Evaluation plan was developed by the Bank preparation team to track the progress and impact of the Project. The plan was part of a broader initiative to better assess the impact of Bank projects, Development Impact Evaluation (DIME), where a number of projects were selected as pilots, including the Bahia Poor Urban Areas Integrated Development Project. As originally planned, the Project implementing agency, CONDER, was to use its management information system and special periodic household surveys to provide the data on which it would produce quarterly results reports. Overall, the Project incorporated a sophisticated approach to M&E in its design, which theoretically included a rigorous impact evaluation. The original 19 intermediate indicators reflected the broad scope of the PDO and ambitious goals related to specific sectors, with indicators measuring achievement in areas ranging from health and education to violence reduction. 36. M&E Implementation. The limitations of the original M&E framework became apparent over the course of implementation, reflecting the broader limitations of overall Project design. Through the restructuring, the results framework was therefore modified in an attempt to better reflect the Project’s new focus and concentration of activities. The restructured Project produced a total of 18 indicators, (five of them PDO indicators, 13 intermediary indicators). Reporting on two of the new PDO indicators required that satisfaction surveys be carried out with beneficiaries, while the remaining three required a geographic analysis of the intervention areas. Although the DIME team provided technical assistance during the results framework review that occurred during restructuring, a comprehensive impact evaluation was never conducted as originally planned. 37. Despite these efforts, the modified results framework remained limited and continued to reflect a misalignment with the actual Project activities carried out, as well as a lack of accurate understanding of the possible achievements within such a particular high density urban context. This is partially due to the fact that baseline values were not fully available at the time of restructuring. When achievement of indicators was assessed after Project closing, and the necessary satisfaction surveys and geographic analyses were carried out and reported, many of the target values were found to be not consistent with and uninformative about the Project’s actual achievements. 20, 21 20 For example, the PDO indicator measuring “Improved infrastructure and social services: Increase in the percentage of households within 250m of paved roads” failed to take into account that a 250m radius is much too large for the dense urban context of the Project communities receiving infrastructure works. When the baseline was eventually calculated, it came to 92.5 percent, and thus the previously defined target indicator value of a 60 percent increase above the baseline became mathematically meaningless (by bringing the target value well over 100 percent). At closing, the measured value of this indicator was 8 percent (increase above the baseline), indicating that 100 percent of households now had access to paved roads. A smaller radius would have been more informative of the number of households benefitting from road paving. Similarly, the PDO indicator measuring “Improved infrastructure and social services: Increase in the percentage of households within a 1km radius of social equipment (improved park, public plaza, soccer fields, community center)” failed to consider that a 1km radius may have been too large to reflect a meaningful impact of Project activities, since most of the households were already located within 1km of these social facilities. 21 The satisfaction surveys carried out after the closing of the Project revealed further limitations of the results framework. The baseline indices of satisfaction calculated for the first two PDO indicators, “Increase in the index of satisfaction with basic sanitation conditions in the intervention areas” and “Increase in the index of satisfaction with the provision of new social equipment in the intervention areas” were 2.18 and 0.33 (on a scale of 0 to 5), respectively, both with target values of a 25 percent increase above the baseline. However, a 25 percent increase would bring the indices to 2.73 and 0.41, which would leave the index of satisfaction within the same range on the classification scale as the original baseline, reflecting no real qualitative improvement in the level of satisfaction. While the first indicator was Not Achieved (with an index of 2.13) and the second was Achieved (with an index of 2.04), the structure of the indicators reveals serious flaws in their overall design, and should be taken into account when assessing the results framework. 9 38. Lastly, the results framework fell short in capturing the depth of social activities carried out by the Project, especially those related to resettlement. Resettlement practices developed by the UGP over the course of the Project were later adopted at the statewide level. The UGP also carried out extensive supplemental activities with affected families related to “productive inclusion,” such as vocational training and skills development to support social and economic inclusion within their communities. These activities are not fully taken into account in the indicators provided. 39. Throughout the Project, the Borrower produced satisfactory quarterly reports on intermediate indicators. Information reported was generated by an automated system with the support of CONDER’s GIS department (INFORMS), producing a highly reliable and detailed source of information. However, it should be noted that while the UGP has carried out careful reporting of indicators, it has also undertook a fairly conservative assessment of its own activities to ensure the accuracy of results reported. For example, the intermediate indicator “Number of households with improved housing conditions,” only counts new homes, improved homes, and monitored indemnifications. Families receiving only cash compensation without additional monitoring by the UGP were not counted, as the UGP could not guarantee whether their new homes purchased were an improvement on the previous residence, or how effectively the compensation money was spent. 40. M&E Utilization. The limitations mentioned above have interfered with possibilities for M&E utilization, since the indicators do not reflect the full range and depth of activities carried out. The results framework has therefore limited the Borrower’s utilization of indicators for communication purposes about the achievements of Project implementation. The inadequacies of the framework also forced the Task Team and the Borrower to develop alternative mechanisms for data collection to supply data for the ICR, such as applying surveys to sub-samples of the population that had received infrastructure works. 2.4 Safeguard and Fiduciary Compliance 41. The Borrower has demonstrated a high quality of resettlement, environmental monitoring, and financial management procedures over the course of the Project. 42. Involuntary Resettlement. Resettlement was guided by the goal of moving households out of high-risk areas to new housing located at a maximum distance of 1 km from the original homes in order to minimize the impact on the community’s social fabric and residents’ income-earning possibilities. Resettled households were presented with three options: (i) cash compensation (or indemnification); (ii) assisted purchase (“monitored indemnification”) of replacement housing; and (iii) monitored resettlement to newly constructed houses. The practice of “monitored indemnification” conducted by the Project was considered an important one. It was incorporated to the Borrower’s overall resettlement procedures during implementation, thus revealing interesting results 22. As of Project closing the two key indicators of resettlement had been surpassed. 23 22 For details, refer to section 3.5c. 23 The two indicators surpassed are: (i) 92 percent of households in high-risk areas or living under unhealthy conditions resettled compared to the target value of 60 percent; and (ii) 1,332 households had been resettled from high-risk areas or unhealthy conditions compared to the target value of 983 households. At the time of Project closing a total of 1,750 families had been resettled of whom 62 percent had opted for new housing, 21 percent for cash compensation, 12 percent for assisted purchase, and 5 percent for continued legal proceedings, possibly in the hope of receiving higher compensation. In the only housing development concluded at Project closing, Feira de Santana, only 4 percent (14 out of 358) of families sold their new housing. The national percentage is 12 percent of 10 43. At Project closing there were still 832 pending cases of resettlement, which has since been reduced to 795 as of January 2014. Of these 795 cases, 661 (83 percent) are tied to the delivery of new housing units, a majority of which (535) are expected to be delivered by August 2014, with the remaining units by August 2015. 24 Refer to Table 2 and Table 3 for more detailed information on pending resettlement and on works to be carried out after the closing date. 44. An additional resettlement challenge was the influx of new families into high-risk or unsanitary areas. 25 To mitigate this impact, almost all Project interventions included design features to prevent the encroachment of new unplanned settlements within the Project areas. 26 Table 2: Pending Resettlement Cases by Type as of January, 2014 Pending by Parque Sussuarana São Encosta Mangueira Avenida type of São III / Baixa da Total Bartolomeu de Pirajá III Anchieta compensation Bartolomeu Paz New housing unit 132 270 8 120 131 0 661 Cash compensation 4 5 2 3 11 1 26 (indemnification) Monitored 27 15 11 0 0 0 0 26 indemnification Judicial 47 32 0 0 3 0 82 compensation Total 198 318 10 123 145 1 795 45. Fiduciary Aspects. Procurement routines and the quality of the documents were overall found to be of good quality during Project implementation. There were no major shortcomings identified by supervision. Financial Management (FM) performance is rated Moderately Satisfactory, consistent with the UGP’s timely and reliable provision of information required to manage and monitor the implementation of the Project. The Project’s financial statements (Interim Financial Reports, or IFRs) were submitted on a quarterly basis and overall information has been accurate. All loan proceeds have been disbursed (100 percent) and properly documented by closing date. Counterpart funds had an 84 percent disbursement and execution ratio by the closing date 28. External audit reports arrived at the Bank in time, with an Unqualified Opinion (CY12). 