Document of The World Bank Report No: ICR1868 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-46900) ON A LOAN IN THE AMOUNT OF US$46.0 MILLION TO THE STATE OF PERNAMBUCO, BRAZIL WITH THE GUARANTEE OF THE FEDERAL REPUBLIC OF BRAZIL FOR A RECIFE URBAN UPGRADING PROJECT August 30, 2012 Sustainable Development Department Brazil Country Management Unit Latin America and the Caribbean Region CURRENCY EQUIVALENTS (Exchange Rates Effective April 24, 2003 and March 2, 2011) Currency Unit = Brazilian Real (R$) R$1.00 = US$0.33 – Board Approval US$ 1.00 = R$3.00 R$1.00 = US$0.60 – Closing Date US$1.00 = R$1.66 FISCAL YEAR January 1 – December 31 ABBREVIATIONS AND ACRONYMS BNDES National Development Bank (Banco Nacional de Desenvolvimento) CAS Country Assistance Strategy CBA Cost Benefit Analysis CDS City Development Strategy CELPE Energy Company of Pernambuco COFIEX External Finance Commission (Comissão de Financiamento Externos) COHAB State Housing Agency COMPESA State Water Company CONDERM RMR Development Council CPS Country Partnership Strategy DRUP Rapid Participatory Urban Diagnostic EA Environmental Assessment EMHAPE State Housing Agency EMP Environmental Management Plans FIDEM Foundation for Municipal Development (Fundação de Desenvolvimento Municipal) FM Financial Management GSP Government of the State of Pernambuco GTZ German Technical Cooperation Agency (Deutsche Gesellschaft fur Tecchnische Zusammenarbeit) ICMS Taxes on the Circulation of Goods and Services IFR Interim Financial Report IPTU Property Taxes M&E Monitoring and Evaluation NGO Non-Governmental Organization OM&R Operation, Maintenance and Repair PAC Growth Acceleration Program (Programa de Aceleração do Crescimento) PDO Project Development Objective PMU Project Management Unit PPA State Multi-Year Plan (Plano Plurianual) PRAFEP State of Pernambuco Fiscal Adjustment Program (Programa de Restruturação e de Ajuste Fiscal do Estado de Pernambuco) PREZEIS Plan to Regularize Areas of Special Social Interest (Plano de Regularizacao das Zonas Especiais de Interesse Social) PROMETROPOLE Infrastructure Program for Low-Income Areas (Programa de Infraestrutura em �reas de Baixa Renda da RMR) RMR Recife Metropolitan Region SEM State Economic Memorandum SEPLANDES Planning and Development Secretariat SIGMA Management Information System UE Sanitation Collection Unit (Unidade de Esgotamento Sanitário) URB Recife Urbanization Company WSS Water Supply and Sanitation WTP Willingness to Pay Vice President: Hasan A. Tuluy Country Director: Deborah L. Wetzel Sector Manager: Anna Wellenstein Project Team Leader: Marcos Thadeu Abicalil ICR Team Leader: Marcos Thadeu Abicalil Brazil Recife Urban Upgrading Project Data Sheet .......................................................................................................................... i INTRODUCTION ............................................................................................................ 1 Section 1. Project Context, Development Objectives and Design .................................... 1 1.1 Context at Appraisal ............................................................................................. 1 1.2 Original Project / Program Development Objective (PDO) and Key Indicators .. 2 1.3 Revised PDO and Key Indicators ......................................................................... 2 1.4 Beneficiaries ......................................................................................................... 2 1.5 Original Components ............................................................................................ 3 1.6 Revised Components ............................................................................................ 4 Section 2: Key Factors Affecting Implementation and Outcomes ................................... 5 2.1 Project Preparation, Design and Quality at Entry ................................................. 5 2.2 Implementation ..................................................................................................... 6 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization ...... 7 2.4 Safeguard and Fiduciary Compliance ................................................................... 8 2.5 Post-completion Operation / Next Phase ............................................................ 10 Section 3: Assessment of Outcomes ............................................................................... 11 3.1 Relevance of Objectives, Design and Implementation ....................................... 11 3.2 Achievement of Project/Program Development Objectives ............................... 12 3.3 Efficiency ............................................................................................................ 15 3.4 Justification of Overall Outcome Rating ............................................................ 16 3.5 Overarching Themes, Other Outcomes and Impacts .......................................... 16 Section 4: Assessment of Risk to Development Outcome.............................................. 17 Section 5: Assessment of Bank and Borrower Performance .......................................... 17 5.1 Bank Performance ............................................................................................... 17 5.2 Borrower Performance ........................................................................................ 18 Section 6: Lessons Learned ............................................................................................ 20 Annex 1: Project Costs and Financing ............................................................................ 22 Annex 2: Outputs by Component ................................................................................... 23 Annex 3: Economic and Financial Analysis ................................................................... 28 Annex 4: Bank Lending and Implementation Support/Supervision Processes .............. 37 Annex 5: Beneficiary Survey Results ............................................................................. 39 Annex 6: Summary of Borrower's ICR and/or Comments on Draft ICR ....................... 40 Annex 7: List of Supporting Documents ........................................................................ 43 Annex 8: Project’s Revised Components ....................................................................... 44 Annex 9: Project’s Results Framework .......................................................................... 45 Annex 10: Project Intervention Map............................................................................... 49 Annex 10: Map (IBRD 33377) ....................................................................................... 50 Table 1: Revised Project Components (Fourth Amendment, June 23, 2010) .................. 4 Table 2: Pending Resettlement Cases ............................................................................. 10 Data Sheet A. Basic Information RECIFE URBAN Country: Brazil Project Name: UPGRADING PROJECT Project ID: P049265 L/C/TF Number(s): IBRD-46900 ICR Date: 08/30/2012 ICR Type: Core ICR ESTADO DO Lending Instrument: SIL Borrower: PERNAMBUCO Original Total USD 46.00M Disbursed Amount: USD 46.00M Commitment: Revised Amount: USD 46.00M Environmental Category: B Implementing Agencies: COMPESA Municipality of Recife Muncipality of Olinda State of Pernambuco Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 06/09/2000 Effectiveness: 10/09/2003 10/09/2003 01/23/2004 10/13/2006 Appraisal: 10/15/2001 Restructuring(s): 03/06/2009 06/10/2010 Approval: 04/24/2003 Mid-term Review: Closing: 03/31/2009 03/02/2011 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: Moderate Bank Performance: Moderately Unsatisfactory Borrower Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Government: Moderately Satisfactory i Unsatisfactory Implementing Moderately Quality of Supervision: Moderately Satisfactory Agency/Agencies: Unsatisfactory Overall Bank Moderately Overall Borrower Moderately Satisfactory Performance: Unsatisfactory Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Project Quality at Entry No None at any time (Yes/No): (QEA): Problem Project at any time Quality of Supervision Yes None (Yes/No): (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Central government administration 5 Housing construction 30 40 Solid waste management 30 5 Sub-national government administration 5 15 Water supply 30 40 Theme Code (as % of total Bank financing) City-wide Infrastructure and Service Delivery 25 Environmental policies and institutions 20 Land administration and management 20 5 Participation and civic engagement 20 10 Urban services and housing for the poor 40 60 E. Bank Staff Positions At ICR At Approval Vice President: Hasan A. Tuluy David de Ferranti Country Director: Deborah L. Wetzel Vinod Thomas Sector Manager: Anna Wellenstein Danny M. Leipziger Project Team Leader: Marcos T. Abicalil Dean A. Cira ICR Team Leader: Marcos T. Abicalil ICR Primary Author: Elizabeth Eiseman ii Emanuela Monteiro Marcos T. Abicalil F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) Improve the wealth and well being of the urban poor of the Beberibe River Basin in the Recife Metropolitan Region by increasing the institutional capacity of public and civic entities to plan for, deliver and maintain basic shelter services for the low-income population in a coordinated and sustainable manner. 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 approval Completion or Target Values documents) Target Years Indicator 1 : Increase in property values by end of Project Increase of Increase of US$ US$2,963 per 7,349.92 per property property in slum Value in slum upgrading upgrading areas and quantitative or areas and increase of increase of Qualitative) US$ 24,277.64 per US$3,774 per property in property in resettlement areas. resettlement areas. Date achieved 03/31/2009 02/01/2012 Comments (incl. % Met achievement) Measure of the quality of housing pre and post-project (a measure of private Indicator 2 : investment) Data for UEs 20-21 and 17 Avg# of floors: 1.19 Avg# of floors: 1.18 (UE (UEs 20-21) and 20-21), 1.09 (UE 17); Walls 1.08 (UE 17); Walls not made of concrete: 9% not made of concrete: Value (UE 20-21), 4.5% (UE 17); 1% (UE 20-21), 0% quantitative or Floors not covered with (UE 17); Floors not Qualitative) concrete or tiles: 8%(UE covered with 20-21), 15% (UE 17); Ext. concrete or tiles: 6% walls w/o plaster: 20% (UE (UE20-21), 5% (UE 20-21), 30% (UE 17) 17); Ext. walls w/o plaster: 20% (UE 20- iii 21), 13% (UE 17) Date achieved 03/31/2009 02/01/2012 Comments MET (incl. % Results of the Borrower's Final Assessment Report show an overall improvement of achievement) housing quality post-Project. Indicator 3 : Measure of the perceptions of the quality of life of project beneficiaries Value Please See Annex 9 - The Please See Annex 9 - quantitative or data does not fit in this The data does not fit Qualitative) section in this section Date achieved 03/31/2009 02/01/2012 Comments MET (incl. % Results of the Borrower's Final Assessment Report show an overall improvement of achievement) perception of quality of life by Project beneficiaries. Indicator 4 : Measure of improvements in cost recovery of OM&R of investments 7,202 connections were installed in the geographic area served by Prometrópole. With Value an average payment quantitative or rate of 84 percent, the Qualitative) annual economic benefit from water distribution was estimated at US$1,735,452. Date achieved 02/01/2012 The marginal increase in the receipts of the water distribution company (COMPESA) Comments was used to estimate the economic benefits of the program. (incl. % *COMPESA was originally supposed to supply this information. The data from achievement) Datametrica was used as a substitute. Program decentralization structure (measure of decentralization of control to local Indicator 5 : levels and community) State: 47.40% (plus Spending by the end .60% Bank fees) Value of Project: 43% Olinda: 18.90% quantitative or executed by State, Recife: 33.10% Qualitative) 57% executed by Overall spending by municipalities municipalities: 52% Date achieved 03/31/2009 02/01/2012 Comments (incl. % Met achievement) iv (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 Indicator 1 : Number of new/legalized water supply and sewer facilities and connections Water supply systems: Total # contracted m 107,878.60, # executed m 103,764.30 (96%). Value 100% of works Household water (quantitative contracted have connections: Total # or Qualitative) been executed contracted m 8,528, # executed 7,434.00 (87%); Reservoirs: Total # contracted 1, executed 1 (100%). Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % Remaining works are being carried out. achievement) Indicator 2 : Number of kms of drainage constructed or rehabilitated Value 100% of works Contracted m (quantitative contracted have 11,292.1, executed m or Qualitative) been executed 10,335.1 (92%). Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % Remaining works are being carried out. achievement) Indicator 3 : Increase in length of thoroughfares with public lighting coverage Contracted 940m, executed 0 (0%). Value 100% of works Will be executed by (quantitative contracted have CELPE: 8,286 m or Qualitative) been executed contracted and executed (100%) Date achieved 03/31/2009 02/01/2012 Comments (incl. % Met achievement) Number of families resettled due to geotechnical risk and/or encroachment into Indicator 4 : environmentally sensitive areas Value 100% of works Contracted 1,674, (quantitative contracted have executed 1,385 or Qualitative) been executed (83%). The v remaining 27% are in execution. Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % Remaining housing development works are being carried out. There are 289 families achievement) still to be resettled. Number of community-based groups integrated into the slum upgrading process in the Indicator 5 : Beberibe River Basin 5 groups (Local Value 7 groups (local Development (quantitative development Councils) with 152 or Qualitative) councils) members. Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % The difference between target value and achieved value refers to the fact that the # of achievement) UEs was reduced in Project restructuring Indicator 6 : Number of kms of pedestrian paths and communal space improved and/or built Value 100% of works Contracted m (quantitative contracted have 37,813.50, executed or Qualitative) been executed 37,039.50 (98%). Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % Remaining works are being carried out. achievement) Indicator 7 : Number of kms of new and/or rehabilitated primary and secondary access roads Value 100% of works Contracted m (quantitative contracted have 27,875.10, executed or Qualitative) been executed 26,887.60 (96%) Date achieved 03/31/2009 02/01/2012 Comments Substantially MET (incl. % Remaining works are being carried out. achievement) Increase in access to recreational activities (pocket parks), including rehabilitation of Indicator 8 : historical infrastructure "Matadouros de Peixinhos" Value 100% of works Contracted 6, (quantitative contracted have executed 4 or Qualitative) been executed Date achieved 03/02/2011 02/01/2012 Comments Substantially MET (incl. % Remaining works are being carried out. achievement) Indicator 9 : Number of new property titles issued Value (quantitative 1,674 1,385 or Qualitative) Date achieved 03/31/2009 02/01/2012 Comments Substantially MET vi (incl. % Remaining titling is being carried out. achievement) Indicator 10 : Direct employment of community member is execution of Project works In all the works): Value 100% of works with Contracts at the (quantitative community period: 46; or Qualitative) members employed Accumulated contracts: 2,120. Date achieved 03/31/2009 02/01/2012 Comments (incl. % MET achievement) Program Cost structure, defined as the proportion of total project costs spent on Indicator 11 : overhead, surveying and planning, infrastructure investments, titling, social services, resettlement etc. Supra-local infrastructure: 12.5%; Supra-local Local infrastructure infrastructure: 12%; in poor areas: 41.5%; Local infrastructure Housing units: Value in poor areas: 46%; 24.6%; Social (quantitative Housing units: 21%; outreach: 4.3%; or Qualitative) Social outreach: 3%; Engineering designs: Engineering 4.6%; Project designs: 5%; Project management: management: 13%. 