46. Some shortcomings in FM were present throughout implementation: (i) financial statements revealed overall low budget execution, and (ii) counterpart spending after Project closing was below what was estimated by the UGP. During the last missions, the Task Team appropriately followed with the State Government on their commitment to Project resale within a short time period. These families had received their houses free-of-charge and were in practice able to transfer ownership, or at least occupancy, to other families, promptly although there is a notional blocking period of at least three years. 24 These units are part of the second stage of the São Bartolomeu urban upgrading. 25 The number of households living in risk (whose situation was to be addressed by the Project, either by resettlement or by other measures) went up from the original baseline of 983 to 1,448. 26 For example, in the Ribeira Azul area, a road (Pista de Borda) was constructed along the perimeter of the community that could serve as a barrier for future unplanned settlements in high-risk or unsanitary areas. In São Bartolomeu Park, fences were constructed around the border of the park to serve a similar function. 27 See section 3.5c for a description of the monitored indemnification practice. 28 See paragraph 29 and Annex 1 for additional information. 11 implementation, including guarantees that the flow of counterpart funds will not be blocked for the payment of remaining contracts in FY14 and FY15. 2.5 Post-completion Operation/Next Phase 47. As of the closing date, Project activities were not fully completed. However, all but one of the remaining works have been contracted with construction underway, and the Borrower is committed to completing the remaining works. 29 Key issues in the post-completion phase include: 48. Completion of resettlement. There are currently 795 pending cases of resettlement, a majority of which are tied to the delivery of new housing units. All pending resettlement activities, including those involving cash compensation, monitored indemnification, and judicial compensation 30, will continue to be monitored by the Task Team in accordance with OP 4.12. 49. O&M of shared infrastructure. The O&M of São Bartolomeu Park and other facilities such as the nursery school and community center will be carried out in accordance with a Management Plan. The Plan has been fully developed, but final agreements between the State, municipalities and private firms remain pending. 50. Overall O&M and land use control in the Project areas. The State of Bahia and municipalities of Salvador and Feira de Santana will assume responsibility for managing different elements within the Project areas. Sanitation works fall under EMBASA, the state water supply and sanitation concessionaire that has proven capacity for the task. Other infrastructure works, such as management of the roads, fall under the municipalities. Once completed, the constructed works are commissioned and delivered to the respective municipality that commits, in a formal agreement, to provide the necessary financial resources and institutional capacity to operate and maintain the works efficiently and sustainably. By the completion of this ICR, Feira de Santana agreement was signed; Salvador one was still pending full completion of the Project works. While the capacity of these municipalities is generally much lower than that of the State, their status as the two largest cities in Bahia supports the expectation that they will be able to manage these works effectively, despite inherent risks to operational sustainability, especially in areas such as solid waste collection and infrastructure maintenance. Table 3: Pending Works Contracts after Project Closing Pending Works Contracts Expected Date of Completion Safeguard Impacts Urban Integrated Infrastructure Completed. None and Housing Development in Sao December 2013 st Bartolomeu – 1 phase 120 housing units Urban Integrated Infrastructure August 31, 2015 Involuntary resettlement and Housing Development in Sao 136 housing units nd Bartolomeu – 2 phase *not yet bid Completed nursery school None Facilities in São Bartolomeu Park Events area: July 31, 2014 29 The second phase of infrastructure and housing construction in São Bartolomeu is the only contract that has not been bid. Refer to Table 3. 30 Judicial compensation is not a resettlement option. It occurs when agreement is not reached based on the alternatives given to the affected population. In these cases, cash compensation is made available in an account specifically opened for this purpose, while the resettlement process follows the judicial step-by-step procedures. 12 Pending Works Contracts Expected Date of Completion Safeguard Impacts Information Center: July 31, 2014 Urban Integrated Infrastructure August 31, 2014 Involuntary resettlement and Housing Development in 160 housing units Encosta de Pirajá Urbanization of São Bartolomeu Completed. None Park December 2013 Urban Integrated Infrastructure August 31, 2014 Involuntary resettlement and Housing Development in 131 housing units Mangueira III Pista de Borda (Ring Road) and April 30, 2014 None related works Urban Integrated Infrastructure June 30, 2014 Involuntary resettlement and Housing Development in 120 housing units Sussuarana III/ Baixa da Paz 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Rating: High 51. The relevance of the Project Development Objectives, Design, and the method of implementation is rated High because of its close fit with the World Bank Group’s former Country Assistance Strategy (CAS) for Brazil covering 2003-2007 31 (effective at the time of appraisal) and current Country Partnership Strategy (CPS) for Brazil covering 2012-2015. 32 The project also supports Target 11 under the Millennium Development Goal 7 that deals with slum upgrading programs. 52. Additionally, the Project has a high relevance to policies at the state and national level. Subsequent PT Governments have continued to stress increased equity and reduced poverty, both objectives which the Project supported through its clear targeting of the urban poor. The Project directly supported Bahia’s statewide development strategy, Plano Estratégico da Bahia, 2003-2020 (PEB), which had similar objectives to the Bank’s former CAS for 2003-2007, and is aligned with the general objectives of the national PAC program (launched during Project implementation, on January 28, 2007) and its successor PAC 2. For the Northeast region in particular, the PAC program allocated 55 percent of total projected costs to social and urban sectors. 31 Brazil CAS 2003-2007 (Report No. 27043-BR), discussed by the Executive Directors on December 9, 2003. 32 Brazil CPS 2012-2015 (Report No. 63731-BR), discussed by the Executive Directors on November 1, 2011: From 2012-2015, the Bank will focus on: (i) increasing the volume and productivity of public and private investment; (ii) improving the quality of public services for low-income households and expanding their provision through public and private channels; (iii) promoting regional economic development through improved policies, strategic infrastructure investments and support for private sector actors in frontier areas; and (iv) further improving the sustainable management of natural resources and enhancing resilience to climate shocks while maximizing contributions to local economic development and helping to meet rising global food demand. 13 3.2 Achievement of Project Development Objectives 33 Rating: Modest 53. The Project Development Objective was to reduce urban poverty in a sustainable manner, targeting the poorest and most vulnerable sections of Salvador and strategic cities of the State of Bahia with access to basic services and improved housing and social support services. Three out of the five PDO indicators have been achieved. Nine out of the 13 intermediate outcome indicators have also been achieved. 54. The rationale behind the proposed Modest rating lies in the fact that: (i) there was a non- negligible amount of uncompleted infrastructure and resettlement activities by the Project’s closing date; and (ii) Component A’s revised IBRD cost estimates (as per the Project Restructuring) account for 73 percent of the total Project costs, but only two of its five intermediate indicators were achieved. 55. Despite the above, several positive aspects should be highlighted: (i) a majority of pending works are due for completion by end of August 2014; (ii) given the shortcomings of the results framework, not all achievements are accounted for in the PDO and intermediate indicators; (iii) although Components B and C comprise a smaller percentage of Project activities than Component A, overall they reflect satisfactory levels of achievement. 3.3 Efficiency Rating: Modest 56. The economic evaluation during appraisal was conducted using cost benefit analysis, measuring the benefits through a hedonic prices approach. For the ICR a cost benefit analysis was applied as well 34. However, the hedonic prices approach was not used due to lack of information regarding increase in price on the benefited properties caused by the Project. Instead, avoided cost, travel time, and willingness to pay approaches were applied based on gains obtained with the Project. The evaluation was carried out for the following interventions: a) Parks and Cultural center; b) Drainage; c) Daycare facilities; and d) Housing. Total investment evaluated corresponded to about 46 percent of total funds implemented. 57. Actual Costs and benefits were transformed to 2005 prices to make them comparable to those used at appraisal. For doing so, actual prices were adjusted by exchange rate fluctuations and inflation rate. This is very important, as the Brazilian real appreciated during the period and construction price increased as well. Results of the analysis of the loan proceeds show that real cost of the works was 46 percent lower than what was expected. In reality, the overall impact on costs was absorbed by the counterpart funding which increased by 89 percent (or 36 percent of total Project cost) during Project life. It was also absorbed by changes in activities during the restructuring and reduced scope. 58. The summary of results shows that the outcome of the Project has had important benefits for Bahia, as overall benefits are 25 percent higher than total costs and return of the Project on 33 It should be highlighted that this performance assessment was made only against the revised design and results framework of the Project. Although the restructuring process took place more than 4 years into Project implementation, at the occasion disbursement was only at 21% of the total loan amount. It accounted for startup activities, as most of the critical activities leading to the PDO would be implemented during the three final implementation years. 34 Refer to Annex 3. 14 average is 16 percent. This is above the opportunity cost of capital of 12 percent. It is, however, lower than what was anticipated in ex-ante economic analysis of the Project in the PAD (in the range of 34 to 38 percent for the areas of Alagados VI and Pau da Lima, and 16 to 25 percent for the other areas). 59. Additionally, the per unit costs of the housing units, even once brought to the 2005 price, remains high, with an average of USD 11,482 per home, excluding underlying infrastructure. This is relatively high if compared with international benchmarks, 35 highlighting a relatively low efficiency of the Project. Table 4: Costs and Benefits, Total Results Present Value of Flows (000R$) Costs Benefits Net Benefit IRR Park and Cultural 21,461 26,382 4,921 16% Center Drainage 9,660 17,136 7,476 20% Daycare facilities 11,031 11,887 856 13% Housing 15,596 17,340 1,744 13% Total 57,749 72,746 14,997 16% Source: Own elaboration 3.4 Justification of Overall Outcome Rating Rating: Moderately Unsatisfactory 60. Based on the Project’s achievement of objectives (Modest), relevance (High) and efficiency (Modest), the Overall Outcome Rating is MU. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 61. The Project targeted a socioeconomically disadvantaged population in Salvador and Feira de Santana. A 2009 study undertaken by Instituto Brasileiro de Geografia e Estatística (IBGE) within the national household survey Pesquisa Nacional por Amostra de Domicílios (PNAD) estimated the open unemployment at 14 percent and the level of self-employment at 44 percent of the economically active population. 62. A Project study found that 74 percent of the targeted households were headed by a woman, of whom about 52 percent earned a living from preparing and selling food out of her own residence. About 54 percent of households earned less than one minimum salary and 14 percent were assessed to have negative incomes. General education levels were low and 86 percent had not had any professional training for their livelihood, nor was there any access to micro-credit. It is not surprising to find that 80 percent of households had no access to social security. 36 35 Un Millennium Project Task Force on Improving the Lives of Slum Dwellers, 2005, A home in the city. This study estimates an average investment per household of USD 3,000. 36 Kraychete (2012). 15 63. The activities under the Project to train these poor and usually self-employed income earners were welcomed by the beneficiaries who reported raised income levels as they became more proficient in calculating the costs of the goods they offered and in seeking out niches with improved economic prospects. (b) Institutional Change/Strengthening 64. About USD 11.66 million was spent on institutional strengthening activities, corresponding to about 15 percent of total executed Project funds. The funds were spent on consultant services, training and equipment to build SEDUR’s capacity in planning and managing urban development projects. SEDUR had only been created in 2003, but gained sufficient capacity over the course of the Project period to manage and monitor its complex activities. SEDUR worked in tandem with CONDER, a State enterprise that had existed since 1973, and had been successful in implementing earlier Bank-financed projects. (c) Other Unintended Outcomes and Impacts (positive or negative) 65. Monitored indemnifications. The UGP developed new and effective practices related to the Project’s resettlement activities, which have since been adopted as standard practice throughout the State of Bahia. The use of “monitored indemnifications,” whereby the UGP assists families receiving cash compensation through the process of finding a new home, became increasingly popular with affected families over the course of the Project. Monitored indemnifications allowed the UGP to better ensure that cash compensation funds were spent effectively, and that the quality of the new homes purchased reflected an improvement over the former residence. 37 Furthermore, oftentimes they represented an option for accelerating negotiations on resettlement during implementation. 66. Training programs for the self-employed. As a result of the Project’s activities in technical training and assistance for low income self-employed households as a means of raising income levels, the Government of Bahia launched a state-level program, Vida Melhor, to continue and expand these activities. 38 67. Synergy with other development programs. Some sub-projects with technical designs that could not be accommodated under the Viver Melhor II Project (due to the financing shortfalls) were implemented with PAC funding that had very similar development objectives. Similarly, there was synergy with programs financed by the national BNDES and by the IDB in the Ribeira area. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops N/A 37 At the time of Project closing, a total of 162 monitored indemnifications had been carried out. Currently, 26 remain pending (refer to Table 2). 38 As established through Decree No. 13.167 on August 11, 2011 16 4. Assessment of Risk to Development Outcome Rating: Substantial 68. Activities completion risk. Counterpart funding in the amount of USD 25.3 million has been fully allocated in the State budget for the completion of remaining works (refer to Table A1- 2). All except one of the pending construction works have been bid, with the majority scheduled for completion by mid-2014 (refer to Table 3). 69. Sustainability of O&M arrangements. The finalization and signing of formal agreements for the O&M plan for São Bartolomeu Park and related facilities is still pending. Furthermore, operational risks exist tied to the limited capacity of the municipalities to operate and maintain the built facilities, as discussed in section 2.5. 70. Political risk. The upcoming statewide elections in 2014 will bring in a new administration, which may have different priorities and perhaps less political investment in the Project’s pending activities. This scenario could negatively affect the sustainability of the infrastructure built under the Project. 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Unsatisfactory 71. Bank performance during preparation is rated Unsatisfactory. Quality at Entry was compromised by the Project’s overambitious PDO and scope, underestimation of risks, and low level of implementation readiness. The likelihood of accomplishing the Project’s stated goal to achieve a sustainable reduction of urban poverty in the most vulnerable population segments in eight municipalities throughout the State of Bahia was therefore low from the beginning. The Bank ensured continuity in staffing, as the core preparation team of the Viver Melhor II was the same who worked in earlier urban development projects in Bahia. Project preparation was also supported by technical assistance grants from the Cities Alliance (through the Italian AVSI grant), and the Japanese-financed PHRD. This, however, did not prove sufficient to ensure a high Quality at Entry of the Project. (b) Quality of Supervision Rating: Moderately Satisfactory 72. Bank performance during supervision is rated Moderately Satisfactory. Continuity in task management was ensured with the same task manager during the two years of preparation and the first three years of supervision. However, the rest of the supervision period was characterized by high TTL turnover, and this was viewed by the Borrower as a factor negatively affecting implementation. In particular, the Bank provided substantial supervision resources to help the Bahia counterparts rebuild the implementation capacity that had been impaired during the years 2007-2008. As a result, the Bank supervision costs rose from an annual level of around USD 100,000 during the first two years of supervision to USD 140,000 during the subsequent two years of restructuring and capacity rebuilding. Supervision missions were well staffed and scheduled at least every six months. Overall, the World Bank and Borrower teams established a strong working relationship with each other, benefitting from a high level of cooperation on both sides. Specifically, the Bank worked 17 with UGP and UTP to adapt to evolving Project circumstances and solve problems as they came up over the course of the Project, especially in the area of procurement. However, restructuring of the Project had some shortcomings, particularly related to the M&E framework. (c) Justification of Rating for Overall Bank Performance Rating: Moderately Unsatisfactory 73. The Rating of Overall Bank Performance is rated Moderately Unsatisfactory, given the U rating for Quality at Entry and the MS rating for Quality of Supervision. 5.2 Borrower Performance (a) Government Performance Rating: Moderately Unsatisfactory 74. Government performance is rated Moderately Unsatisfactory. Although the ownership of the Project design was high at the outset, and overall the Project was supported throughout the implementation period, the high turnover of counterpart staff during the first two years of implementation motivates the MU rating. The turnover disrupted the continuity from earlier political administrations and impaired implementation capacity. Additionally, the Project was not viewed as a priority by the new administration, further impairing Project implementation. Lastly, the Bahia State government was late in providing counterpart funds in spite of commitments made in the last year of Project implementation, and the delays negatively affected the pace of activities. (b) Implementing Agency or Agencies Performance Rating: Moderately Satisfactory 75. The performance of the implementing agencies is rated Moderately Satisfactory. SEDUR and CONDER exhibited a strong track record in building and applying implementation capacity. There was however a non-negligible inter institutional transaction cost involved in every Project procedure throughout the implementation period, due to the dual arrangement (UGP & UTP). Despite this, and especially after mid-2011 (when internal political alignment within the State was reached), the two agencies were also able to develop a new model of collaboration and coordination which did not exist previously. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Unsatisfactory 76. The Overall Borrower Performance is rated Moderately Unsatisfactory due to the shortcomings during the important first two years of project implementation and delays in providing counterpart funding on a timely basis. 6. Lessons Learned 1. An investment loan dealing with slum upgrading should have an integrated approach to social and physical urban interventions. The Project was correct in matching the improvements in physical infrastructure with social infrastructure delivery and the build-up of capacity of community-based organizations. The deep-rooted socio- cultural causes of the problems of the lowest income populations living in slums require a 18 holistic approach to development. An investment project attempting to tackle challenges of slums should include activities in areas such as education, vocational training, health, crime reduction and violence prevention. These areas however should be properly linked to the project in terms of financial incentives and institutional set up, including all due consents from different levels of administration (to be cleared by project appraisal). For future projects, Bank teams should carefully consider: (i) the right balance and tradeoffs between social and physical interventions, and ‘being comprehensive’ versus ‘being more operationally realistic’; (ii) the sustainability of social interventions after project completion; and (iii) the limited capacity to monitor the impact of social interventions in the medium and long term. 2. Setting realistic expectations, balancing scope and depth of project. Given the long legacy of the development problems and opportunities in slum areas, the design of effective solutions require a coherent long term strategy and adequate time and resources to pursue it. The Bank in particular might need to calibrate its expectations regarding advances that could realistically be achieved in the course of a five-year project (as seen in this and in several other slum upgrading projects in Brazil and elsewhere 39) and clarify what is meant by poverty reduction. Alternatively, the Bank may need to be prepared to undertake longer or multi-phase projects from the start. Assessing the importance of achieving gradual and small scale improvements (such as providing new homes, improving sanitation coverage, or increasing access to educational facilities) could help in setting expectations for the project as a whole, and later in assessing the project’s impact and its contribution to sustainable poverty reduction in a targeted area. 3. It is important to size the right project for the right level of administration. Slum upgrading requires physical interventions whose scope is often within municipal authority. It is risky to support the state to take on housing construction and micro- infrastructure supply (such as roads and drainage), as well as social facilities whose operation will be shared, since these are not always the state’s real area of competence, expertise, jurisdiction and/or strength. In the case of the Viver Melhor II Project, this can threaten the availability of funds for completion of project activities. It can also pose future risks to the sustainability of the investments. Further and more broadly speaking, the relationship between the state and municipalities is often dependent on political alliances which may shift over elections, impacting the implementation and/or sustainability of the project. In light of this experience, the Bank should reframe its urban agenda in Brazil to engage more directly with large municipalities or state capitals, in order to strengthen institutional urban management expertise at the municipal level, notably on aspects related to land use control. On the other hand, the Bank could choose to engage with the States to impact urban development in a larger number of small cities; however, in so doing, not only the type of intervention (micro versus trunk infrastructure) should be carefully evaluated, but also strong institutional and financial linkages (State & Municipalities) should be guaranteed from the beginning. 4. Project preparation of an investment loan should strike the right balance between readiness to implement and flexibility in project design. It is advisable that during 39 Examples are Recife Urban Upgrading Project (P0499265), closed in March 2011 with uncompleted activities after 7 and half years of implementation (including a 2 years extension); Sao Luis Enhancing Municipal Government and Quality of Life Project (P094315) currently under implementation and with projected 7 years implementation period (including 2 years extension). The Barrio Ciudad Project (P088319) in Honduras also implemented (but not fully completed) over 8 years. 19 project preparation, efforts be focused on getting the project ready for an early roll out, by defining activities in detail and dedicating resources to prepare detailed technical designs, a social cadaster of targeted areas, topographies and geo-referencing of works to be undertaken. This will leverage early mobilization of counterpart funding, show early results of the project, help manage expectations among stakeholders, and elevate momentum. On the other hand, it is also advisable to leave some measure of flexibility into the overall design, including dividing works into modules to maximize articulation of project execution and to ensure that some activities can be adjusted over time if conditions (social, political) require it. Flexibility also needs to be taken into account when agreeing on the project’s table of loan categories, amounts allocated, and percentages of expenditures eligible for financing, as this can support faster project execution. 5. Bank engagement in urban and slum upgrading projects should include work on housing policy and finance at local level. When preparing and implementing urban and slum upgrading projects, the Bank should further consider: (i) to prioritize (to the extent possible under OP.BP 4.12) coordinated solutions for resettlement with existing low income financing programs at National or State level, rather than allocating direct financing for it; (ii) to engage and work with the Clients on aspects related to housing policy and finance – development of cost efficient and innovative solutions; and development, regulation and implementation of existing tools and instruments for urban development, land management and private sector participation. 6. Teams dealing with projects targeting at-risk and vulnerable areas should carefully consider variables that may be difficult to control. A project like Viver Melhor II includes a number of variables that are hard to control in terms of the impact they could have on execution – time, cost, effectiveness. At the outset, these variables need to be carefully examined, identified as risks whenever necessary, and mitigation plans need to be set up. Particularly, a team should assess the potential risks and impacts associated with: (i) organized crime and violence in the targeted areas; (ii) different strategies needed to handle different local communities, their expectations and demands; (iii) difficulty in document collection from the targeted population (this had strong impact on titling issuance under the Project); and (iv) resettlement complexity (assessing the availability of compensation options at entry and taking into account regular processing and judicial recourse times is critical). 7. Sequencing in core activities linked to slum upgrading projects could be the key to a successful project implementation. In slum upgrading projects, there is a need for a greater understanding during project preparation of the sequencing of activities in the implementation process, and their impact on the overall project timeline. Specifically, resettlement activities were found to dictate the timeline for project implementation. It is thus fundamental to allocate the right amount of energy during project preparation (and ideally by appraisal) in: (i) defining eligible areas of intervention; (ii) developing reliable topography and soil prospections; (iii) getting at least part of the detailed technical designs completed (the ones for the priority investments); and (iv) preparing social cadasters. Jointly, these will allow for a qualified baseline to assess costs and overall resettlement needs (including compensation options and sources of funds). The project ideally should be ready for a prompt roll out once it becomes effective (streamline of the negotiations with the families, and preparation of bidding documents); or, otherwise, these aspects should be factored in project length (additional 1-2 years of 20 implementation), as well as expected disbursements and communication strategies adjusted accordingly. 8. Bank fiduciary approaches to project supervision could be better integrated, harmonized and congruent to clients’ capacities, in order to reduce project management costs and facilitate day to day routine supervision and implementation. FM and procurement were applied to Viver Melhor II project execution with different underlying modalities and visions. The former had (and in general is characterized by) a more holistic, comprehensive approach that focuses on ex-post ‘remedies’ (for instance funds ineligibility) as deemed necessary; the latter has more of a day to day accompanying approach, focused on step by step processing and contracting. It could be conceived that the two be more integrated and applied holistically, potentially bringing some savings to project management. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies (b) Co-financiers (c) Other partners and stakeholders 21 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) 40 Table A1-1: Original and Revised Allocations of Project Funds (PAD vs. Project Restructuring, 2011), in USD (M) Revised Original Project Original Share Revised Total Share Total Component IBRD (%) IBRD (IBRD + (%) (IBRD + CPF) CPF) A Urban infrastructure 30.83 51.40 62.53 35.79 60.66 73.39 delivery A1 Community 22.01 36.70 44.65 29.66 50.37 60.94 infrastructure A2 Macro-level works 6.75 11.25 13.69 4.87 8.44 10.21 A3 Housing alternatives 0.90 1.50 1.82 - - - A4 Engineering designs 1.14 1.90 2.31 1.26 1.85 2.24 B Social services 11.67 19.45 23.66 4.59 8.01 9.69 delivery B1 Social services 4.47 7.45 9.06 1.31 2.71 3.28 B2 Social service 6.84 11.40 13.87 2.89 4.86 5.88 facilities B3 Social project design 0.36 0.60 0.73 0.38 0.44 0.53 C Institutional strengthening and 4.65 7.75 9.43 8.79 12.20 14.76 Project Management C1 Studies and plans 0.60 1.00 1.22 0.63 0.91 1.10 C2 Institutional strengthening (training, 1.20 2.00 2.43 0.96 1.98 2.40 consultants, equipment) C3 Project management 2.85 4.75 5.78 7.19 9.31 11.26 Contingencies 2.09 3.48 4.23 1.66 2.01 Front End Fee 0.12 0.12 0.15 0.12 0.12 0.15 Total Project Costs 49.30 82.20 100.00 49.30 82.65 100.00 40 Cutoff date is the end of the grace period of the Loan (February 28, 2014). Exchange rate applied was USD 1.00 = BRL 1.87. 22 Table A1-2: Allocation of Project Funds through Project Completion, in USD Project Component IBRD Counterpart Total A Urban infrastructure delivery 35,244,282.51 22,313,566.36 57,557,848.87 B Social services delivery 4,536,668.79 1,682,410.96 6,219,079.75 Institutional strengthening and C 9,391,808.19 2,263,511.79 11,655,319.98 Project Management Front End Fee 123,240.00 0.00 123,240.00 Total Executed 49,295,999.49 26,259,489.11 75,555,488.60 Projected Additional Costs to 41 0.00 25,294,543.98 25,294,543.98 Complete Project Activities Total Project Costs 49,295,999.49 51,554,033.09 100,850,032.58 (b) Financing Appraisal Actual/Latest Type of Percentage of Source of Funds Estimate Estimate Cofinancing Appraisal USD (M) USD (M) Borrower 32.90 51.55 156.00% International Bank for Reconstruction 49.30 49.30 100.00% and Development TOTAL 82.20 100.85 123.00% 41 These amounts are not officially reported in the FM due instruments. They correspond to what has been planned for and allocated in the Borrower’s budget to complete the Project’s activities after the closing date (difference between the total estimated additional counterpart of USD 56.34 million and the executed additional counterpart of USD 31.05 million). 23 Annex 2. Outputs by Component Component A – Urban Infrastructure Delivery Description: This component financed priority infrastructure in poor urban communities such as (a) Infrastructure that would provide basic services such as drainage, roads, street paving, water supply and sewerage, solid waste collection, and housing replacement; and (b) macro-level infrastructure investments to integrate the upgraded settlements into the larger surrounding urban fabric. Level of Achievement: Moderately Unsatisfactory Strengths • The activities and works to reclaim the 1500 hectare São Bartolemeu park are of great ecological, cultural and recreational importance for the entire city of Salvador, not only for the adjacent communities of urban poor. The park was previously under threat to be permanently impaired through encroachment of unplanned communities, which has been prevented by the Project’s upgrading activities in the area; • The constructed urban infrastructure will in many cases benefit a far larger number of households beyond the immediate communities targeted by the Project; • Most of the pending works to be completed are now contracted, and scheduled to be occupied by resettled households by mid-2014, boosting achievement under sub- component A1. Pending • Necessity of integration of the urban infrastructure with the O&M procedures in the Challenges municipalities of Salvador and Feira de Santana to prolong benefits; • Monitoring and resolution of the tendencies for continued encroachment of unplanned urbanization in the Project areas to prevent the emergence of new unplanned communities. There are signs that encroachment is continuing albeit at a relatively low rate. • Addressing education, healthcare and job training/income generation activities to expand the potential for urban poverty reduction in the target communities. Achievement of Intermediate Results defined in the Loan Agreement’s RF for this component: At Project closing, three out of five intermediate outcome indicators for Component A were achieved: A1: 1,009 out of the targeted 2,000 households had been provided with improved housing through (i) receiving new housing; (ii) purchased replacement housing; and (iii) cash compensation. (50% achievement); A2: 135 or 92 percent of the households in high risk areas or unhealthy conditions had been resettled as compared to the targeted 60%. (225 or 153% of achievement); A3: 1,332 households had been removed from high risk areas or unhealthy conditions as compared to the targeted 983 households (136% achievement); A4: 419 households had been provided with secure ownership of their homes as compared to the targeted 2,500 households (17% achievement); A5: 1,093 households had benefitted from urban infrastructure investments as compared to the targeted 6,332 households. (17% achievement). Component B – Social Services Delivery Description: This component financed social investments to strengthen the community based organizations (CBOs), activities to strengthen income through training, and communal facilities such as day care facilities. Level of Achievement: Satisfactory Strengths • A large community center directly adjacent to the São Bartolomeu Park promises to develop into an important hub for educational and recreational activities; • A large public plaza with stalls constructed for small businesses to serve food near the São Bartolemeu community center will provide the resettled households with enhanced possibilities to continue their income-generating food preparation and distribution businesses; • Football fields adjacent to the high crime communities near the São Bartolemeu Park provide needed recreational facilities for the community’s youth; 24 Pending • Crime associated with the drug trade continues strong in the São Bartolomeu area, Challenges which has been identified as a drug transit route Achievement of Intermediate Results defined in the Loan Agreement’s RF for this component At Project closing, all five intermediate outcome indicators for Component B were achieved: B1: Four out of four planned Local Technical Offices had been installed (100% achievement); B2: 14,080 direct beneficiaries as compared to the targeted 12,812 beneficiaries had been registered (110% achievement); B3: 904 beneficiaries attended training activities as compared to the target of 600 (151% achievement); B4: 467 young people (ages 11-29) as compared to the targeted 420 had directly benefitted from social activities of the Project (111% achievement); B5: 119 social organizations strengthened by the Project as compared to the targeted 72 (165% achievement). Component C – Institutional Strengthening and Project Management Description: This component financed a series of activities and goods for (a) building capacity in participating municipalities to enable them to better manage the process of urban poverty reduction projects; (b) a variety of studies to underpin project implementation; and (c) project management including monitoring and implementation. Level of Achievement: Moderately Satisfactory Strengths • The large number of community meetings held to engage and inform the target beneficiary communities on Project progress; • Considerable institutional capacity developed within SEDUR (as a relatively new government agency) to manage implementation of highly complex urban development projects; • Enhanced institutional cooperation between the SEDUR and CONDER agencies, resulting from coordination of activities and resources through the Project implementation process Pending • Maintenance of the Project preparation and implementation capacity within CONDER Challenges and SEDUR. A potential for a greater degree of coordination and integration of nationally financed programs such as PAC 2 may be possible. Achievement of Intermediate Results defined in the Loan Agreement’s PRF for this component At Project closing, two out of three intermediate outcome indicators under Component C were achieved: C1: None of the O&M plans (out of a target of four) have been finalized (0% achievement). However, two months after Project Closing the first such O&M plan was prepared for Feira de Santana for a 25% achievement; C2: 100% of the Project Management Tools had been commissioned (100% achievement); C3: A total of 720 community meetings as compared to the targeted 558 had been held (129% achievement) 25 Annex 3. Economic and Financial Analysis 1. The Bahia Poor Urban Areas Integrated Development Project (Viver Melhor II Project) comprised investments in (i) urban infrastructure services; (ii) social services; and (iii) improved institutional capacity. The project benefited a wide range of the population. There were people directly benefitting from new housing, those benefitting from an improved urban environment and provision of urban services (including streets newly paved, drainage, water and sewerage works, and safeguarded parks); and those benefitting from social infrastructure, such as day-care centers, markets, and community meeting facilities. 2. The economic evaluation during appraisal was conducted using cost benefit analysis, measuring the benefits through hedonic prices approach. A sample of surveyed households were asked to assign values to the houses they occupied (either as tenants or as owners) and, in addition, were asked to report on a large number of urban characteristics in the areas they lived in. Through regression analysis the variations in housing values were then associated with a large number of attributes of urban infrastructure where the variations in the housing prices could be then taken as a proxy for variations in the quality of the houses themselves and the surrounding infrastructure. 3. For the ICR a cost benefit analysis was applied as well. However, hedonic prices approach was not used due to lack of information regarding increase in price on the benefited properties caused by the project. Instead avoided cost, travel time, and willingness to pay approaches were applied based on gains obtained with the project. The evaluation was carried out for the following interventions: a) Parks and Cultural center; b) Drainage; c) Daycare facilities; and d) Housing. Total investment evaluated corresponded to about 50 percent of total implemented. 