11.8%Bank fees: 0.6% Date achieved 03/31/2009 02/01/2012 Comments (incl. % MET achievement) G. Ratings of Project Performance in ISRs Date ISR Actual Disbursements No. DO IP Archived (USD millions) 1 06/04/2003 Satisfactory Satisfactory 0.00 2 12/04/2003 Satisfactory Satisfactory 0.46 3 06/04/2004 Satisfactory Satisfactory 0.86 4 12/02/2004 Satisfactory Satisfactory 1.01 5 05/11/2005 Satisfactory Moderately Satisfactory 1.01 6 01/04/2006 Moderately Satisfactory Moderately Unsatisfactory 2.07 7 05/19/2006 Moderately Satisfactory Unsatisfactory 2.38 8 12/15/2006 Moderately Satisfactory Moderately Unsatisfactory 3.45 9 03/30/2007 Moderately Satisfactory Moderately Satisfactory 3.45 10 06/04/2007 Moderately Satisfactory Moderately Satisfactory 5.05 11 12/18/2007 Moderately Satisfactory Moderately Satisfactory 9.89 vii 12 06/26/2008 Satisfactory Satisfactory 18.11 13 12/17/2008 Satisfactory Satisfactory 25.85 14 06/22/2009 Satisfactory Satisfactory 32.37 15 11/23/2009 Satisfactory Satisfactory 35.45 16 05/18/2010 Satisfactory Satisfactory 39.09 17 09/23/2010 Satisfactory Moderately Satisfactory 45.39 H. Restructuring (if any) ISR Ratings at Amount Board Restructuring Restructuring Disbursed at Reason for Restructuring & Key Approved PDO Date(s) Restructuring Changes Made Change DO IP in USD millions To create a second disbursement 01/23/2004 S S 0.86 category for goods and expand the Project description for Part E. To create an additional subcategory for works and 10/13/2006 MS U 2.94 reallocate loan proceeds among disbursement categories. To include engineering designs for waste water collection micro- basins, incorporate electronic bidding among the procurement 03/06/2009 S S 28.96 methods acceptable to the Bank, reallocate loan proceeds and disbursement percentages among categories, and extend the Closing Date for a period of 18 months. To extend the closing date by five months, to reallocate loan proceeds 06/10/2010 S S 39.09 and change the disbursement percentages and to revise the Project's scope. viii I. Disbursement Profile ix INTRODUCTION The Recife Urban Upgrading Project was one of the first integrated urban upgrading Projects in Brazil. It was appraised in October 2001, approved in April 2003, became effective in October 2003 and closed in March 2011. The Project was extended for a total of 23 months due to design-related delays, political impediments, unanticipated cost escalations, and a dramatic increase in the number of families to be resettled. The Project was a very complex operation; it was multi-jurisdictional, relied on four implementing agencies that had varying levels of capacity, and involved interventions in very challenging communities. The Project reached its development outcomes and produced many lessons learned that have played an important role in the development of subsequent Bank and Borrower projects. Section 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1. Country and Sector Background. At the time of appraisal, the living conditions in the low-income areas in the Recife Metropolitan Region (RMR) were among the worst of any major metropolitan region in Brazil. These areas were characterized by makeshift housing, irregular development patterns, a lack of basic urban services, a lack of physical and social infrastructure, and insecure land tenure. In addition, the low-income settlements were often located in environmentally sensitive and high risk areas. The Project was focused on the Beberibe River Basin, which had one of the highest concentrations of poverty in the RMR. Sixty six percent of the Basin’s 550,000 residents lived on less than two minimum salaries. 1 Residents were living in areas of severe blight, poor drainage and had poor or no sanitary services. Some of the main sector issues in the Beberibe River Basin 2 were: (i) intermittent provision of water and un- metered / illegal connections; (ii) 29 percent of households connected to the sewerage network; (iii) untreated effluents; (iii) insufficient drainage systems; (iv) rain water mixed with sewage; and (v) about 60 percent of solid waste collected door-to-door. 2. In order to address the challenges facing the low-income population, the Government of the State of Pernambuco (GSP) led a number of initiatives to manage interventions in low-income and irregular urban areas. The programs, however, were carried out on a small scale, did not approach interventions in an integrated manner and had a limited impact. As Recife and Olinda became increasingly urbanized and informal settlements became increasingly common, the municipalities felt a growing sense of urgency to develop a framework to systemize urban upgrading and land management. In 1999, the GSP approached the Bank with a proposal to carry out a large scale, urban upgrading project that would improve the wealth and well being of the people along the Beberibe River Basin and strengthen the municipal institutions’ ability to manage informal settlements. 3. Rationale for Bank Assistance. The Bank was well positioned to help the GSP develop and carry- out the Project. In addition to having worked with 14 cities in the RMR to create a City Development Strategy and having developed a sanitation plan for Recife under the Federal Brazil Water Quality and Pollution Control Project, the Bank had experience in large scale urban upgrading projects. 4. The Project also directly supported the Bank’s 2003 to 2007 Country Assistance Strategy’s (CAS) development objective to reduce urban poverty. The CAS highlighted the following critical 1 At the time of the Project’s preparation, two minimum salaries was an income measure of poverty, consistent with the World Bank's definition of the poverty line of less than one dollar per day per person. 2 The basin spans over the Recife, Olinda and Camaragibe municipalities, out of the 14 that comprise the RMR. 1 problems affecting the urban poor: (i) a lack of urban services, in particular sanitation; (ii) a high incidence of water-borne diseases; and (iii) a qualitative and quantitative housing deficit exacerbated by a lack of access to land through normal market mechanisms and the ensuing problems of insecure land tenure, particularly in squatter settlements. The CAS also highlighted the need for assistance in Northeastern Brazil. The Project addressed all of these issues. 1.2 Original Project / Program Development Objective (PDO) and Key Indicators 5. The Project Development Objective (PDO) aimed to improve the wealth and well being of the urban poor of the Beberibe River Basin in the Recife Metropolitan Region by increasing the institutional capacity of public (state and local) and civic entities to plan for, deliver and maintain basic shelter and urban services for the low-income population in a coordinated and sustainable manner. 6. The PDO’s key performance indicators were: (i) Measure of the changes in property values pre and post project; (ii) Measure of the quality of housing pre and post project (a measure of private investment); (iii) Measure of the perceptions of quality of life of Project beneficiaries; (iv) Measure of improvements in cost recovery of OM&R of investments; and (v) Program decentralization structure (assessing the division and devolution of responsibilities from state to municipalities). 3 1.3 Revised PDO and Key Indicators 7. The PDO was not revised during the Project. The key indicators were not changed, but the measurement standards for these indicators were restructured in June 2010 as follows: (i) “Measure of the changes in property values pre and post Project� was based on a Cities Alliance survey for the baseline and a similar methodological study for the end of Project measurement; and (ii) “Measure of improvements in cost recovery of OM&R of investments� measured only WSS tariffs as other variables were irrelevant for the low income areas. 8. To better capture the expected results, the following changes were made to the output indicators: (i) The indicator that measured the “Percent change in conduction capacity of drainage� was removed. The team felt that the result of the Project’s drainage interventions would be more effectively captured by the “Number of kilometers of drainage constructed or rehabilitated�; (ii) The indicator that measured the “Number of new retaining walls constructed� was also removed due to the fact that this activity was excluded from the Project; and (iii) The indicator that measured the “Adoption of a streamlined process of land tenure regularization in the informal settlements� was removed, as this was outside of the scope of the State’s responsibility. The Municipalities of Recife and Olinda have departments which are responsible for land tenure regularization. 1.4 Beneficiaries 9. The Project’s primary target group was the low-income population of the Beberibe River Basin. The investments were intended to directly target 35,000 such families (150,000 individuals) and to indirectly reach 500,000 people (the total population of the basin). 10. The expected benefits of the Project included, inter alia: (i) improved access to basic services; (ii) improved WSS services; (iii) the resettlement of people living in risk prone areas; (iv) the reduction of flooding as a result of the macro and micro drainage investments; (v) environmental and health 3 In the PAD, the Project indicators in the main text and the annex are not aligned. The ICR team decided to analyze the Project indicators as provided in Annex 1 – Project Design Summary. In the Annex, however, there is an editorial mistake: “availability of housing� should read “quality of housing.� 2 improvements; (vi) increased social capital; and (vii) improved institutional capacity of the Borrower’s agencies to address planning and shelter needs in low income areas. 1.5 Original Components 4 11. The Project components, as presented in the Loan Agreement, are as follows: Part A: Beberibe River Basin Investments Carrying out of infrastructure and environmental projects for the overall improvement of the built and natural environment of the Beberibe River Basin, consisting of: (a) creation or rehabilitation of parks, including the ecological park along the margins of the Beberibe River, pocket parks within each of the Sanitation Collection Units (UEs – Unidades de Esgotamento Sanitário) and the Matadouro de Peixinhos complex of historic buildings; (b) construction or rehabilitation of roads and accesses, including improvements in the general flow of traffic; (c) improvements in macro drainage to reduce flooding; and (d) improvements in the transport systems of sewage to be treated by the Peixinhos waste water treatment plant. Part B: Water Distribution 1. Installation of pipes and carrying out of the associated works to improve water supply and distribution within the UEs in the Beberibe River Basin. 2. Acquisition of micro meters for household installation in the UEs. 3. Installation of macro meters and the carrying out of associated works required to isolate the water distribution network within the UEs. Part C: Low-Income Investments Carrying out of urbanization projects in groups of low-income settlements located in the UEs 03 (R), 04 (R), 08 (R), 17 (R), 19 (R), 20 (R), 21 (R) and 23 (R), which are in Recife, and 07 (O), 12 (O), 13 (O), 15 (O) and 17 (O), which are in Olinda, consisting of water supply and sanitation improvements and household connections, secondary road improvements, micro drainage works, installation of public lighting, creation or rehabilitation of public parks and open spaces, resettlement of population from risk prone areas, and other related investments. Part D: Complementary Urbanization Activities Carrying out of activities complementary to the activities under the preceding Parts of the Project, consisting of: (a) engineering design services for the Subprojects; (b) development of social outreach and community participation activities; (c) provision of environmental education to the communities; (d) feasibility studies for a sites and services project; (e) studies for expanding to other areas in the RMR the investments made under the Project; (f) Project monitoring and evaluation activities; (g) a feasibility study for the establishment and operation of a micro credit program for housing improvement and small business development in the Beberibe River Basin (and provision of technical assistance for the operation of such micro credit program); and (h) formulation of the Environmental Management Plan (EMP). Part E: Project Management Provision of technical assistance to CONDEPE/FIDEM and the Municipalities in the management, implementation and coordination of the Project. 4 The Project components described in the PAD and in the Loan Agreement differ. The ICR team decided to utilize the components provided in the Loan Agreement – Schedule 2, Description of the Project. 3 1.6 Revised Components 12. The Project’s scope needed to be tightened due to a lack of funds during implementation. Although the GSP and the municipalities of Olinda and Recife increased their counterpart financing by US$36 million to ensure that the PDO was substantially realized, some of the planned activities had to be removed from the Project. The 55 percent appreciation of the Brazilian Real against the US Dollar 5, the 132 percent increase in construction sector costs 6 in Brazil, and the inconsistencies between cost estimations at Appraisal and actual prices during implementation were the main reasons for the Project restructurings. The Project’s components were revised as follows (See Annex 10 – Revised Project Components): Table 1: Revised Project Components (Fourth Amendment, June 23, 2010)7 Main changes Reasons for the changes The reference to: (i) the creation or rehabilitation of parks, including the ecological park along the margins of Although originally included in Part Part A: the Beberibe River; (ii) pocket parks A, these activities were included / Beberibe River within each of the UEs; and (iii) the carried out under Part C of the Investments Matadouro de Peixinhos complex of Project. historic buildings was removed from Part A. The available loan resources were restricted and interventions were prioritized in UEs 03, 20, 21, and 23 The interventions in UEs 04, 08, 17, (in Recife) and in UEs 07 and 17 (in Part C: Low Income and 19 (in Recife), and in UEs 12, 13, Olinda). In 2007, the Federal Investments and 15 (in Olinda) were removed Government launched the Growth from the Project. Acceleration Program (Programa de Aceleração do Crescimento – PAC), a massive housing and infrastructure program. With the support of the PAC, the GSP financed the works that were taken out of the Project’s scope. The reference to the micro credit Part D: The restricted available resources program for housing improvement Complementary were targeted at completing the and small business development in Urbanization engineering designs for the the Beberibe River Basin was Activities subprojects. removed from the Project. 5 At the time of Board Approval, US$1.00=R$3.00, and at the Closing Date US$1.00=R$1.66. 6 According to the National Research System on Construction Costs and Indexes (SINAPI), from the time of Appraisal (October 2001) to the Closing Date (March 2011) the construction costs in the Northeast region increased by 132%. 7 All restructurings were approved by the CMU. 4 Section 2: Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 13. Soundness of background analysis. The Bank team performed a solid background analysis of the intervention area. In addition to conducting a detailed analysis of the living conditions in the Beberibe River Basin, the team analyzed related projects in the region and incorporated lessons learned in the Project’s design. The knowledge gained from these past operations helped the Bank team develop three alternative structures for the Project that ranged from integrated interventions to sector specific investments. The Bank then selected the structure of the Project based on, among other criteria, the number of poor attended, the visibility and urban impact, the sustainability of the investments, the impact on habitability and risk reduction, and the ability to replicate the Project. The central weakness of the background analysis was the lack of a comprehensive assessment of the institutional capacities of the implementing agencies. 14. Assessment of the Project’s design. This kind of integrated urban project is by nature very complex as it is multi-sectoral, multi-jurisdictional, and includes interventions in some of the most challenging urban communities. The design was appropriate for the Project’s objective to “improve the wealth and well being of the urban poor of the Beberibe River Basin.� First, the design targeted the poorest neighborhoods in the RMR. Second, the design integrated urban housing and WSS to ensure that beneficiaries would have access to services, which are inseparable from physical well being. Nevertheless, the target related to this objective (improve the life of 150,000 low-income people living in the Basin) was overly ambitious given the Project’s scope and financial resources (US$84 million). 15. The design also included complex institutional arrangements (execution was shared between the GSP, Olinda, Recife and COMPESA). The arrangement was key to the Project’s objective to “increase the capacity of public (state and local) and civic entities to plan for, deliver and maintain basic shelter and urban services for the low income in a coordinated and sustainable manner� as it provided the municipalities and the state with a hands-on experience from which to build a precedent for future urban upgrading projects. The complexity of the arrangement, however, posed risks to the successful and efficient implementation of the Project as the implementing agencies depended on political support and had varying levels of financial and technical capacity. Despite these risks, the Project’s design did not emphasize institutional support and allocated less than five percent of its resources to project management and institutional development activities. In addition, the Project design did not include a clear method for evaluating the degree to which the involved entities’ capacities increased (See Section 2.3 – Monitoring and Evaluation (M&E) Design, Implementation and Utilization). 16. Financial aspects. From a financial stand point, the design did not support efficiency in the Project. First, as mentioned above, the Project costs were underestimated. Second, the financial resources allocated to support institutional strengthening were insufficient. Third, the manner in which the many financial categories were structured caused unnecessary rigidity, resulting in delays, complex implementation and multiple restructurings. 17. Safeguards. The safeguards triggered, Environmental Assessment (OP / BP 4.01) and Involuntary Resettlement (OP / BP 4.12), were considered appropriate given the issues posed by the Project. 18. Resettlement and titling. The Project design underestimated the complexity and scale of the resettlement and titling issues. The time required for the identification and the due legal process for land acquisition was not taken into account, resulting in a delay in the resettlement schedules. In Brazil land regularization is based on complex federal and municipal legislation. The Project did not identify the need 5 to provide technical assistance to the implementing agencies to navigate the process of issuing titles to existing houses. 