4. Actual Costs and benefits were transformed to 2005 prices to make them comparable to those used at appraisal. For doing so, actual prices were adjusted by exchange rate fluctuations and inflation rate. This is very important as the Brazilian real appreciated during the period and construction price increased as well. Results show that real cost of the works was 46 percent lower than what it was expected. The overall cost impact was absorbed by the counterpart funding which increased by 89 percent (or 36 percent of total project cost) during project life. It was also absorbed by changes in activities during the restructuring. 5. The summary of results show that the outcome of the project has had important benefits for Bahia, as overall benefits are 25 percent higher than costs and return of the project is 16 percent. This is above the opportunity cost of capital of 12 percent. It is however, lower than what was anticipated in ex-ante economic analysis of the project in the PAD (in the range of 34 to 38 percent for the areas of Alagados VI and Pau da Lima, and in the range of 16 to 25 percent in other areas). 6. Additionally, the per unit costs of the housing unit, even once brought to 2005 price remains high, to an average USD 11,482 per home, excluding underlying infrastructure. This is relatively high if compared with international benchmarks, 42 highlighting a relatively low efficiency of the project. 42 Un Millennium Project Task Force on Improving the Lives of Slum Dwellers, 2005, A home in the city. This study estimates an average investment per household of USD 3,000. 26 7. Summary of results show that the outcome of the project has had important benefits for Bahia, as overall benefits are 25 percent higher than costs and return of the project is 16 percent. Table A3-1: Total Results Present Value of Flows (000R$) Costs Benefits Net Benefit IRR Park and Cultural 21,461 26,382 4,921 16% Center Drainage 9,660 17,136 7,476 20% Daycare facilities 11,031 11,887 856 13% Housing 15,596 17,340 1,744 13% Total 57,749 72,746 14,997 16% Source: Own elaboration Methodology 8. The economic evaluation during appraisal was conducted using cost benefit analysis, measuring the benefits through hedonic prices approach. For the ICR a cost benefit analysis was applied as well; however hedonic prices approach was not used to measure the benefits due to lack of information regarding increase in price on the benefited properties caused by the project. Instead avoided cost, travel time, and willingness to pay were applied based on actual gains obtained with the project. Costs and benefits were included as they actually occurred during and after the implementation. To make costs and benefits comparable with those used at appraisal, they were adjusted to 2005 prices by the impact of inflation and exchange rate differences. Net benefit of the project was estimated as the incremental benefit from two scenarios: with and without project. The without project scenario was projected assuming that the situation existing at time of appraisal remained; while with project scenario was projected including benefits already attained with the works. The cash flows were discounted using the same discount rate used at appraisal, which was 12 percent. Benefits Analysis 9. Among the benefits already seen from the works are: a) drainage works have reducted risks and losses to property due to flooding; b) urban upgrade has shown productivity gains in improved urban areas; c) the park of São Bartolomeu has shown environmental improvement and increase access to recreational activities and training; c) public space infrastructure has allowed access to daycare facilities in the project area; d) road rehabilitation has facilitated access to some communities; and e) improved institutional capacity. 10. At the time of the ICR there was not enough information to evaluate all the components, but the evaluation includes the following: a) Parks and Cultural center; b) Drainage; c) Daycare facilities; and d) Housing. The cost of theses interventions comprises about 50 percent of total project cost. 11. The following approaches were used to measure the benefits: a) avoided damage cost for the drainage component; b) willingness to pay and saving costs were applied for the park; and c) travel time and savings costs for daycare facilities. Results were added to avoid double counting benefits. The avoided damage costs were based on data from previous flooding events. Travel time and savings costs were based on surveys carried out on the field 27 comparing with and without project situations. Willingness to pay was based on the case studies conducted by Hildebrand (2001) to estimate benefits of urban parks in Curitiba. 43 Costs Analysis 12. Costs consisted of actual costs of the works plus the institutional component, which is 18 percent of works costs. For this evaluation, this percentage was added to the actual cost of each component. 13. Without including counterpart funds, expected costs at appraisal were R$ 127M (USD 49M) and actual costs were R$ 90M (USD 49M) which corresponds to 0 percent increase when expressed in US dollars, and a 29 percent decrease when expressed in Reais. However costs at appraisal are expressed in 2005 prices, while actual costs are expressed in prices of the implementation year, and therefore they are not comparable and the resulting price increase is not accurate. Table A3-2: Investment costs (without counterpart), Expected at Appraisal and Actual Costs At appraisal Actual costs (000 USD) (000 R$) (000 USD) (000 R$) Urban Infrastructure 35,243 64,771 Delivery Social services delivery 4,537 8,338 Institutional Strengthening, studies, 9,515 17,487 contingencies Total Project Costs 49,300 127,18 49,295 90,596 Average Exchange Rate 2.58 R$:USD 1.84 R$:USD 14. During the period 2005-2013, the currency exchange rate fluctuated significantly and the inflation rate also rose. As Figure A3-1 shows, the exchange rate went from R$ 2.58 in 2005 to R$ 1.61 to USD in 2008, and then R$ 2.01 in 2013; and the construction price index increased 73 percent during the period. 44 43 Hildebrand, Elisabeth (2001). Avaliacao Economica dos Beneficios Gerados por los parques urbanos: Estudo de Caso en Curitiba- PR. 44 According to IBGE, Consumer Price Index during the period was: 2005: 5.69%, 2006: 3.14%, 2007: 4.46%, 2008:5.90% , 2009: 4.31%, 2010: 5.91% , 2011: 6.50%, 2012: 5.84% , 2013: 5.91%. Construction Price Index was: 2006: 5.4%, 2007: 5.2%, 2008: 9.1% , 2009:7.7% , 2010: 6.5% , 2011: 7.75%, 2012: 7.0% , 2013: 8.0%. According to the Brazilian Central Bank Annual average exchange rates: 2005:2.53; 2006: 2.25, 2007: 1.93, 2008: 1.62; 2009: 21.96; 2010: 1.81; 2011: 1.59; 2012: 2.05; 2013: 2.17. 28 Figure A3-1: Actual Implementation Costs Source: Own elaboration 15. To compare expected and actual cost, actual prices of investment were transformed to 2005 prices eliminating exchange rate variation and inflation rate. Currency fluctuation is measured adjusting prices with inflation indexes and exchange rate fluctuation; any remaining amount corresponds to real increase of works. 16. If we run the analysis focusing on the WB proceeds (without counterpart funds) - results show that: a) 29 percent increase was due to the appreciation of the Brazilian Real to the USD. At appraisal the exchange rate was about R$ 2.58 per USD and the average obtained during implementation period was 1.84; b) 44 percent increase was due to inflation rate. This inflation was calculated as the average between construction price index and consumer price index. The construction price increased 73 percent from 2005-2013, while the consumer price increased 50 percent in the same period. The works were implemented along the period with an average price increase of prices of 44 percent (construction prices 46 percent and consumer prices 31 percent); and c) the remaining 46 percent increase is explained in higher costs of works. Table A3-3: Investment Cost, Expected and Actual Appraisal Actual Increase % USD (000) 49,295 49,295 0% Breakdown of the cost increase: Appreciation of the Brazilian 2.58 1.84 29% Real Inflation Rate 44% Real cost increases -46% Source: Own elaboration Evaluation of Interventions São Bartolomeu Park and Cultural Center 17. The interventions around the São Bartolomeu Park consisted of: a) enhancement of urban infrastructure around a park of 150 hectares, b) construction of a cultural and educational center for the community and additional facilities for park management. The park does not 29 currently have an entry fee. Once the Park will be fully operational, the flow of expected visitors is estimated at 312,000 people. These people will come to relax, enjoy natural attractions offered by the park, participate to local festivities linked to São Bartolomeu (in July each year a big festival is held for instance) and so on. The cultural center has benefitted from the acquisition of diversified equipment; it has a service provision digital center for citizens, dedicated space for dance and theatre classes, manual arts classes, music classes, digital production room, a library, a play center and toy library, as well as dedicated space for exhibitions. It is estimated that about 156,000 people per year will visit the facilities. Before implementing the project there was no park or training facility in the neighborhood. 18. This evaluation used two parameters to measure the benefits: a) willingness to pay for the park; b) price of renting rooms with same characteristics of those available at the educational center; c) gains on time when distances to a training center decreased. Particularly, the considerations below were used to build assumptions for the analysis: a) Willingness to pay based on the case studies conducted by Hildebrand (2001) to estimate the benefits of urban parks in Curitiba, Brazil. Results show that people were willing to pay an average of $ 2.10 per person, which corresponds to R$ 2.98 in 2005 prices. b) Renting prices. The educational center has been used for courses and events. It is reasonable to assume that if the education center had not been built, other spaces would have to be rented in other locations. The current facility has 4 meeting rooms and 2 auditoriums. The rooms are used weekdays all yearlong; while the auditoriums are used weekdays during the school calendar, which is 9 months per year. The price of renting a room of the same characteristics is estimated to be R$ 500 per day. The prices were transformed to 2005 prices and applied to the number of rooms and auditoriums. 