19. Adequacy of Government’s commitment. At the time of preparation, the Project was a clear priority of the GSP. In 1999, the GSP and the Federal Government signed the State of Pernambuco Restructuring and Fiscal Adjustment Program (PRAFEP). The Project was one of four external credit operations included in the PRAFEP. The GSP, demonstrating its strong commitment to the Project, oversaw technical studies for institutional and financial arrangements, analyzed project alternatives, and assessed the degree of poverty in the proposed Project areas. The GSP also selected two areas (out of the total thirteen) for pilot projects early in preparation and met with the proposed beneficiary communities to develop preliminary improvement plans. The GSP was also fully cooperative with the Bank in appraising the Project and made unanticipated concessions. Among those was the GSP’s agreement with the Bank’s recommendation to contract private operators to handle water services in the Project area at a time when discussing possible privatization was still very sensitive politically. 20. Assessment of risks. The Project team identified the majority of the risks that affected the Project’s implementation but did not propose strong enough measures to mitigate the risks. The risk to the development outcome at the time of evaluation was rated as substantial. The most significant risk (rated as high) identified was the institutional capacity of the municipalities and FIDEM to execute the Project. To mitigate this risk, the Project team proposed hiring a consulting firm to assist in the coordination of the Project and to provide support to the PMU. The Borrower felt that this mitigation measure was effective in providing managerial support to the Project’s implementation, but should have included more structured technical assistance on complex issues, such as inter institutional coordination, urban legislation and land regularization. As discussed in the Assessment of the Project Design Section, the risk assessment did not foresee the extent to which the institutional arrangement would affect implementation. 21. The risk assessment also underestimated the risk of maintaining costs within 15 percent of estimates. This risk was rated as negligible to low but was a key aspect affecting the actual delays and efficiency. Another risk that was underestimated was the delay in counterpart funding. An unidentified risk was the technical risk associated with not being able to properly assess the land where works were planned to take place (people were oftentimes living on this land). This inability to perform proper assessments led to problems in the execution of the works, increasing the costs and delaying the progress of the Project. 2.2 Implementation 22. Although all of the loan proceeds were disbursed by the Project’s Closing Date, the Project faced significant obstacles and delays during implementation. The shared implementation arrangement among the GSP, the municipalities of Recife and Olinda, and COMPESA made execution difficult, as all of the entities were autonomous and had varying levels of capability. As a result, the decision making process among these entities was very challenging. 23. Another issue affecting implementation progress was the Project’s vulnerability to electoral cycles. In 2004, just after Project effectiveness, municipal elections were held, resulting in a political disconnect between the Mayors and the Governor. A full Municipal and State political alignment was only reached in 2008, when new Mayors from the same coalition as the Governor were elected. Thereafter, the willingness of the executing entities to work together propelled the Project forward. After 2008, the Governments’ support for the Project did not waiver. 24. Implementation also suffered from design related delays. At effectiveness (2003), the Project only had engineering designs for two pilot UEs. The first bidding processes for these works, launched only in 6 2005, failed because of underestimated costs. The development of the engineering designs for the additional UEs took two more years to complete (as late as 2007), and the bidding processes for the related works were not concluded until 2008. Project designs became outdated over the years of delays, necessitating revisions, which increased costs and further slowed the Project’s progress. In December 2006, three years after the Project’s effectiveness, only 10 percent of the loan had been disbursed, and Implementation Progress (IP) was rated “Moderately Unsatisfactory� (MU). The Borrower felt that the initial five year implementation period was not realistic given the Project’s level of ambition. 25. In addition, the rigidity of the Project’s cost allocation, as described earlier, delayed the Project progress, as it could not respond to price variations and changes in the implementation without formal amendments. 26. As mentioned before, the Project faced a significant increase in costs (See Annex 1: Project Costs and Financing). This was further exacerbated by a 37 percent increase in the number of families to be resettled (See the Safeguard and Fiduciary Compliance Section for more detail). In addition, COMPESA’s standards for WSS projects became more stringent, raising the overall costs of the Project. To ensure that the PDO would be substantially realized despite the increase in costs, the GSP, Olinda and Recife increased their counterpart funding by US$36 million. Securing complementary funding at the Federal level, however, further delayed the Project’s progress, as the submission process for funding requests was time consuming. Despite the delay, additional funding had a very positive impact on Project results. 27. To mitigate the effects of the delays and the increase in costs, the Bank team and the PMU restructured the Project four times. Key changes included: (i) extending the Project’s closing date by 23 months, allowing for the completion of many activities that otherwise would have been left unfinished; (ii) narrowing the scope of the Project, focusing the limited resources on successfully implementing outputs in priority UEs; and (iii) simplifying disbursement categories, easing Project implementation. These adjustments were essential to the Project’s successful implementation, but some issues were not addressed, namely: (i) more accurately defining the PDO; (ii) revising the number of beneficiaries; and (iii) further extending the closing date so that all Project activities could be completed. 28. Despite the implementation challenges, there were many factors that contributed positively to implementation. The GSP and the Municipal Governments’ commitment to carrying out the Project from 2008 onwards was crucial to successful implementation. Other factors that positively impacted implementation include: (i) the continuity of the PMU’s technical team, which remained largely the same throughout implementation; (ii) the hiring of a management firm to support implementation provided a means to avoid public bureaucracy and promote communication and exchange among the stakeholders; and (iii) community participation during preparation and implementation facilitated decision making and made it possible for the teams to enter and operate inside violent communities. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 29. M&E Design. The first set of outcome indicators – “Measure of the changes in property values pre and post project�; “Measure of the quality of housing pre and post-project (a measure of private investment)�; and “Measure of the perceptions of quality of life of project beneficiaries� – were designed to measure wealth and well being. These outcomes were directly linked to the Project’s investments and were good measures of the impact and effectiveness of the investments. 30. The second set of outcome indicators were designed to measure the institutional capacity gains. The first indicator – “Measure of improvements in cost recovery of OM&R of investments� – was aligned to the Project’s objectives, but it did not reflect the fact that in the targeted low income areas, a large 7 percentage of the population is de jure exempt from paying property taxes and solid waste fees, which inherently affects cost recovery. The second indicator – “Program decentralization structure� – provided a straightforward means to assess the extent to which responsibility was being shared by the different implementing agencies. The ICR team, however, feels that this indicator is an intermediate indicator rather than an outcome indicator as it does not reflect the municipalities’ institutional gains. 31. The intermediate output indicators were directly tied to the achievement of the outcome indicators. Nevertheless, they did not provide the PMU with clear short and long term goals. The only quantitative guideline for the majority of the output indicators was that 100 percent of the works contracted had to be executed by the end of the Project. It is important to note that at the time that this Project was designed the practice of using results frameworks was relatively new and was much less advanced than today. 32. M&E Implementation. M&E was a challenge for the majority of the Project’s implementation. The PMU only began providing the Bank with Project Progress Reports in 2009, after a new project team leader informed the Borrower that it was required in the Loan Agreement. Supervision missions also became more frequent. The Project’s M&E implementation shortcomings related to low institutional capacity, a challenging institutional arrangement structure and a lack of familiarity with M&E methodologies. Since 2009, however, data on social outreach and community participation activities, contracts and works supervision and monitoring, and intermediate and output indicators was provided regularly and in a satisfactory manner. This was the result of diligent Bank supervision, which will be reflected in the later sections of this ICR. 33. M&E Utilization. The PMU began utilizing M&E to guide their actions over the last two years of Project implementation. The proper use of M&E instruments, including progress reports, helped both the Borrower and the Bank proactively address implementation obstacles. The production of the Borrower’s Completion Report, which the PMU developed hand-in-hand with an external consulting firm, demonstrated the PMU’s newfound capacity for M&E. The Report provided the Bank with a reliable reference for the development of this ICR. Notably, the PMU is including M&E as a key instrument to be used in the planned next phase of the Project, the PROMETROPOLE II. 2.4 Safeguard and Fiduciary Compliance 34. Fiduciary Compliance. There were no major fiduciary issues during the Project’s implementation. Although disbursements were minimal for the first four years of implementation, the loan was fully disbursed by the Project’s Closing Date. Procurement ratings remained Satisfactory or Moderately Satisfactory throughout the Project’s implementation. The main procurement challenge was the large number of amendments that had to be made to contracts during implementation because of inadequate planning, inadequate engineering designs, delays in implementation, and cost escalations. Others issues faced by the Project with respect to procurement were: (i) delays, especially in the civil works’ procedures, due to long internal approval procedures for issuance of bidding documents and contract awards and (ii) the learning curve associated with the use of the Bank’s guidelines and procedures; the Borrower was well versed in the Brazilian National procurement system, but was not familiar with Bank guidelines. 35. For the most part, the PMU complied with the Financial Management (FM) terms of the Loan Agreement and the applicable laws and regulations. FM problems related to: (i) delays in the delivery of the 2008 audit report; (ii) delays in the implementation of auditors’ recommendations; (iii) shortcomings in the reporting of counterpart funded activities and monitoring of the implementation financed by counterpart funds; and (iv) instability in SOEs and IFRs periodicity and quality. FM improved after the implementation of several FM actions plans that the Bank team and the PMU agreed on. The Project 8 closed with a “Moderately Satisfactory� rating for procurement and a “Moderate� rating for FM. The last audit report (January 1, 2010 – July 2, 2011), however, identified some inconsistencies in the IFRs and in accounts conciliation, which are currently being corrected by the Borrower and followed by the Task Team. 36. Safeguard Compliance. Although overall safeguards compliance ratings remained “Satisfactory� throughout the Project’s implementation 8, cumulative delays and unconcluded housing development and resettlement activities over the last year presented the Bank team with concerns. The Project was implemented under a Resettlement Policy Framework (RPF) dated November 2001. Additional Resettlement Action Plans (RAPs) were prepared for each specific resettlement project. A comprehensive census was carried out in 2003 and was revised in 2006 and in 2008. The resettlement was carried out in stages, but the cutoff dates were not enforced. Despite the advice of the Bank, the implementing agencies, for political and social reasons, opted to accept families for resettlement even when they had arrived just before resettlement began (not distinguishing between “good faith� and “bad faith� squatters). While the additional number of families resettled by the Project hindered its efficiency, it promoted efficacy and social commitment. 37. Regardless of whether the extra resettlement was positive or negative, the implementing agencies carried it out competently. Following best practices: (i) after construction, homes were distributed by lottery among eligible families, and families were allowed to trade locations with each other to be close to friends and relatives; (ii) a wide consultation and participatory process was held during the planning and implementing resettlement stages; (iii) community-based groups were integrated into the process; and (iv) a team of social workers worked closely with the legal, technical, and engineering teams to present and address the concerns of the community. 38. The increase in the number of families to be resettled, and the fact that construction sometimes took place in the same space as the original housing, made it necessary to place some families in provisional subsidized rental housing. The State and the Municipalities are still subsidizing these families because of delays in the completion of the housing developments. This has negative economic and social impacts (Governments are spending more and families are still waiting for their permanent houses). 39. Olinda currently has a housing deficit of 245 units. Despite the delays, the ongoing works and related resettlement activities are finally on track and are expected to be completed by December 2012. Recife, however, due to recent contract issues, has not started constructing its 44 pending houses. The Bank will continue to provide supervision until the process is satisfactorily completed (See Table 2 – Pending Resettlement Cases for details). 40. During the life of the Project there were no environmental safeguards issues. Rather, positive results in terms of urban environment improvement (not captured through the Project indicators) can be attributed to the Project interventions. For example, before the Project, the Jacarezinho Canal, which flows into the Beberibe River, was surrounded (and even covered) by informal houses that deposited sewage and solid waste directly into it. The area lacked adequate drainage and floods were frequent, hazardous and toxic. To address this issue, the Project resettled the families that lived along the canal, connected them to the collection networks, re-designed the canal’s drainage system and built urban infrastructure around the canal to prevent the reoccurrence of informal settlements. The Jacarezinho Canal was not an isolated case. The Project’s targeted areas no longer have raw sewage running through the streets, as there is a formal collection system in place. The paving of roads also improved the urban environment by reducing dust in the air. In addition, the Project increased the green spaces and urban amenities (pocket parks and sports courts). 8 For additional information on the last ISR ratings, please refer to Section 3 – Assessment of Outcomes. 9 Table 2: Pending Resettlement Cases Recife – Pending Resettlement Cases Olinda – Pending Resettlement Cases 44 245 Since April 2012 no progress has been made on the construction of the pending housing units. The pending houses were contracted through These houses are part of the integrated Federal Programs. 184 of these houses are urbanization works contract for UEs 20 and 21. currently under construction. The timeline for The contracted firm is facing legal and financial completion is as follows: 144 are to be completed difficulties, and the Municipality of Recife has not by September 2012 and 101 are to be completed yet taken a final decision on whether to insist that by December 2012. the current firm completes the works or to terminate the contract and hire a new firm. 2.5 Post-completion Operation / Next Phase 41. Several activities under the Project have not been fully completed to date: • UE 20-21 (Recife): 84.70% complete. Although the Borrower expects this work to be concluded in November 2012, the Bank team is skeptical given the above mentioned contractual problems (See Table 2 – Pending Resettlement Cases). The remaining works include: paving Joao Roma Street and Regeneração Avenues; Jacarezinho Canal; construction of 40 housing units in Jacarezinho; and establishing individual sewage connections; • UE 3 (Recife): 83% complete; expected to be concluded in September 2012. The remaining works include: sewerage network along Beira Rio and Construção Aves; paving of Beira Rio Avenue; and wastewater pumping facility; • Olinda: construction of 245 housing units; large majority expected to be concluded in December 2012; • Presidente Kennedy Avenue, in Olinda (GSP): 93% complete. The remaining works include adding road signs, which can only start after the conclusion of various repairs in the street pavement, urban median strip, storm drains and drainage conduits. These repairs (which are to be financed out of different, complementary contracts, outside of the Project) were started, but stalled due to sanitation works held by COMPESA; • UE 17 Sewerage Treatment Facility, ETE Minerva (GSP): Although construction is complete, the facility is not yet operational. Before COMPESA takes over its operation, final tests and repairs need to be concluded; • UE 17 Water Supply Reservoir (GSP): Defects were identified and need to be repaired; expected to be concluded in the near future. 42. The Task Team will continue supervision and implementation support of these activities given the sensitive nature of the ongoing and/or pending housing development and resettlement activities under OP 4.12. To this end, the Team has carried out seven post-Closing Date supervision missions (February 2011; May 2011; June 2011; November 2011; March 2012; May 2012; and June 2012). During the course of these missions, various action plans were agreed on. The targets established in these plans, however, were not fully met and required reprogramming and adjustment. Nevertheless, the GSP has demonstrated its commitment to complete the remaining works: (i) progress reports continue to be turned in; (ii) the PMU was kept in place (the GSP passed a decree that places the unit within Secretariat of Cities, outlines the responsibilities of the team, and describes the next steps required for preparation of the proposed PROMETROPOLE II); (iii) post-occupation and social work activities are also being continued; and (iv) budget allocations are secured under the State Budget and/or Federal Government Programs. The GSP 10 also secured complementary investments for the original UEs that were taken out of the Project through the Federal PAC I and II Programs. 43. As part of the Borrower’s Final Project Assessment, the Borrower produced a printed summary of the assessment and is planning on producing a video targeted at the marketing of the Project’s outcomes and outputs. Other than the marketing itself, these deliverables will play a key role in disseminating the results and lessons learned from the Project. 44. The GSP is supporting an initiative to prepare a second phase of the Project entitled PROMETROPOLE II. PROMETROPOLE II has secured the approval of its “Carta Consulta� by the Federal Guarantor, is in the GSP’s 2012-2015 Multiyear Plan, and is included in the GSP’s 2012 Budget. It draws on the knowledge and lessons gained during the first phase of the Project and focuses on: (i) complementary investments in the Beberibe River Basin; (ii) expanding its geographical reach to the Capibaribe River Basin (9 UEs in Sao Lourenço da Mata and Camaragibe municipalities); (iii) providing local social and economic development opportunities; and (iv) enhancing metropolitan governance and building municipal institutional capacities (land use regulation and disaster risk management). The GSP intends to simplify the implementation arrangements of PROMETROPOLE II and to prepare a series of agreements in advance that will accelerate the Project’s implementation. Also, to expedite the expansion of the wastewater services in the RMR, the GSP is preparing a PPP concession contract. Section 3: Assessment of Outcomes 45. The last Project ISR, submitted in October 2010, rated the Project as follows: PDO (S); IP (MS); Overall Safeguards Compliance (S); Environmental Assessment (S); Involuntary Resettlement (S); Financial Management (MS); Project Management (MS); Counterpart Funding (S); Procurement (MS); and Monitoring and Evaluation (MS). The Task Team performed the last supervision mission at the end of February 2011, which coincided with a period of numerous back-to-back missions. By the time the ISR was completed (late in March) it could not be submitted in the system because the closing date had passed. In addition, the social and environmental safeguards specialist was not able to join the last official mission due to a conflicting agenda and visited the Project in May 2011 (2 months after the closing date). After these missions, the Task Team felt that the following rating adjustments should have been made: PDO (MS); IP (MU); Overall Safeguards Compliance (MS); and Involuntary Resettlement (MU). 3.1 Relevance of Objectives, Design and Implementation 46. The Project objectives and design are highly relevant to current national priorities and the Bank’s partnership strategy. Despite the recent optimistic macroeconomic growth figures, Brazil faces challenges related to inadequate housing and service provision. In 2000, over 12 million people (3.2 million households) lived in precarious housing within slums (favelas) in Brazil. The country’s housing deficit is estimated to be 7 million units today and new demand is forecast to reach 23 million units by 2023. One third of the housing deficit is concentrated in Brazil’s 11 largest metropolitan areas. Addressing these challenges is a top government priority. In 2010 the Federal Government launched the second phase of the PAC, which envisions investments in housing, energy, social and urban infrastructure, community development, transportation and universalizing access to water and electricity. Between 2011 and 2014, Minha Casa Minha Vida, a subset of PAC II focused on housing issues, expects to invest US$155 billion (US$17 billion in slum upgrading, US$40.2 billion in housing subsidies, and US$97.8 billion in the housing financing market). Metropolitan governance and coordination between several municipalities and the GSP will be a key issue in these urban upgrading projects. 47. One of the four strategic objectives of the Bank’s 2012 to 2015 Country Partnership Strategy (CPS) with Brazil is to improve the quality and expand the provision of public services for low income 11 households. The CPS also highlights the need to maintain focus on poverty eradication in Northeastern Brazil and to keep the Bank’s engagement at the sub-national level, supporting, among others, strategic social and infrastructure investment programs. The Project’s objectives and design remain strongly aligned with the country’s priorities and the 2012 to 2015 CPS. The objectives are also closely linked to the Pernambuco State Multiyear Plan for 2012 to 2015 (Plano Plurianual – PPA 2012-2015), whose strategic themes include “Quality of Life – A better life for all,� and objectives: “to universalize the access to WSS,� and “to improve livability and mobility.� The PPA also focuses its actions on “the most vulnerable population.� 48. Given the above, the ICR rates the relevance of the Project as “high.� 3.2 Achievement of Project/Program Development Objectives 49. The outcome that most reflects the overall impact the Project had on improving the wealth and well being of residents is the increase in property values of the upgraded and new housing as it encapsulates the impacts of all the Project’s urban upgrading investments and measures the sum impact in market terms (See Annex 3 – Economic and Financial Analysis for more details). The increase in property values of the upgraded and new housing far exceeded the Project’s targets (See Annex 9 – Project’s Results Framework). The average value of the housing for the direct beneficiaries was estimated at US$16,928 pre-Project. The Project target outlined in the PAD was to increase the value by US$2,963. Results show that post-Project houses are evaluated at US$24,278, representing a 43 percent rise in value (increase of US$7,350). External factors, including the steep rise in real estate prices, the growth of employment and income, among others, undoubtedly contributed to the rise in prices, but the spillover effects of these factors would not have been felt as strongly by the beneficiaries had urbanization and new and upgraded housing works not taken place. 50. This increase in property value was substantiated by an overall improvement in the quality of housing and quality of life. The consulting firm hired by the Borrower (Datametrica) conducted a comprehensive survey in the six UEs, interviewing 3,835 people and assessing urban upgrading works. The survey was conducted in December 2011 and February 2012. The reach of the survey went beyond the outcome indicators in the PAD, and the following analysis draws on this comprehensive database. The quality of housing indicator was measured by analyzing the houses’ physical gains, including improvements in the walls, in the floors, increases in the number of rooms and pavements per house. Highlights included: • In UE 20-21 (Recife), the percent of households with 3 to 6 rooms increased from 47 percent in 2003 to 72 percent in 2011; • In UE 23 (Recife), the percent of households with concrete walls went from 60 percent in 2003 to 94 percent in 2011. Wood walls, which are a common mark of slum dwellings in Brazil, went from 30 to 4 percent; • In UE 17 (Olinda), external walls covered with plaster went from 70 percent in 2003 to 87 percent in 2011; • In UE 17, a predominance of ceramic floors in houses went from 30 percent of households in 2003 to 56 percent of households in 2011. 51. The quality of life outcome indicator also showed remarkably positive results. In the survey, residents rated, among other factors that contribute to quality of life, the provision of public services, water quality, sanitation, drainage, access to paved roads, public transportation, solid waste collection, and security (both land property security and safety perception). Results included: 12 • Water. In UE 17 (Olinda), regular/daily water service increased from 64 percent in 2003 to 92 percent in 2011, while in UE 20-21 (Recife) service increased from 35 percent in 2003 to 94 percent in 2011. In both UEs (17 and 20-21) the percentage of families without water provision was reduced to 0 (from 8 percent in UE 17 and 2 percent in UE 20-21). The general quality of the water utilities’ service also improved. The percentage of families that reported good service in UE 20-21 went from 39 percent in 2003 to 76 percent in 2011; • Sanitation (UE 20-21). The percentage of families evaluating sanitation services positively went from 17 percent in 2003 to 52 percent in 2011. The percentage of families without access to sanitation services went from 45 percent in 2003 to 4 percent in 2011. These results are reflective of the fact that approximately 40 percent of the overall investments made by the Project for expanding the network and 47 percent of the overall investments in new connections were made in UE 20-21; • Drainage and Pavement (UE 20-21). The prevalence of paved roads increased dramatically – in 2003, 43 percent said there was no paving, while in 2011 only 3 percent did. Similarly, 45 percent said drainage was not available in 2003, while only 2 percent did in 2011. Fifteen percent rated drainage services ‘good’ in 2003, while 53 percent did in 2011. These results in drainage highlight the effectiveness of the Project’s investments; UE20-21 received 55 percent of the overall investments of the Project in micro-macro drainage and 83 percent of overall investments in canals. 52. In addition to conducting surveys on the quality of housing and quality of life, Datametrica conducted focus group interviews to better understand the impact of the Project on a personal level. In general, the focus groups participants felt that the Project had raised their quality of life, improving, among other areas, living conditions, dignity, health conditions and accessibility. The focus groups also identified issues related to the Project, such as difficult adaptation in the newly resettled areas, size and design of the homes, a shortage of wastewater treatment (the wastewater treatment plant that the Project financed has been constructed, but is not operational) and insufficient social and education activities post- Project completion. 53. As discussed in the Assessment of the Project Design and in the M&E Design sections, the second objective of the Project – “to increase the institutional capacity of public (state and local) and civic entities to plan for, deliver and maintain basic shelter and urban services for the low-income in a coordinated and sustainable manner� – was not directly linked to the Project’s investments. Rather, institutional capacity was built via hands on experience. By isolating results shown from the Program decentralization structure outcome indicator, the GSP executed / spent 47.40 percent of the Project against a 43 percent target, and the municipalities executed 52.6 percent against a 57 percent target. Given the limitations of such indicator and considering there were no other outcome indicators that measured the impact of this experience, the ICR Team relied on the Borrower’s final report and personal anecdotes supplied by the institutions to assess this outcome. 54. In general, the Municipal and State entities felt that through the Project they enhanced their ability to deliver and maintain basic shelter and urban services for the low-income, particularly in the areas of resettlement. In regard to resettlement, the GSP and the municipalities developed a standardized approach (that includes community participation and social outreach) that they are currently utilizing in the execution of PAC-financed urban upgrading projects. Olinda described the approach as a “paradigm shift.� The Municipality now recognizes: (i) the need to develop resettlement plans; (ii) the need to pre- define the population that will be affected; (iii) the importance of systematically following up on the situation of the families; (iv) the benefit of having complementary social teams as part of the works contracts; and (v) the need to improve the transparency of the process. 13 55. The achievement of the PDO was also analyzed through assessing supporting information on: (i) Project coverage/number of beneficiaries; and (ii) physical outputs of the Project. 56. Beneficiaries. The estimated number of direct beneficiaries of the Project, as assessed by the Borrower’s Final Report,9 is 22,180 (6,440 households / families). If the indirect beneficiaries are taken into account, the Project benefited approximately 321,511people (89,597 households / families). Although these numbers differ considerably from the Project’s original targets (to directly benefit 35,000 families / 150,000 people), the ICR team considers that they are consistent with the final restructured scope of the Project given that: • The restructured Project only targeted six of the thirteen UEs that were originally included in the Project. As already mentioned, the number of beneficiaries was not restructured to align it with the scope tightening; • The Borrower’s methodology to define Project beneficiaries was overly conservative. Within each targeted UE only the people that benefited from the integrated urbanization works (Part C of the Project) were counted as direct beneficiaries. All of the people who benefited from Part B of the Project were not included in the direct beneficiary count 10 (See Annex 10 – Project Intervention Map); and • Finally, the Team highlights that the majority of the integrated urbanization activities in the UEs that were cut from the Project in the last restructuring were funded independently by Federal Programs. This population was only part of the indirect beneficiaries count. 57. Physical Outputs. The Project’s outputs provide a good sense of the progress the Project made towards achieving the PDO “to improve the wealth and well being of the urban poor in the Beberibe River Basin� (See Annex 2: Outputs by Component and Annex 9: Project’s Results Framework for more detail). One of the most important outputs of the Project was the 1,385 housing units constructed (plus land titles issued) to resettle families who were living in areas of severe blight, poor drainage and poor or no sanitary services. Other outputs that contributed to the achievement of the development outcome included: (i) the expansion of the WSS network and new household connections; (ii) the construction of a new water reservoir; (iii) macro and micro drainage and canal upgrading interventions; (iv) the implementation of public and urban facilities, such as new and rehabilitated access roads, new and improved pedestrian paths and communal spaces, a bus terminal, Matadouro de Peixinhos Cultural Center, pocket parks within the subproject’s areas, and street lighting; and (v) the employment of 2,120 community members. During supervision missions, contractors told the Team that 10 to 15 percent of the workers who were hired from the local community acquired marketable construction skills and found careers as construction workers. 58. The rehabilitation of Presidente Kennedy Avenue, which was part of the Project’s Part A, was an exception to the positive outputs discussed above. This intervention, aimed at improving the traffic flow of an important transport corridor in Olinda, was not concluded and did not achieve its intended results. The beneficiary survey conducted by the Borrower shows overall unsatisfactory results from the pedestrian and commercial stakeholders. The pedestrians feel that road security, public transport flow and traffic organization have not improved. When asked about possible improvements in the business 9 Avaliação do Programa de Infraestrutura em �reas de Baixa Renda da Região Metropolitana do Recife – Prometropole, May 2012 (Datamétrica Consultoria, Pesquisa e Telemarketing). 10 At preparation, the Project estimated that Part B (Water Distribution / Sectorization) would benefit approximately 95,000 people, in the historical center of Olinda and in ELO Alto do Ceu areas. Although a post-Project count was not conducted, all works included under Part B were finished successfully. The number of beneficiaries is expected be around 95,000. 