19. Results of the analysis show benefits 23 percent higher than costs and return of 16 percent. Table A3-4: Results, São Bartolomeu Park Present Value of Flows (000R$) Costs Benefits Net benefit IRR Total 21,461 26,382 4,921 16% Drainage 20. The drainage component contributed to reducing the risk of flooding in the targeted areas. Benefits were measured in the form of losses that are averted due to floods mitigation works implemented with the project. Damage costs were estimated for residential properties based on information from previous events. The evaluation was carried out for five localities with a total of 3,158 beneficiaries. Table A3-5: Drainage Beneficiaries Total drainage Beneficiaries Sao Bartolomeu 500 Avenida Anchieta 320 Encosta de Piraja 1,983 Mangueira 146 Sussaurana/Baixa da 209 Paz Total 3,158 30 21. Previous events showed average damage cost per house of R$ 500 per year and repair costs of R$ 7,000 every five years. Additional cost of time was included of about 6 days per year per household, to cover the time lost to attend the emergency and risk of illness caused by flooding. Assumed time cost was the average wage in the zone of R$ 66 per day with 50 percent adjustment to include kids, for a total of R$ 33 per person per day. 22. Results of the analysis show that benefits double the costs and return is 26 percent much higher that 12 percent discount rate. Table A3-6: Results, Drainage Total Present Value of Flows (000R$) Costs Benefits Net benefit IRR Sao Bartolomeu 2,383 7,056 4,674 28% Avenida Anchieta 2,303 4,960 2,657 24% Encosta de Piraja 1,599 1,542 (57) 11% Mangueira 123 387 264 47% Sussaurana/Baixa da paz 3,251 3,191 (61) 12% Total 9,660 17,136 7,476 20% Daycare Facilities 23. The evaluation was made based on the cost of time and number of kids. Before the project was implemented kids and her parents had to travel 6 km to get to another daycare facility. About 300 kids have benefited with the facilities (between day care, and after hours schooling). Cost of time was the same used for the other interventions. 24. Results of the analysis show returns of 13 percent and benefit 10 percent higher than costs. Table A3-7: Results, Daycare Facilities Present Value of Flows (000R$) Costs Benefits Net benefit IRR Total 11,031 11,887 856 13% Housing 25. The average cost per house built or improved is as follows: Table A3-8: Housing Costs Actual cost per Actual cost per Actual cost per unit Houses unit R$ unit 2005 R$ 2005 US Sussuarana 120 54,553 35,588 13,794 Mangueira 131 56,702 36,990 14,337 Encosta de Piraja 160 69,337 45,233 17,532 Feira de Santana 358 19,189 12,518 4,852 Sao Bartolomeu 120 70,254 45,831 17,764 Total 889 45,409 29,623 11,482 26. All of these houses were located at risk areas, and some had severe damages when flooding. For the evaluation it was assumed (conservatively) similar damage to those in the drainage 31 component, even though the damages were more severe. The time lost cost was assumed higher, 10 days repairing damages and 3 workers needed. Results of the analysis show return of 12 percent and economic benefits of R$ 273,000. Table A3-9: Results, Housing Present Value of Flows (000R$) Costs Benefits Net benefit IRR Total 13,167 11,887 273 12% 27. It is worth noting that for a good part of the units produced, the investment cost includes house interior features upgrade and possibility of expanding the size of the houses (up to doubling the entire size of the homes) through assigned adjacent space already pre-set for such purpose (with electricity, water and other utilities already connected). Therefore potential benefits for the units delivered go beyond those presented here. Overall final analysis 28. Considering all elements presented above, the summary of results benefits 25 percent higher than costs and overall return of investment is of 16 percent. Table A3-10: Total Results Present Value of Flows (000R$) Costs Benefits Net benefit IRR Park 21,461 26,382 4,921 16% Drainage 9,660 17,136 7,476 20% Daycare facilities 11,031 11,887 856 13% Housing 15,596 17,340 1,744 12% Total 57,749 72,746 14,997 16% 32 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Financial Susana Amaral Financial Management Specialist LCSFM Management Judy L. Baker Lead Economist EASIN Economist Fernando Andres Blanco Cossio Senior Economist AFTP4 Economist Jose Augusto Carvalho Consultant LCSPT Eleoterio Codato Sector Manager AFMAO Financial Tulio Henrique Lima Correa Financial Management Specialist LCSFM Management Maria Emilia Freire Consultant WBIUR Daniel R. Gross Consultant AFTCS Gerard Martin La Forgia Lead Health Specialist EASHH Health Kathy A. Lindert Sector Leader LCSHD Human Development Rogerio F. Pinto Consultant AFTHE Alberto Rodriguez Sector Manager, Education ECSH2 Education Sustainable Maria Angelica Sotomayor Araujo Sector Leader LCSSD Development Bernice K. Van Bronkhorst Sector Manager SASDC Disaster Risk Luciano Wuerzius Procurement Specialist LCSPT Procurement Supervision/ICR Sinue Aliram Procurement Specialist LCSPT Procurement Financial Susana Amaral Financial Management Specialist LCSFM Management Judy L. Baker Lead Economist EASIN Etel Patricia Bereslawski Aberboj Senior Procurement Specialist LCSPT Procurement Fernando Andres Blanco Cossio Senior Economist AFTP4 Economist Cidalia Brocca Finance Analyst CTRLN Finance LCSUW- Marcus Vinicius Ferreira Da Silva Consultant Engineer HIS Nicolas Drossos Consultant EASNS Maria Emilia Freire Consultant WBIUR LEGLA- America Teresa Genta Fons Lead Counsel Legal HIS Ivo Imparato Senior Urban Development Specialist TWIWP Urban (TTL) Daniel R. Gross Consultant AFTCS Social Specialist Jose C. Janeiro Senior Finance Officer CTRLA Finance LEGLA- Jorge Kamine Counsel Legal HIS Marta Elena Molares-Halberg Lead Counsel LEGES Legal Catarina Isabel Portelo Senior Counsel LEGLE Legal Luis R. Prada Villalobos Senior Procurement Specialist MNAPC Procurement Sustainable Maria Angelica Sotomayor Araujo Sector Leader LCSSD Development Bernice K. Van Bronkhorst Sector Manager SASDC Disaster Risk (TTL) Sameh Naguib Wahba Sector Manager, Urban UDRUR Urban Anchor (TTL) Luciano Wuerzius Procurement Specialist LCSPT Procurement Taimur Samad TTL / Senior Urban Economist EASIS Urban (TTL) Tatiana Cristina O. de Abreu Finance Analyst CTRLN Finance Controller Souza Francesco Notarbartolo di Consultant UDRUR Social Specialist Villarosa 33 Infrastructure Finance Alessandra Campanaro Senior Infrastructure Finance Specialist LCSDU (TTL) Emanuela Monteiro Urban Specialist LCSDU Urban (Co-TTL) Alberto Costa Senior Social Development Specialist LCSSO Social Development Clarisse Dall Acqua Senior Environmental Specialist LCSEN Environment Danilo Pereira de Carvalho Procurement Specialist LCSPT Procurement Paulo Fantini Consultant LCSDU Infrastructure Klas Ringskog ICR Consultant LCSDU Infrastructure Water Supply and Luz MarÍa Gonzalez Economic Analysis Specialist LCSWS Sanitation Fabiana Meacham Consultant LCSDU Urban (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY04 24.21 140.99 FY05 40.31 236.23 FY06 8.28 53.96 Total: 72.8 431.18 Supervision/ICR FY06 12.21 72.40 FY07 14.56 99.31 FY08 13.92 89.84 FY09 17.83 139.83 FY10 21.53 172.23 FY11 6.86 79.37 FY12 12.23 104.99 FY13 16.31 86.38 FY14 7.6 46.28 Total: 123.05 890.63 34 Annex 5. Beneficiary Survey Results N/A 35 Annex 6. Stakeholder Workshop Report and Results N/A 36 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR *** Full content of the documents was uploaded in the Project files *** (a) Summary of the Borrower's ICR Context The Project was conceived as part of the efforts of the Government of Bahia to reduce poverty in the state, specifically among urban populations of higher social vulnerability. Implementation of the Project was complex, characterized by a series of issues outlined below, the majority of which arising from the preparation phase including (i) the Project implementation time in relation to the complexity of its design; (ii) ambitious main objective and scope; (iii) underestimation of costs and risks; and (iv) complexity in the proposed institutional and implementation arrangements. Project Results and Objectives The Borrower substantially agrees with the Bank evaluation of the results framework. For more details, refer to the full document, available in WB Docs. World Bank Performance during Project Preparation and Implementation The main negative aspects of World Bank performance during the project preparation phase were: i) Excess of planning instruments prepared ii) Absence of integrated urban upgrading and infrastructure detailed design projects iii) Ambitious PDO and scope iv) Absence of an implementation plan with realistic costs and timeframes v) Complex design relative to the proposed implementation period for the Project vi) Underestimation of the risks of the Project—in particular institutional coordination and cooperation, and the risk inherent in cooperation across the political spectrum vii) Complex institutional arrangement The main positive aspects of World Bank performance during the project preparation phase were: i) Integrated approach to slum upgrading ii) Implementation support provided by the Bank (team) during project preparation The main negative aspects of World Bank performance during the project implementation phase were: i) Limitations in the restructuring process of the project ii) Changes in the World Bank’s project leaders (TTLs) iii) Excessive bureaucracy in dealing with fiduciary matters (procurement) for the project The main positive aspects of World Bank performance during the project implementation phase were: 37 i) Technical assistance made available by the project team Government Performance during Project Preparation and Implementation The principle negative aspects of Government performance during the project preparation phase were: i) Underestimation of the Project costs ii) Lack of integration between UTP (Unidade Tecnica do Projeto – Project Executing Unit) and the rest of the organizational structure within CONDER (Companhia de Desenvolvimento Urbano do Estado da Bahia – State of Bahia Urban Development Company) The main negative aspects of Government performance during the project implementation phase were: i) Slow pace of implementation of the project during the first years (2006 to 2008) ii) Changes in project management (UGP / Unidade de Gestao do Projeto – Project Management Unit, within SEDUR; and UTP, within CONDER iii) Modifications made to the detailed design projects for the urban upgradings in the PIFs (Poligonais de Intervenção Física -Physical Intervention Areas) iv) Delays in the payment of invoices and bills for services rendered v) Difficulties in making payments between May and September, 2013 negatively affected the implementation pace of the works The main positive aspects of Government performance during the project implementation phase were: i) Government’s commitment in allocating the additional counterpart needed for the completion of the project’s targets ii) Project resources invested in the preparation of detailed designs made it possible for the Government to reach alternative sources of funding through Caixa Economica Federal / PAC II and BNDES / PROINVEST (2014-2016) Lessons Learned During the preparation and implementation of the Project, the main lessons learned were the following: Importance of a policy for slum upgrading that integrates the various dimensions of urban development – a state policy for slum upgrading should integrate environmental, social, economic, and cultural dimensions of urban development, with the aim to physically integrate those parts of the urban fabric with the formal city through (i) creation and improvement of accessibility; (ii) implementation of public facilities and spaces in the design of slum upgrading; (iii) expansion of basic sanitation, respecting the support capacity of natural ecosystems. Beyond these aspects, sectors such as education, health, violence reduction and prevention, job and income creation, and social inclusion should be promoted, as a way to improve quality of life for the resident population. 