14 environment, most of the commercial stakeholders refer to the intervention as having negatively affected their businesses. 59. The ICR team believes that the efficacy of the Project was “substantial.� The Project met all the outcome indicators (housing prices, quality of housing, quality of life) that were tied to the Project’s objective of improving the wealth and well being of the residents of the Beberibe River Basin. In addition, the Project met one of the two outcome indicators (decentralization structure) that was tied to the institutional capacity building objective (as discussed earlier, data on the second OM&R indicator was never collected). The Project also received qualitative feedback from both the beneficiaries and implementing agencies that reinforced the positive results of the outcome indicators. The ICR team rated efficacy as “substantial� as opposed to “high� because of the unconcluded infrastructure works and the related pending resettlement. 3.3 Efficiency 60. The Cost Benefit Analysis (CBA) was carried out by the consulting firm Datamétrica. Datamétrica estimated the economic benefits of the project indirectly, via the increase in property values, based on the hedonic price function. With this methodology, consumers’ willingness to pay for goods and services not available on the market could be estimated indirectly, using the increase in real property values in areas benefited by Prometrópole as a proxy. For the purpose of evaluating Prometrópole, data on real estate prices was obtained from the residents of properties in the areas where the Project operated. Then, the impact of various characteristics of the properties (size, number of bathrooms, type of walls, etc.) were analyzed to understand how they contributed to the value of the property. Datamétrica estimated how much Prometrópole contributed to the increase in property prices in areas where it was directly involved through estimating the difference in value for properties with and without those characteristics. Characteristics that positively impacted price included brick walls, plastered interior walls, ceramic tile floors, and slab roofs. 61. The economic benefits of water service, on the other hand, were estimated based on a census survey conducted among communities in the areas benefited by Prometrópole, using the following variables: average water bill, number of residences benefited, and percentage of residences with water meters. 62. The results of the CBA were estimated based on a comparison between the total economic benefit of the Project and the total economic costs of capital and maintenance and operation of the physical investments. The main results are as follows: (i) the value of the housing in the targeted areas exceeded the target projected in the PAD by 9.3 percent (US$157,345,158 versus US$143,967,703); (ii) water sectorization and distribution was estimated to generate an economic benefit of US$1,735,451.67 per year; (iii) the internal rate of return was estimated to be 40 percent, 3 percent higher than the target projected in the PAD (cash flow based on total benefits and total costs over a 20 year period); and (iv) the net present value of the Project, calculated using two different scenarios for the opportunity cost of capital in Brazil, demonstrated that the Project was economically viable. In addition, the CBA could not fully capture the positive, qualitative responses of the Project’s beneficiaries. 63. The ICR Team rates the Project’s efficiency as “substantial� given the positive results of the CBA. Efficiency was not rated as “high� because of design and implementation delays that negatively affected the efficiency of the Project. For instance, the GSP and the municipalities are still providing subsidies for to-be-resettled families living in rental housing (See Annex 3 – Economic and Financial Analysis for more details on the CBA). 15 3.4 Justification of Overall Outcome Rating Rating: Moderately Satisfactory The ICR rates the overall outcome of the Project as “Moderately Satisfactory� based on a “high� rating for relevance and a “substantial� rating for efficacy and efficiency. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 64. Although small business development activities were excluded from the Project in one of its amendments due to scarce financial resources, the Project’s engagement towards promoting a wide dialogue with civil society, the community and other development organizations provided unanticipated benefits. The Bank supported and facilitated the GSP to access funds from the European Union and other sources to implement an urban poverty and social vulnerability reduction initiative in Olinda. The €936,660 “Projeto Comunidade Viva – Projeto de Redução da Pobreza no Município de Olinda,� involved community strengthening, training and income generation activities and had a particular focus on gender issues. It was implemented from 2009 to 2011 through a partnership between the GSP, Olinda, the AVSI Foundation and the “Cooperação para o Desenvolvimento e a Moradia.� Through this Project 150 adults and 90 young apprentices were trained; community centers were refurbished and/or upgraded; and 6 projects proposed and implemented by the communities were supported. The training and capacity building courses (especially on the construction sector) targeted women’s groups. 11 Many of the women who participated in these courses began careers in construction, a sector that had been relatively closed to women. In addition, the Project, following common practice in Brazil, empowered women by giving them land titles for the new houses whenever possible. 65. The Project also facilitated the involvement of: (i) SEBRAE (“Serviço Brasileiro de Apoio a Micro e Pequenas Empresas�), which was key to providing business training opportunities to small business holders; (ii) SESI (“Serviço Social da Indústria�) and SENAI (“Serviço Nacional de Aprendizagem Industrial�), which held citizenship enhancement activities in the Project’s targeted areas; and (iii) the University of Pernambuco (Universidade Federal de Pernambuco), which developed environmental education materials for new home owners. (b) Institutional Change/Strengthening 66. As described in paragraphs 33, 53, and 54, the Project played a significant role in building institutional capacity to carry out resettlement in a systematic manner, to coordinate large scale urban upgrading projects and to monitor and evaluate projects’ progress. (c) Other Unintended Outcomes and Impacts 67. An unexpected impact of the Project has been an increase in the level of public investment in the Beberibe River Basin. As one of the first large scale integrated infrastructure projects in the RMR, the Project called both public and private attention to the area. For example, large investments financed by the Federal PAC Program were made in the River Basin following the same approach as the Prometropole Project, some of them based on engineering designs financed by the Bank, while others are being made to complement the quality of the urban environment, such as the ongoing dragging of the Beberibe River. 11 Gender focus under capacity building and job training activities is being mainstreamed in the Bank’s urban portfolio in Brazil. 16 The inter-institutional and coordinated implementation arrangement for the Project was reinforced and is currently used to discuss additional needs and investment priorities in the area. Section 4: Assessment of Risk to Development Outcome Rating: Moderate 68. Considering the general nature of the Project’s development outcome, the ICR Team decided to analyze the risk associated with sustaining the actual improvements the Project made towards the wealth and well being of the urban poor. The risk to the sustainability of the benefits and outputs are as follows: 69. First, the risk of not concluding the remaining activities is low. The PMU is actively monitoring Project progress, resources have been allocated for all pending activities, works have been contracted and are under execution and local politicians have voiced their support for the works. The upcoming 2012 municipal elections and the recent contract issues discussed earlier (See Table 2: Pending Resettlement Cases), however, may affect the flow of funds and the pace of implementation. The Task Team is helping manage this risk through continuing supervision of the Project. 70. Second, the risk to the sustainability of the infrastructure investments is moderate. The investments made in macro and micro drainage, roads and access improvements, for instance, will require OM&R from the Government during their useful lives. The sustainability of the WSS works will depend on the service provider’s ability to continue providing service to the new clients and maintain the new systems created through this Project. A PPP is being prepared for COMPESA which will help mitigate the risk. 71. Third, the inherent social risks of the Project are significant. The sustainability of the infrastructure works in the short term depends largely on the community’s involvement in maintaining and protecting them. For example, a small plaza that was financed by the Project was destroyed shortly after its completion because a gang wanted to use the space to play soccer. Also garbage is frequently thrown in the canals. In addition, the families’ precarious economic situations may hinder their ability to maintain the new homes, do repairs, and pay for WSS services. The municipalities also do not have a good track record of maintaining public spaces. A limited number of social activities to provide environmental education and guidance to reinforce community ownership and local economic development were included under the Project and, in hindsight, may not have been sufficient to adequately mitigate this risk. New Federal Programs, such as “Brasil sem Miseria,� and the ongoing technical assistance provided by the World Bank to the Ministry of Cities to strengthen the guidelines of social work within the context of Federally-supported urban development and slum upgrading projects, however, have the potential to significantly reduce these social risks. 72. Finally, the risk of re-favelization of the areas from which and to which the beneficiaries were resettled is moderate. The works in the canals, the construction of roads and sidewalks along the river banks, and the desire of the beneficiaries to maintain their new living conditions mitigate this risk. Nevertheless, the municipalities lack control over land use within the new housing areas and there is an intense demand for land in the RMR. In many of the housing areas home extensions were built over the common / public areas. The GSP is aware of these issues and has included mitigation measures in the next envisaged phase of the Project, PROMETROPOLE II. Section 5: Assessment of Bank and Borrower Performance 5.1 Bank Performance 17 Bank Performance in Ensuring Quality at Entry Rating: Moderately Unsatisfactory 73. The Bank team performed a comprehensive background analysis and integrated lessons learned in the country and sector in the design of the Project. The approach the Bank developed for this Project was innovative and is still relevant. There were, however, a number of serious shortcomings in the design of the Project. First, the Project’s cost assessment was unrealistic – costs were underestimated. Second, there was a disconnect between the actual investments and the institutional capacity building PDO – the original design devoted less than 5 percent of resources for institutional strengthening. Third, the results frameworks’ limitations posed challenges to M&E implementation and utilization. Finally, Project design underestimated the complexity and scale of resettlement and titling issues. Bank Performance in Quality of Supervision Rating: Moderately Satisfactory 74. The Bank Team demonstrated great flexibility and determination throughout supervision, successfully guiding the Project through years of delays, financial uncertainty, and resettlement booms without losing sight of the development objectives. For example, although the Project closed in March 2011, the Bank Team has continued closely supervising the remaining resettlement cases and unfinished works, which has been key to ensuring adequate Government commitment to complete implementation. In addition, the Bank team restructured the Project four times to respond to Project challenges and the Borrower’s needs. The Bank Team also went beyond pure supervision, advising the Borrower on areas outside of the Project’s immediate scope (such as COMPESA’s PPP) that would benefit the RMR. The Borrower’s report (See Annex 6 – Summary of the Borrower’s ICR and Comments on Draft ICR) emphasizes the high value that the implementing agencies’ attached to the team’s guidance throughout implementation. Shortcomings in supervision included: (i) In hindsight, in the early stages of the Project, a somewhat earlier-timed restructuring could have helped address some implementation issues ; and (ii) the Bank could have played a more active role in the early stages of the Project in advising the Borrower to restructure the M&E framework and supervise its implementation. These shortcomings are considered moderate vis-à-vis the overall Bank role in supervision and its contribution in implementation support. Justification of Rating for Overall Bank Performance Rating: Moderately Unsatisfactory 75. The Bank’s performance at entry is rated as Moderately Unsatisfactory and the Bank’s supervision is rated Moderately Satisfactory, on account of the extensive implementation support and proactivity manifested in this complex, multi-sectoral and multi-jurisdictional Project, which lead to a combined rating of Moderately Unsatisfactory. 5.2 Borrower Performance Government Performance Rating: Moderately Satisfactory 76. The GSP and the Municipalities of Olinda and Recife provided clear evidence of strong political and administrative commitment throughout preparation and during the last years of implementation. The Project was considered to be a priority for both the State and the Municipalities. When faced with cost increases, all institutions made an effort to ensure adequate resources were mobilized and made available to support the planned Project activities. The Municipalities, however, faced setbacks because of political and management changes and fiscal difficulties. In response, the GSP provided additional counterpart funds and helped coordinate efforts with the two municipalities. The Governor has been involved in 18 resolving financial issues with both municipalities as PROMETROPOLE was part of the state monitoring system, which helped the State act in a proactive and timely manner to solve coordination, contract and financing issues. The main shortcoming that affected the GSP’s performance was the lack of a formal and efficient enabling environment for metropolitan governance and coordination between the State and the municipalities. Implementing Agency or Agencies Performance Rating: Moderately Unsatisfactory 77. The PMU was responsible for coordinating the institutional agencies’ actions, monitoring and evaluating implementation, and strengthening the capacity of the municipalities to implement the Project. PMU staff members were able to maintain financial management arrangements compatible with Bank requirements, and counterpart funds were available throughout the last years of Project implementation. Project accounts (including contracts for consultants, goods, and works), as well as the deposits and withdrawals from the Special Account, were audited each year under terms and conditions satisfactory to the Bank. 12 Procurement activities were also carried out in accordance with the Bank's Guidelines. The PMU’s efforts, however, were not always successful; difficulties they encountered included: (i) financial shortcomings in the initial years of implementation; (ii) challenges in assuring adequate exchange and flow of information; and (iii) obtaining key documentation from the municipalities in a timely manner. In addition, the PMU often lacked the ability to harmonize lines of action between the Governor and the Mayors. 78. Although the Project’s initial shortcomings were at the State level, later shortcomings were at the municipal level. The Municipality of Olinda lacked capacity and experience in executing large scale infrastructure works and managing resettlement. Recife faced financial shortcomings, slow decision making processes, and difficulties in defining areas for resettlement. COMPESA did not prioritize the Project, possibly because the works were deemed too small in comparison to the company’s portfolio to be given priority. The Municipalities also struggled to report on the development of the activities in their respective areas. 79. If the ICR Team were able to rate the State, the Municipalities and COMPESA separately, the State would receive a rating of “Moderately Satisfactory, while the others would receive “Moderately Unsatisfactory.� Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 80. Despite the shortcomings discussed above, the ICR Team rates the overall Borrower’s performance as “Moderately Satisfactory.� The Project’s outcomes were largely achieved; the Borrower’s commitment grew over the life of the Project and is expected to be maintained until all pending activities are concluded; the Borrower demonstrated an extraordinary effort to gather resources to mitigate the effects of the increase in Project costs; and the PMU has managed to overcome design-related institutional complexity in a fairly satisfactory manner. 12 There are some auditing issues that need to be responded to by the Borrower. They relate to aspects, such as: financial management system, accounting of amounts within the SIAFEM system, and bank conciliations of Recife and Olinda municipalities. These aspects are registered in two management letters sent to the Borrower (February 27 and April 4, 2012), both filed in WBDocs. 19 Section 6: Lessons Learned 81. Integrated urban upgrading, WSS and housing approach. The integrated urban upgrading approach used during the Project improved efficiency, reduced Project risks, promoted synergies between the various upgrading activities and increased the magnitude of the results. It allowed the implementing agencies to build their capacity to work in a coordinated manner. In addition, the Project found the division of the targeted areas into UEs to be immensely useful. 13 The UEs were based on technical aspects rather than political or administrative ones, improving the Project’s legitimacy. They also promoted an integrated approach at the territorial level and facilitated contracting processes and the management of works. Another integrated aspect of the Project that functioned very well was the hiring of one firm to manage all upgrading works in each UE, including housing and payment for provisional subsidized rents. Having a social team as part of the works contracts also improved the ability of the engineers and the social workers to coordinate their action. As one firm was left accountable for results, the risk of work being left unfinished or of varying degrees of quality was lessened. The integrated approach is already being replicated in other Projects, including the Recife Urban Development and Social Inclusion Project (P089013 – “Capibaribe Melhor�). 82. Reducing design complexities. The involvement of the municipalities as both implementing agencies and co-financiers proved to be important for the outcomes of the Project and for future sustainability. However, this management model also led to difficulties in implementation and coordination. Issues such as municipal autonomy, distinct technical and managerial capacities and financial constraints contributed to Project delays and required high level intervention from the Governor and Mayors. Also, the long list of expenditure categories made the Project’s financial management very difficult. Several adjustments to the Loan Agreement were required to simplify these categories during the life of the Project. Simplified institutional and implementation arrangements would have kept the municipalities engaged and would have facilitated project implementation. 83. Enforcing clear cut-off dates for resettlement. The Bank’s Resettlement Policy (OP/BP 4.12) recommends setting a cutoff date at the same time that the census is completed. After this date, no entitlements should be offered to people who settle in the area or make improvements to their homes or businesses. Under the Project, resettlement was carried out in stages, and the Project did not enforce the cutoff dates for resettlement eligibility. As a consequence, many people occupied land in affected areas after the census had been carried out. When a second census was carried out from 2006 to 2008 there was a 37 percent increase in families eligible for resettlement. This was brought to the attention of the Project team during supervision, but the Borrower decided that excluding families from benefits would be sensitive politically. The Borrower thus accepted the additional families for resettlement even if they had arrived just before resettlement began. As a result of this unexpected increase in the number of families to be resettled, the Project closed with a considerable number of families living in subsidized rental housing, which placed an unexpected burden on the State and Municipal budgets. While it is politically difficult to deny eligibility to new squatters, the lack of an established cutoff date makes it impossible to plan. The best way to overcome this issue is to communicate clearly with the community involved as well as local leaders that the cutoff date will be enforced. 84. Discussing proper urban planning and land use control mechanisms during Project Preparation. In Brazil land use and building regulations are controlled at the municipal level. Although there is an advanced overall legal framework, the available instruments for construction in low income 13 These territorial management / intervention units were identified in the Borrower’s “Estudo de Consolidação e Complementação de Diagnóstico Sobre a Qualidade das �guas, Relativo à Preparação do Programa de Investimentos nas Bacias dos Rios Beberibe, Jaboatão e Ipojuca (PQA/PE)�, financed under the Bank’s Federal Water Quality and Pollution Control Project. 20 settlements need to be developed and approved on a case by case basis. Housing development was one of the key themes targeted by Prometropole. Following good practice, the Borrower decided to build single family dwellings and town houses to preserve the traditional way of living of poor families. The design of the houses, however, did not allow the formal construction of business spaces due to rigid municipal urban regulations that prevent mixed use buildings in residential areas. As a result, and despite the regulation standards, many residents expanded their homes for residential purposes or to accommodate small business, such as bars, hair dressing and convenience stores. The available design patterns did not reflect the preferences and needs of some of the residents of the low income areas. Insisting on these designs posed an additional risk to the sustainability of the investments given that new home owners continued to build additions, oftentimes over public spaces, thus affecting the overall urban and environmental quality of the housing complexes. Drawing on this learning experience, the ICR team recommends that Bank financed operations incorporate TA in the early preparation and implementation stages in order to: (i) support a “sophisticated� design that will increase environmental and social sustainability of investments and the beneficiaries’ “say-so� in the design process (perhaps through giving each beneficiary family a list of design options to choose from); and (ii) conduct a wide institutional dialogue and review necessary regulations to maximize the success of the housing developments. 85. Rebalancing pre and post construction activities. The Project focused on improving the living conditions of the residents of the Basin through infrastructure and housing upgrades. The Project did not place equal emphasis on how to sustain its results and promote the autonomy of the families. Slum upgrading projects usually have more intensive social and community development at the planning and implementation stages than at the post-implementation stage. Continuing the social and environmental initiatives and including economic development initiatives would have improved ownership, social control and commitment to asset management and would have reduced the risk to the PDO. The PROMETROPOLE II integrates slum upgrading with revisions on land use planning and construction codes, economic development, income generation opportunities, and institutional capacity building activities. Section 7: Comments on Issues Raised By Borrower/Implementing Agencies/Partners 86. There are no substantive differences between the Borrower and the Bank’s evaluation of the Project. The Borrower’s analysis and comments were incorporated into the ICR. The only issue raised that requires Bank reaction was the effect the large turnover in Task Team Leaders (TTLs) – four in total – and Consultants – innumerous – had on the Project. The Borrower indicated that this turnover hindered the timely progress of the Project because of the differences in the TTL’s philosophies and methods for approaching the Project. The Bank team faced more turnover than normal because of the length of the Project, which from the concept review to closure ran 12 years. The Bank team, however, is taking note of this comment and will investigate ways in which hand-offs between TTLs can be managed more smoothly. The Borrower also noted that the initial Bank teams were rigid at times and were slow in reviewing additional terms and work contracts, which contributed to the overall delay of the Project. The Borrower, however, highly complemented the Bank’s flexibility, technical assistance and responsive supervision during the later years of the Project (See Annex 6 – Summary of the Borrower’s ICR and Comments on Draft ICR). 21 Annex 1: Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Latest Estimate 14 (USD millions) Actual Percentage of Appraisal Estimate Components As per the 4th (USD millions) Appraisal (USD millions) Amendment to the June 30, 2012 (Actual) LA Part A 11.80 6.52 5.39 46% Part B 5.50 8.27 7.38 134% Part C 52.05 85.17 83.71 161% Subtotal 69.35 99.95 96.48 139% Part D 10.05 6.91 5.28 52% Part E 4.14 13.00 14.62 353% Total Baseline Cost Physical Contingencies 0.00 0.00 - Price Contingencies 0.00 0.00 - Total Project Costs 83.54 119.87 116.38 139% Front-end fee IBRD 0.46 0.46 0,46 - Total Financing Required 84.00 120.33 116.84 139% (b) Financing Latest Estimate Actual Percentage Appraisal (USD (USD of Appraisal Type of Estimate Source of Funds millions) millions) (Actual) Cofinancing (USD As per the 4th June 30, 2012 millions) Amendment to the LA Borrower 38.00 74.33 70.84 186% State of Pernambuco 21.00 47.55 45.57 217% Olinda 3.50 10.66 10.38 296% Recife 13.50 16.12 14.89 110% International Bank for Reconstruction and 46.00 46.00 46.00 100% Development Total 84.00 120.33 116.84 139% 14 The “Latest Estimates� column represents the expected Project costs when all works are finished. 22 Annex 2: Outputs by Component BRAZIL: Recife Urban Upgrading Project (Prometropole, P049265) As of June 30, 2012 Status (Bidding, In Borrower IBRD Component/Activity progress, TOTAL Financing Financing (After Loan Amendments) Complete, or (US$) Excluded / (US$) (US$) Cancelled) Part A: Beberibe River Investments 3,962,345 1,427,438 5,389,783 A Construction and Rehabilitation of 2,911,656 1,427,438 Roadways Presidente Kennedy Avenue In progress 2,911,656 1,427,438 •Implementation of a stop exclusively for the (93%) omnibus; • Improvement of pavement in CBUQ for the entirety of the roadway; • Drainage along the Giriquiti road; • Construction of 8 stops; • Relocation of wired electric network; • Horizontal and vertical signalling; • Social attendance for those who had land expropriated • Drainage at omnibus stops 1 through 8; • Earthwork and concrete pavement at the omnibus stops 1 through 8; • Expropriations. Acquisition of land for the Presidente Complete 1,050,689 ­ Kennedy Avenue Rehabilitation (UT Prometrópole) Part B: Water Distribution 4,481,027 2,901,318 7,382,345 B1 Installation and Improvement of Networks 4,367,365 2,762,397 Improvement in the distribution of water in the Complete 2,686,306 1,880,218 Region of ELO Alto do Céu Improvement in the distribution of water in Complete 1,681,059 882,179 Olinda's historical area B2 Acquisition of micro-meters Complete 100,135 122,387 B3 Acquisition of macro-meters Complete 13,527 16,534 23 Status (Bidding, In Borrower IBRD Component/Activity progress, TOTAL Financing Financing (After Loan Amendments) Complete, or (US$) Excluded / (US$) (US$) Cancelled) Part C: Beberibe Low-Income 55,165,014 28,546,092 83,711,106 Investments C1 Urbanization of the Pilot Areas 15,812,931 13,055,465 UE 23 - Campo Grande - Capilé - Recife Complete 4,284,383 4,164,346 • Construction of 187 housing units (179 units in Capilé and 8 in Saramandaia); • Roads in living areas; • Construction of a bridge above the Vasco da Gama/Arruda Canal; • Construction of a footbridge above the Vasco da Gama/Arruda canal; • Improvement of the water supply system; • Implementation of a wastewater system; • Earthwork, drainage and paving of the roads; • Extension of the Av. Prof. José dos Anjos (between Av. Jerônimo Vilela and Rua Farias Neves); • Repair of the lining of the Farias Neves Canal • Social attendance; • Environmental education. UE 23 - Campo Grande - Saramandaia - Complete 7,044,598 5,632,017 Recife • Construction of 292 housing units in Saramandaia and 8 in Capilé; • Roads in living areas; • Construction of a sewage pumping station and outfall in ETE Peixinhos; • Construction of a pedestrian bridge over the Farias Neves canal in order to link the existing Capilé road with a new, planned section of the Capilé road; • Improve the water supply system; • Implement a wastewater system; • Earthwork, pavement and drainage of the roads; • Extension of the Av. Prof. José dos Anjos, that lines the embankments of the Canal Vasco da Gama/Arruda canal (stretch between the Farias Neves Road and the Av. Correia de Brito). • Construction of a local office and a soccer field. • Social attendance; • Environmental education. 24 Status (Bidding, In Borrower IBRD TOTAL Component/Activity progress, Financing Financing (After Loan Amendments) Complete, or Excluded / (US$) (US$) (US$) Cancelled) UE 17 - Passarinho - Olinda In progress 4,483,950 3,259,102 • Construction of 39 housing units, (99%) • Roads in housing areas; • Earthworks, paving and drainage of the Estrada do Passarinho and other roadways; • Drainage system; • Implementation of the sewage pumping station and outfall of 200mm; • Implementation of first model of the sewage treatment center (E. T. E. Minerva); • Improvement of the water distribution network; • Construction of a water reserve with a capacity of 550m3; • Reform of a house for the implementation of a local office; • Implementation of an omnibus station to accomodate 8 operating buses and store 12; • Macrodrainage - Rua 20B Canal; • Social attendance; • Environmental education. Acquisition of land for the pilot area (UT Complete 2,278,465 ­ Prometrópole) C2 Urbanization of the rest of the areas 33,918,755 14,771,246 Jacarezinho - Recife Canal (*) Complete 8,117,085 6,182,301 • Construction of 299 housing units; • Improvement of the water supply system; • Implementation of a drainage system; • Earthwork, pavementing and drainage of roadways; • Macro draingae: Construction of 780m of the Jacarezinho Canal; • Living area roads; • Drainage and pavement, drainage of 24 roadways perpendicular to the canal and 2 on the banks of the canal; • Implementation of sewerage collectors (Transport of sewage for ETE peixinhos). (*) Although Jacarezinho is located in UE 21, the works are separated from the ones in UE 21 because they were executed under different contracts. 25 Status (Bidding, In Borrower IBRD TOTAL Component/Activity progress, Financing Financing (After Loan Amendments) Complete, or Excluded / (US$) (US$) (US$) Cancelled) UEs 20 Arruda and 21 Campina do Barreto In progress 7,235,997 3,074,337 - Recife (Jacarezinho removed) (84.70%) • Construction of 282 housing units, 227 located in UE23; • Implementation of a water supply system in the area; • Implementation of a sewerage system; • Paving of the Av. Prof. José dos Anjos, stretch of the Desenhista Eulino José dos Santos/ Rio Beberibe Road; • Earthwork, paving and drainage of 16 roadways; • Conclusion of the continuation of the Jacarezinho Canal (Rua das Moças/Rua Bom Conselho/Rua da Regeneração); • Transposition of a water main (800mm) above the canal; • Paving sections of roads; • Household connections. UE 03 Dois Unidos - Recife In progress 4,790,516 2,239,506 • Construction of 92 housing units; (83%) • Roads in living areas • Improvement of water supply systems; • Implement a sewerage system; • Earthwork, pavement and drainage of 12 roadways; • Execution of macro-drainage works (lining of canals); • Urbanism and landscaping; • Construction of a sewage pumping station. • Reform a house for the implementation of a local office; • Paving Av. Beira Rio; Acquisition of Land - UEM Recife Complete 2,239,987 ­ UE 07 Varadouro - Olinda (BIRD) Complete 3,395,718 3,275,102 • Construction of 103 living units; • Water supply system; • Sewerage system; • Earthwork, pavement and drainage of the roads; • A stretch of Park along the Canal da Malária; UE 07 Varadouro - Olinda (OGU) Complete 10,379,439 ­ • Construction of 128 living unites for resettlement; • System of Macro-Drainage (lining of the canal); • Drainage and Earthwork; • Implementation and paving roadways; • Urbanization of an area along the Canal da Malária; • Local office. Acquisition of land - UEM Olinda Complete 326,291 ­ A1 Parks 588,585 719,381 Ecologic Park along the banks of the Beberibe Excluded River 26 Component/Activity Status Borrower IBRD TOTAL (After Loan Amendments) (Bidding, In Financing Financing progress, Complete, or (US$) (US$) (US$) Excluded / Cancelled) Cultural Center/Matadouro de Peixinhos Complete 588,585 719,381 Park Part D: Complementary 1,834,077 3,449,832 5,283,909 Urbanization Activities D1 Basic/Executive Projects UEs Part C 1,336,772 2,907,052 Urbanization Projects UE 03 Dois Unidos - Complete 112,741 263,062 Recife Urbanization of the Jacarezinho - Recife Canal Complete 33,483 78,128 Urbanization Projects UEs 20 Arruda e 21 Complete 278,022 648,718 Campina do Barreto - Recife Urbanization Projects UEs 04, 08, 17 e 19 - Complete 591,398 1,379,928 Recife Urbanization Projects UEs 11, 12 e 13 - Olinda Complete 321,128 537,217 D2 Basic/Executive Projects Parts A and B and 497,305 542,780 Complementary Studies PSI UE 07-Olinda,Project ETE Minerva and Complete 88,846 184,242 Evaluations of Expropriations Projects that complement the living units in Complete 8,130 18,971 UE 23 - Campo Grande - Recife Management System Matadouro de Peixinhos Complete 12,788 29,839 Supra-Local Road System (Av. Presidente Complete 65,950 153,883 Kennedy) Publishing the city's strategy Complete 30,724 64,521 FM System Complete 7,838 17,350 English translation ICB 01/2005 Complete 2,565 5,985 Attendance works Matadouro de Peixinhos Complete 7,977 18,613 Formation of the Prometrópole II In progress Evaluation of the Prometrópole Program In progress 194,043 Audit of the Prometrópole Program Complete 78,444 49,376 Part E: Project Management 6,995,504 7,622,652 14,618,156 E Equipment 166,641 277,487 UT Prometrópole Equipment Complete 82,758 152,112 UEM Recife Equipment Complete 63,780 84,806 UEM Olinda Equipment Complete 20,103 40,570 E3 Management consulting for the Project 6,828,863 7,345,164 UT Prometrópole Management In progress 4,740,556 4,228,205 UEM Recife Management Complete 690,859 1,027,328 UEM Olinda Management Complete 1,397,448 2,089,631 Interest Bank Administration ­ 460,000 460,000 Adjustements in the Bank financing percentages made through the -1,592,669 1,592,669 0 4º Amendment to the LA 70,845,298 46,000,000 116,845,298 27 Annex 3: Economic and Financial Analysis 1. This section contains a cost-benefit analysis of the Prometrópole program. In the analysis, whenever possible, a comparative analysis of the parameters established in the 2003 Project Appraisal Document (PAD) was undertaken to measure the expected (ex ante) results of the Prometrópole program against the actual (ex post) results when the Prometrópole program ends. METHODOLOGY 2. This cost-benefit analysis of Prometrópole estimated the economic benefits of the Project indirectly, via the increase in property values, based on a hedonic price function. With this methodology, consumers’ willingness to pay for goods and services not available on the market can be estimated indirectly, using the increase in real property values in areas benefited by Prometrópole as a proxy. The hedonic price function is thus able to capture—through the increase in property values following the Project’s interventions—the improvement in the welfare of the population attributable to the various benefits generated by Prometrópole. 