38 Relevance of an integrated approach to slum upgrading – it is advisable for slum upgrading projects to balance demand for physical interventions with activities that support social development. Slum upgrading projects should (i) integrate low-income settlements with the formal city through creation and improvement of accessibility; (ii) value the implementation of social public facilities and spaces; (iii) be aware of the support capacity of natural ecosystems in the design of slum upgrading projects; (iv) expand basic WSS services; (v) work within (and reveal) the characteristics of the existing social fabric; (vi) develop support activities for families affected by physical works interventions or relocated from risk areas, principally those related to job and income generation as well as group coexistence. It is worth highlighting that new procurement documents from CONDER, including terms of reference for carrying out slum upgrading works and social development projects, make this new concept explicit, mandating an integrated approach to slum upgrading. Importance of a social and environmental safeguards framework for the implementation of slum upgrading projects – the improvement of involuntary resettlement processes and procedures for families affected is essential for the implementation of slum upgrading projects. The involuntary resettlement policy of the Project is cautious with the social fabric and forms part of the principle that guarantees the reestablishment of the quality of life of families affected by physical interventions regarding, (i) the loss of dwelling; (ii) the loss of financial income (interruption of productive activities); (iii) losses stemming from the disruption of social support and relationships within a neighborhood. The legal framework put in place by CONDER for procedures concerning assisted purchase (indenizacao monitorada) amounted to a lesson learned and is being applied in new slum upgrading projects. Concerning environmental safeguards, the Project developed instruments for monitoring these aspects, which are available at CONDER for application in other urban development projects. Downsizing the institutional arrangements and integrating project management and execution functions in a single unit/instance – in the case of similar projects financed by the World Bank, it is advisable to simplify the institutional structure, with a focus on project implementation and involve a reduced number of interlocutors and agents. It is also advisable to implement a single coordinating unit to avoid instances of rework and overlapping roles, and to propel the project forward during its implementation. Necessity of alignment within diverse levels of public administration for the implementation of slum upgrading projects – the difficulties encountered during the implementation process related to (i) obtaining environmental licenses for the execution of physical interventions; (ii) obtaining construction permits for the works (in some intervention areas permits have still not been issued), (iii) resolving the legal status of public owned land in order to allow the issuance of titles to beneficiaries; and (iv) the signing of legal terms for the delivery of the completed works and social facilities to the due level of administration (the municipalities), these all suggest the necessity of alignment of diverse levels of public administration, by means of the formalization of processes and procedures for the implementation of slum upgrading projects and consequent sustainability of investments made. Alignment of methodology for addressing fiduciary aspects – the methodology for supervision of the financial management and procurement fiduciary aspects of a project should be adjusted to the institutional capacity of the borrower and approaches should be aligned. Depending on the capacity of the client, supervision arrangements should give more or less autonomy to the execution of administrative processes and procedures. That is, revise beforehand step by step the procedures for the contracting of services, or provide more wide ranging and complete support, 39 with emphasis on the systematic control of revisions and possible penalties, which reduce the management costs and streamlines implementation of the project. Necessity of adopting preventative measures to address exchange rate fluctuations – in the case of projects financed in foreign currency, financially subject to exchange rate fluctuations, it is advisable to evaluate risks beforehand and adopt mitigation measures, with the aim to minimize impacts, assure the completion of implementation, and reach goals and objectives. (b) Comments made by the Borrower on Draft ICR Most of the suggestions made by the Borrower were addressed in the final review of the ICR. Key aspects raised are the following: 1. On the Results Framework: 1.1. Intermediate Outcome Indicator A1 – “Number of households with improved housing conditions”: Different from what is mentioned in the ICR, the Borrower disagrees with the comment from the Bank team on intermediate indicator A1 qualifying the monitoring approach as ‘conservative’. The client specifically reiterates that regular cash compensations provided to beneficiary of resettlement policies should not be part counted as a household with improved housing conditions given the impossibility to monitor the use of funds transferred to the beneficiary. 1.2. Intermediate Outcome Indicator A2 – “Percentage of households in high risk areas or unhealthy conditions resettled”: The Borrower comments that a different methodology – which was a part of the Project’s Monitoring and Evaluation Plan and approved by the Bank in 2013 – was undertaken to measure this indicator. By so doing, the actual achievement / measurement would be 111%, as a result of the relation between the number of households resettled within the PIFs (407 45) and the number of households taken out of high risk areas or unhealthy conditions (366 46). 2. On Key Factors Affecting Implementation and Outcomes: 2.1. Crime and violence: The Borrower does not concur with the ICR’s assessment that crime and violence have negatively impacted Project implementation, due to drug dealing gangs in Project target areas. 3. Assessment of Risks to Development Outcomes and Project Costs and Financing: The Borrower raises the fact that an estimated amount of additional counterpart funding (USD 56.3 million) – of which USD 31.05 million were executed before Project closing – was fully budgeted for, among other, and included the completion of Project works after its closing date. 45 49 from Alagados III PIF and 358 from Avenida Anchieta PIF. 46 49 from Alagados III PIF and 317 from Avenida Anchieta PIF. 40 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders N/A 41 Annex 9. List of Supporting Documents Project Appraisal Document (Report No. 31856-BR) – October 31, 2005 Restructuring Paper 1 (Report No. 58588-BR) – January 4, 2011 Restructuring Paper 2 (Report No. 73500-BR) – November 12, 2012 Operational Manual – June 11, 2010 Environmental Impact Assessment – March 21, 2005 Resettlement Framework – March 21, 2005 Beneficiary Study: Kraychete Gabriel (2012), Fomento à Economia dos Setores Populares, Urban Development Secretariat and Urban Development Company, State of Bahia, 2012 Country Assistance Strategy 2003-2007 (Report No. 27043-BR) – December 9, 2003 Country Partnership Strategy 2012-2015 (Report No. 63731-BR) – September 21, 2011 Aide Memoire – October 2013 Aide Memoire – March 2013 Aide Memoire – July 2012 Aide Memoire – October 2011 Aide Memoire – May 2011 Aide Memoire – December 2010 Aide Memoire – February 2010 Aide Memoire – September 2009 Aide Memoire – May 2009 BTOR Midterm Review – December 2008 Aide Memoire Midterm Review – December 2008 Aide Memoire – August 2008 Aide Memoire – March 2008 Aide Memoire – November 2007 Aide Memoire – May 2007 Aide Memoire – May 2004 Aide Memoire – February 2004 Aide Memoire – November 2002 Implementation Status Report 15 – November 2013 Implementation Status Report 14 – June 2013 Implementation Status Report 13 – October 2012 Implementation Status Report 12 – January 2012 Implementation Status Report 11 – August 2011 Implementation Status Report 10 – February 2011 Implementation Status Report 9 – May 2010 Implementation Status Report 8 – December 2009 Implementation Status Report 7 – June 2009 Implementation Status Report 6 – December 2008 Implementation Status Report 5 – June 2008 Implementation Status Report 4 – December 2007 Implementation Status Report 3 – June 2007 Implementation Status Report 2 – December 2006 Implementation Status Report 1 – June 2006 42 World Bank (2005-2013), Additional Documents in Project’s Electronic File, including Credit Agreement, Back-to-Office Reports, Project Status Reports, Project Financial Assessments, and Project Procurement Assessments; Washington, D.C. Hildebrand, Elisabeth (2001). Avaliacao Economica dos Beneficios Gerados por los parques urbanos: Estudo de Caso en Curitiba-PR. Kraychete Gabriel (2012), Fomento à Economia dos Setores Populares, Urban Development Secretariat and Urban Development Company, State of Bahia, 2012 43 44