15 3. The economic benefits of water service, on the other hand, were estimated based on the census survey conducted among communities in the areas benefited by Prometrópole, using the following variables: average water bill, number of residences benefited, and percentage of residences with water meters. The marginal increase in the receipts of the water distribution company (COMPESA) was used to estimate the economic benefits of the program. 4. The payment of urban property taxes (IPTU) and taxes on the circulation of goods and services (ICMS, electricity) was not considered in calculating the economic benefits of the program. Most households in the areas where Prometrópole operates are exempt from IPTU, pursuant to the special law benefiting low-income households and small dwellings. This fact is borne out by the census survey, which indicated that 85 percent of households do not pay IPTU. As for the levying of the ICMS on the electric bill, a decree signed in 2007 exempts households consuming up to 220 KW per month from the payment of ICMS. According to census survey findings, the average consumption of households located in areas where Prometrópole operates is 105.62 KW, which exempts them from paying ICMS on their electric bills. 5. The capital cost of the investments made and the operating and maintenance costs were considered in estimating the program’s economic costs. The capital cost is defined as the sum of the financial investments made by the Prometrópole program, which, according to the disbursement spreadsheet of December 29, 2011, amounts to R$205,153,268.00, or US$113,974,038. The maintenance cost was estimated at 1 percent of total infrastructure investments. The operating cost was estimated at 40 percent of the maintenance cost. The percentages used are in line with the pattern of maintenance and operating costs of Brazilian public entities. ECONOMIC BENEFITS 6. A program such as Prometrópole generates a wide range of positive benefits for communities involved directly in the areas of intervention and for neighboring regions that receive such improvements indirectly. Some of these benefits are: • Improved water and sanitation services; 15 For more information on the hedonic price methodology, see Freeman (1993), Court (1941), Ridker and Henning (1967), Griliches (1971) and Rosen (1974). 28 • Better public health, with the reduction of various diseases such as diarrhea, leptospirosis, verminosis, dengue fever, etc.; • Improved natural environment owing to the reduction of pollution caused by the lack of sanitation; • Greater accessibility as a result of drainage and paving, which reduces the cost of transportation and travel time; • Improved habitability conditions and the consequent reduction of the risks of landslides and flash floods; • Provision of high-quality public services, such as street cleaning and public lighting; • Higher levels of personal and property security; • Improved leisure and sport activities; • Expanded income-generating opportunities; • Enhanced self-esteem and sense of citizenship among residents. 7. The benefits generated by a project such as Prometrópole are innumerable and have a major impact in terms of improving the welfare of the population. However, measuring all these benefits in monetary terms is not feasible; first, because the goods and services in question are intangible, and, second, because their prices are not determined by the market, as most of them are public goods. 8. For this reason, analyzing the benefits of a social project such as Prometrópole requires an indirect approach in order to infer, in monetary terms, the economic gains generated by the new environment created by the program. For this analysis the Hedonic Price Method was chosen to estimate the economic benefits generated by the Prometrópole program. 9. The Hedonic Price Method originates in consumer theory. It is assumed that the price of a good or service is a function of a series of characteristics that consumers take into account in making a decision. In the case of the price of a residence, the consumer is willing to pay a certain amount, based on a set of structural characteristics (size of the house, type of construction, number of bedrooms and bathrooms, etc.), the availability of public services (water and sanitation, paved streets, trash collection, public lighting, security, schools, childcare facilities, health centers, etc.), and environmental characteristics (risk of landslides, pollution, neighborhood, public common areas, etc.). 10. The choice of this methodology to estimate the economic benefits of Prometrópole is relevant, as most of these characteristics are directly or indirectly affected by the program initiatives. Thus, determining the prices of residences located in these areas synthesizes the economic gains derived from the series of positive, post-Prometrópole benefits. INCREASE IN PROPERTY VALUES 11. The first step in the hedonic price method is to obtain information on real estate prices, either directly by surveying owners and residents or indirectly by obtaining data from the municipal IPTU office, for example. For purposes of evaluating Prometrópole, data on real estate prices was obtained from the residents of properties in the area where the program operates. One of the specific aims of the census survey was to obtain three basic types of data: (i) in applicable cases, the purchase value of the property offered by third parties; (ii) the value assigned to the property by the resident; and (iii) the value of a property with the same characteristics located in the same community. Based on these data, the property value used in the hedonic function estimate is the simple arithmetic mean of the three values. The choice of this objective measurement reduced possible biases in the determination of property values. 29 12. For real property values, the hedonic price method consists of estimating how various characteristics of the properties (e.g., size, number of bathrooms, type of walls, floor, roof, etc.) contribute to those values. Formally, a hedonic function can be expressed as: P = Xβ + ε where P is the value of the property (as a natural logarithm) obtained in the census survey and discussed above, X is a matrix containing various characteristics of the property and the community that could affect property values, and β is a measurement of the importance of each of the characteristics of the properties and ε represents elements not included in the estimate, i.e., the part not explained by the variables included in X. 13. An important aspect of this methodology is choosing which variables (i.e., characteristics of the properties and the community) are likely to be major determinants of property values. It is assumed that various characteristics, both quantitative and qualitative, are important, but in many cases it is difficult to obtain an accurate measurement of such characteristics, and when they are sufficiently dispersed among properties the contribution may not be statistically significant. The choice made in this assessment was to include in the estimate only those variables that are statistically significant. 14. The hedonic function was estimated in two stages. In the first stage, a number of interesting variables were included among the possible explanatory variables of real property prices. With these variables in hand, the regression model was estimated using the ordinary least squares method. Based on the results of the regression, the statistically significant variables were identified. The second stage involved estimating the model only with the variables that proved to be significant in the first stage. It is also worth noting that before the estimates were made, the data were statistically processed to exclude observations that were far above or far below the average (discrepant observations). 15. The variables included in the second stage were: stories indicates the number of stories (floors) of a property; rooms represents the number of rooms in the residence; freehold title, which is a binary variable (i.e., only two values are possible; it has a value of one when the residence has that characteristic and a value of zero when it does not) indicating whether or not the owner has freehold title to the property; masonry is a binary variable indicating whether the walls of the residence are made primarily of bricks; interior plasterwork is a binary variable used to indicate whether the interior walls are plastered; tile floor indicates whether the floor is a tile floor; slab roof is a binary variable indicating whether residences have a slab roof; and number of bathrooms indicates the number of bathrooms in each residence. 16. Several other variables in addition to the above were included: water meter is a binary variable indicating whether the residence is equipped with a water meter, which in this case serves as a proxy for water and sewage services; Celpe (Energy Company of Pernambuco) is a binary variable indicating whether owners pay their electric bills on time. The purpose of the latter two variables is to indicate not only the availability of services but also whether households have a meter and pay their electric bills. However, the census survey identified some households that use electricity supplied by Celpe, but their connections were illegal. The Celpe variable therefore refers only to households with legal connections. The underlying idea is that properties that have metered services and whose owners are in good standing tend to be worth more. 17. Variables such as the following were also included: mail service, which is a binary variable indicating whether or not mail is delivered to the residence by the postal service (the idea is that mail service can indicate ease of access both for the residents themselves and for the receipt of goods and 30 services); landline telephone, which is a binary variable indicating whether the residence has a residential landline telephone; IPTU, which is a binary variable indicating whether the owner pays IPTU (property tax) on the property, which, like ownership, reveals the degree of formal marketing of the property; and, finally, drainage, which indicates the quality of drainage services available to the residence. 18. The results of estimating the hedonic price function for the area in which Prometrópole operates are shown in Table 1 below. Table 1: Hedonic Price Regression (dependent variable: natural logarithm of the average value of real properties) Variable Coefficient Standard error t statistic P>|t| Stories 0.0469 0.0270 1.7300 0.0830 Rooms 0.0175 0.0042 4.1600 0.0000 Masonry 0.1918 0.1023 1.8800 0.0610 Interior plasterwork 0.1501 0.0387 3.8800 0.0000 Ceramic tile floor 0.2005 0.0339 5.9100 0.0000 Slab roof 0.1042 0.0363 2.8700 0.0040 Number of bathrooms 0.1561 0.0357 4.3700 0.0000 Water meter 0.0844 0.0332 2.5400 0.0110 Celpe (electric company) 0.1459 0.0394 3.7000 0.0000 Mail service 0.3414 0.0363 9.4100 0.0000 Landline telephone 0.2506 0.0347 7.2100 0.0000 IPTU (property tax) 0.2713 0.0463 5.8600 0.0000 Drainage 0.0935 0.0305 3.0600 0.0020 Freehold title 0.1430 0.0298 4.8000 0.0000 Constant 8.8550 0.1095 80.9000 0.0000 R2 – adjusted 0.3418 Number of observations 1707 Source: Prepared by Datamétrica based on census survey. Note: Semi-log model. 19. Some conclusions can be drawn from the results shown in Table 1. First, as expected, properties whose owners have freehold title are worth more. This was expected, inasmuch as holding proper title is an essential condition for selling a property on the formal market. The number of rooms and the number of stories, which were used as approximate measurements of the size of the property, also contribute positively to property values. 20. Characteristics of the quality of a dwelling, such as whether the walls were made of bricks, whether the interior walls were plastered, whether the floor was a ceramic tile floor, and whether the roof was a slab roof were also included in the estimate, and the results show that all these characteristics add to property values. Consequently, properties with these characteristics are worth comparatively more than those without them. Another variable that appeared as a major determinant of property values was the number of bathrooms and, in this case, the contribution was also positive. 21. The results of estimating the hedonic function also revealed that properties with water meters and owners who pay electric bills on time and are current in the payment of IPTU to the prefecture are also worth more. Moreover, properties with access to mail service and where the evaluation of drainage 31 services was positive also appraise higher. It is worth noting that these characteristics will be taken into account in Prometrópole’s interventions. 22. With the results of estimating the hedonic function in hand, the next step was to estimate how much Prometrópole contributed to the increase in property prices in areas where it is directly involved. The method consisted of estimating the difference in value between properties that have characteristics related to Prometrópole’s initiatives and those without such characteristics. The following table shows the results of this analysis. Table 2: Increase in property values in areas benefited by Prometrópole Average Value US$ Residence R$ (Exchange Rate: US$1.00 = R$1.80) Pre-urbanization 30,469.89 16,927.72 Post-urbanization 43,699.75 24,277.64 13,229.86 7,349.92 Average increase in value 43% 23. Properties without characteristics such as drainage, freehold title, mail service (proxy for degree of urbanization of the residence), etc. were valued, on average, at R$30,469.89, or US$16,927.72. Conversely, properties with all the characteristics (post-urbanization) were valued at R$43,669.75, or US$24,277.64. The difference, therefore, was R$13,229.86 (US$7,349.92), representing an increase in value of 43 percent. 24. For resettled households, the estimated increase in property values was, on average, equal to the average value of a property included in the project, i.e. R$43,669.75 (US$24,277.64). These values will be used in the cost-benefit analysis of the Program. BENEFITS BY COMPONENT Investment in the Beberibe Basin 25. The Project’s infrastructure investments had a positive impact on all areas located in the Beberibe Basin and on neighboring areas as well. To estimate the impact of this component of the Project, an estimated total universe of 89,706 households benefited by Prometrópole investments was selected. Using the coefficients estimated in the hedonic price regression, the estimated benefits are R$1,748.00 per household, or US$971.11. The estimated investment cost per household is R$101.20, or US$56.20. 26. A comparison with the results of the PAD (2003) shows that the benefits obtained were greater than initially estimated. The estimated benefit, measured by the increase in property values, amounted to US$377 for 85,455 benefited households. The per-household investment cost of this component was calculated as US$207, for a total investment of US$17,689,185. 27. The rise in real estate prices following the launch of Prometrópole was far greater than the increase estimated in the PAD (2003). It must be stressed that, in addition to the gains achieved through program-related improvements in characteristics that increase the value of real property, macroeconomic conditions in Brazil have improved substantially since the PAD (2003) was prepared. The economic growth observed in recent years, with positive repercussions in terms of job creation and distribution, have led gradually to a major change in the economic conditions of a large segment of the Brazilian population. The corollary of this phenomenon is a steep rise in real estate prices on the Brazilian market, 32 sustained by the growth of employment, income and credit. It is possible, then, that the estimated post- program rise in real estate values also reflects the current upward movement of the national real estate market. 28. However, the spillover effects of the increase in Brazilian real estate prices on properties located in areas benefited by the program are only possible thanks to the positive externalities generated by urbanization and the availability of public services in localities that previously lacked them. 29. Although the per-household investment cost was lower than initially projected in the PAD (2003), the fact is that the investments included in the original scope of the program, totaling close to US$17.6 million for the component “Investment in the Beberibe Basin,� turned out to be much smaller, amounting to about US$5.0 million. As a result, the per-household investment cost of US$56.20 was well below the PAD (2003) projection of US$207. Investment in low-income areas 30. Investments in this component were estimated to benefit 35,000 households. The increase in property values was estimated in the PAD (2003) at US$2,963 per residence and the per-household investment cost was estimated at US$1,867. 31. The results in terms of the number of households benefited were well below the original projection. The number of households that benefited directly from Prometrópole’s interventions in low- income areas was 5,040. This development is explained by the fact that the original scope of the program was reduced, as discussed in the previous section. 32. It should also be noted that, as discussed in previous sections, projects prepared for the Prometrópole program but not implemented are being implemented through other funding sources, particularly the PAC. Thus, although the number of households benefited in low-income areas was lower than initially projected, it may be that when all the projects are concluded, thanks to the use of other funding sources, the number of households benefited will be larger. 33. The increase in value estimated with the hedonic price function was R$13,229.86, or US$7,349.92, which is larger than the value calculated in the PAD (2003). This outcome points to a significant increase in the value of real property in geographic areas benefited by Prometrópole. In addition, the estimated per-household investment cost was R$16,605.58, or US$9,225.32. This is well above the US$1,867 originally projected in the PAD (2003), inasmuch as the number of households benefited (5,040) was much lower than the initial estimate of 35,000 households. Moreover, civil construction costs in Brazil have risen much faster than inflation, leading to an overall increase in Prometrópole’s project budgets. Resettlement 34. The original scope of the Prometrópole program called for the resettlement of 2,132 households. Adjustments to the program limited the number of households resettled to about 1,400. These households, the most economically and socially vulnerable segment of the population benefited by the Project, received the greatest economic benefits. The resettlement of these household groups to standard housing with a full range of public services produced the greatest increases in property values relative to the other components of the program. 35. The increase in this group’s property values was estimated at R$43,699.75, or US$24,277.64 per household, well above the US$3,774 per household projected in the PAD (2003). This outcome points 33 unequivocally to the transformation of the lives of these people in terms of improved welfare, following the investments made by Prometrópole. Water Distribution 36. One of the components of Prometrópole consisted of investments in water distribution and the sectorization of water distribution. The marginal increase in the receipts of the water distribution company, COMPESA, was used to estimate the economic benefits of the program. The sum of R$43.03, or US$23.91 was chosen as the average water bill, based on the findings of the census survey. In addition, according to data provided by COMPESA, 7,202 connections were installed in the geographic area served by Prometrópole. With an average payment rate of 84 percent, the annual economic benefit from water distribution was estimated at R$3,123,813.00, or US$1,735,452. 37. As mentioned above, the economic benefits generated by the increase in the IPTU and the ICMS on electric bills were not taken into account, owing to the exemptions from these taxes provided by law for the low-income population. Table 3. Economic Benefits by Component Components Benefits - Value (US$) Part A: Beberibe Basin 87,113,996 Part B: Water distribution 34,709,033 Part C: Low-income areas 70,231,162 Part D: Supplementary activities NA Part E: Project management NA TOTAL 192,054,191 ECONOMIC COSTS 38. Total investment costs and maintenance and operating costs were used to estimate the economic costs of the Prometrópole program. The capital cost was defined as the sum of financial investments, amounting to R$205,153,268.00, or US$113,974,038.00, as indicated in the financial disbursement spreadsheet dated December 29, 2011. The following table provides a detailed breakdown: 39. The maintenance value was estimated at 1 percent of the infrastructure components (Parts A, B and C) and the operating value was estimated at 40 percent of the maintenance value. Table 4: Economic Costs by Component Components Cost - Value (US$) Part A: Beberibe Basin 5,041,863 Part B: Water distribution 7,382,345 Part C: Low-income areas 81,882,201 Part D: Supplementary activities 5,089,865 Part E: Project management 14,117,864 TOTAL 113,974,038 RESULT OF THE COST-BENEFIT ANALYSIS 40. The result of the cost-benefit analysis of the program is estimated based on a comparison between the total economic benefits of the project and the total economic costs of capital and maintenance/operation of the physical investments. To that end, a number of assumptions were made: • The timeframe for analysis of the project was 20 years; 34 • The number of households benefited was estimated based on the census survey conducted in the first four months of 2012; • A total universe of 96,037 households benefited directly and indirectly by Prometrópole was estimated, with 6,440 having benefited directly and 89,706 having benefited indirectly. Included in the total number of directly benefited households are 1,400 resettled households. Table 5: Economic Costs and Benefits by Component Components Cost - Value (US$) Benefits - Value (US$) Part A: Beberibe Basin 5,041,863 87,113,996 Part B: Water distribution 7,382,345 34,709,033 Part C: Low-income areas 81,882,201 70,231,162 Part D: Supplementary activities 5,089,865 NA Part E: Project management 14,117,864 NA TOTAL 113,974,038 192,054,191 41. Two economic benefits figure in Prometrópole’s cash flow: a) The increase in real property values in areas served by the program, used as a proxy of the total intangible benefits of the new environment created by the program. The total increase in real property values in the Beberibe Basin was R$156,805,192.00 (US$87,113,996). In the low- income areas, the increase amounted to R$65,236,442.00 (US$36,242,468); in the resettlement areas the increase was estimated at R$61,179,650.00 (US33,988,694). The total economic benefit of the increase in real property values, then, was R$283,221,285.00, or US$157,345,158.00. The increase in real property values projected in the PAD (2003) was US$143,967,703.00. Thus, the actual increase in value was 9.3 percent higher than originally estimated. b) Establishment of the water distribution and sectorization network was deemed to have generated economic benefits for the project. The annual value was estimated at R$3,123,813.00, US$1,735,451, based on an average monthly bill of R$43.03 (US$23.91), obtained from the census survey, multiplied by the 7,202 new connections installed. The value was annualized and a payment rate of 84 percent was used. 42. The economic costs were estimated as follows: a) The total capital cost was R$205,153,268.00, or US$113,974,038, representing the investments in all components of the project. b) The maintenance cost was estimated at 1 percent of the infrastructure investments; c) The operating cost was calculated as 40 percent of the maintenance cost. 43. Based on the above, the cash flow generated by the annual net economic benefits was calculated as the difference between the total benefits and the total costs involved in all aspects of the project over the course of 20 years. Prometrópole’s internal rate of return (IRR) was 40 percent, indicating the economic viability of the project, i.e., the economic benefits generated by the program outweighed the economic costs. The 40 percent IRR was above the PAD (2003) estimate of 37 percent, indicating that the Government’s decision to plan and implement the Prometrópole program was correct. 44. The Net Present Value (NPV) of Prometrópole was calculated based on two scenarios for the opportunity cost of capital in Brazil. In the first scenario, the discount rate used was 12 percent per annum. The resulting NPV was R$54,422,181.00, or US$30,234,545.00. In the second scenario, the discount rate was 9 percent per annum, the current Special Clearance and Escrow System rate (Brazil’s overnight rate, SELIC). Thus, the NPV rises to R$62,652,667.00, or US$34,807,037.00. In both situations, the NPV is positive, demonstrating the economic viability of the Prometrópole program. 35 45. The net cash flow generated by COMPESA’s water supply and sectorization operation was also calculated. The IRR obtained was 21.7 percent. Comparing this percentage with the 19 percent IRR estimated in the PAD (2003) shows that the economic viability of the water distribution component of the program was fully realized. The NPV using a discount rate of 12 percent per annum was R$8,715,346.00, or US$4,841,859.00, and with a rate of 9 percent per annum the NPV was R$13,589,419.00, or US$7,549,678.00. This indicator also confirms the economic viability of the program. 46. In sum, the cost-benefit analysis of Prometrópole confirms its economic viability and shows that the decision of the government officials involved and the World Bank to implement it was correct. The increase in real property values observed in areas benefited by the program is, without a doubt, key to understanding the impact of a program such as Prometrópole in low-income geographic areas. The provision of essential services by the Government, such as paving, sidewalks, drainage, water and sewage systems, trash collection, public lighting, dredging, etc., clearly contributes to a substantial improvement in the quality of life and well-being of populations previously excluded from these basic aspects of urban life. The economic viability of the Prometrópole program holds enormous potential for social projects designed to improve the living conditions of extremely disadvantaged individuals in large urban centers in Brazil, which in turn increases the well-being of society as a whole. 36 Annex 4: Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Luis Baltar Engineer (Consultant) Alexandre Borges de Oliveira Procurement Specialist LCSPT Procurement Patricia Caicedo Civil Engineer Jose Augusto Carvalho Lead Counsel LEGLA Legal Roberto Chávez Lead Urban Specialist (peer reviewer) Urban Dean A. Cira Senior Urban Specialist LCSUW Task Team Leader Tulio Correa Financial Management Specialist LCSFM Financial Management Paula Dias Pini Senior Urban Development Specialist LCSUW Urban Martin P. Gambrill Senior Water and Sanitation Engineer LCSUW Water and Sanitation George Gattoni Lead Urban Specialist Urban Ivo G. P. Imparato Lead Urban Specialist LCSUW Task Team Leader Jose C. Janeiro Senior Finance Officer LCSFM Financial Management Efraím Jiménez Lead Procurement Specialist LCSPT Procurement Kirsten Oleson Operations Analyst Team member Juan David Quintero Lead Environmental Specialist LCSEN Environment Mariangeles Sabella Council LEGLA Legal M. Vitor Serra Lead Urban Specialist LCSUW Co-Task Team Leader Maria Angelica Sotomayor Araujo Senior Economist LCSUW Economic Analysis Cristina Velazco-Weiss Language Program Assistant LCSUW Team Member Carlos Velez Lead Economist LCSUW Economic Analysis Dorte Verner Senior Economist LCSOS Economic Analysis Supervision/ICR Marcos T. Abicalil Sr. Water and Sanitation Specialist LCSUW Task Team Leader Sinuê Aliram Procurement Specialist LCSPT Procurement Alexandre Borges de Oliveira Senior Procurement Specialist LCSPT Procurement Cidalia Brocca Financial Analyst CTRLN Team member Dean A. Cira Senior Urban Specialist LCSUW Task Team Leader Túlio Correa Financial Management Specialist LCSFM Financial Management Resettlement / Paula Dias Pini Senior Social Development Specialist LCSUW Environment Nicolas Drossos Consultant LCSFM Financial Management Elizabeth H. Eiseman Junior Professional Associate LCSUW Team Member Marcus Vinicius Ferreira da Silva Consultant LCSUW Infraestructure Martin P. Gambrill Senior Water and Sanitation Engineer LCSUW Water and Sanitation Environmental and Daniel R. Gross Consultant LCSUW Social Safeguard Ivo G. P. Imparato Senior Urban Specialist LCSUW Task Team Leader Jose C. Janeiro Senior Finance Officer LCSFM Financial Management Efraím Jiménez Lead Procurement Specialist LCSPT Procurement Emanuela Monteiro Consultant LCSUW Urban Sarah Anthony Murphy Consultant LCSUW Team member Francisco Notarbortelo di Villarosa Consultant LCSOS Social 37 Responsibility/ Names Title Unit Specialty Margaret Rosa Pajaro Consultant Luis Prada Vilallobos Senior Procurement Specialist LCSPT Procurement Maria Angelica Sotomayor Araujo Senior Economist LCSUW Economic Analysis (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD (including travel and No. of staff weeks consultant costs) Lending FY1998 0.00 187,854.22 FY1999 0.00 52,779.72 FY2000 33.71 136,552.87 FY2001 9.70 38,071.06 FY2002 42.35 198,283.60 FY2003 18.88 80,644.08 FY2004 2.53 3,806.47 FY2005 0.00 30.54 Total: 107.17 698,022.56 Supervision/ICR FY1999 0.00 673.33 FY2000 0.43 1,147.86 FY2001 0.05 63.06 FY2002 0.00 74.61 FY2003 0.00 5,331.65 FY2004 20.65 74,831.80 FY2005 21.11 84,8181.21 FY2006 22.99 91,313.31 FY2007 26.24 124,402.69 FY2008 12.38 84,781.98 FY2009 14.35 81,821.05 FY2010 10.33 53,916.59 FY2011 5.26 40,427.02 FY2012 7.35 26,176.29 Total: 141,14 669,779.45 38 Annex 5: Beneficiary Survey Results The Borrower hired an independent consulting firm to develop the Project’s Final Assessment (Prometropole_Relatorio de Avaliacao_25-06-12). The evaluation included a comprehensive beneficiary survey (Prometropole_Resultados da Pesquisa Censitaria). Both documents can be found in the Project files. 39 Annex 6: Summary of Borrower's ICR and/or Comments on Draft ICR 1. This is a summary of the Evaluation Report on the Infrastructure Program in Areas of Low Income in the Recife Metropolitan Region – PROMETRÓPOLE, Service Contract 003/2011, signed by the Pernambuco State Planning and Research Agency – CONDEPE/FIDEM, and Datamétrica Consultoria, Pesquisa e Telemarketing, for the execution of specialized services to evaluate the aforecited program. 2. This evaluation primarily sought to understand and assess whether and how Prometrópole has met its objectives, identify unforeseen outcomes that have been important for improving quality of life for the beneficiary population and the natural and built environment, and pinpoint the causes of any failures or shortcomings with a view to learning lessons that will be relevant for future projects. 3. Three basic approaches were used to meet the objectives of the evaluation. The first was a review and analysis of the most important documents relating to the design, implementation, and monitoring of the work done by the program. The second was a qualitative study using two techniques: in-depth interviews and focus groups. The third involved conducting and analyzing a quantitative census survey of the program’s beneficiaries. A total of 3,835 questionnaires were administered to households located in sanitation units [unidades de esgotamento UEs] 3, 20, 21, and 23 in Recife and 7 and 17 in the municipality of Olinda over the period from December 13, 2011, to February 11, 2012. 4. The evaluation focused on five main tasks: (i) analysis of the efficacy of the program in terms of meeting its targets; (ii) analysis of its efficiency in implementation of the projects; (iii) evaluation of the efficiency of management and administration; (iv) consideration of its sustainability; and, finally, (v) identification of the lessons learned. Below is a brief discussion of each of these points. 5. Evaluation of Efficacy. One of the aspects evaluated was the number of direct and indirect beneficiaries. The former were the people directly benefited by any intervention at the local level (urbanization of the UEs), while the latter were those who benefited from supra-local projects (e.g., upgrading of Presidente Kennedy Avenue, restoration of the Peixinhos Slaughterhouse historic site, sectorization of the Compesa water utility) and those benefited indirectly by interventions at the local level. It is estimated that 6,440 households, or a total population of 22,180 people, were direct beneficiaries, while 89,597 households, or a total of 321,511 individuals, were indirect beneficiaries. The benefits resulting from the investments made by Prometrópole may be summed up as follows: increased property values (the average value of properties with urban amenities is estimated to be 43% greater than that of those without such benefits); improved quality of housing, with an increased percentage of dwellings with masonry walls, plastered interior and exterior walls, ceramic tile roofing, etc.; perceived improvements in the quality of life of beneficiaries thanks to greater and better access to such public services as piped water, sewer connections, street drainage, public lighting, etc.; and improvements in the natural and built environment in Beberibe Basin following removal of dwellings in at-risk areas and subsequent resettlement of the families, creation of new access routes, restoration of the Nascedouro de Peixinhos historic site, and other interventions. These changes led to a positive evaluation of the program by the beneficiary families (60% positive evaluation), despite the many challenges involved in implementing them. 6. Evaluation of Efficiency. The efficiency of the management and administration of Prometrópole was deemed to be satisfactory. The institutional arrangement between the project manager, the state government through UT-Prometrópole and other co-executors, the city governments of Recife/Olinda, and Compesa worked satisfactorily. A number of in-depth interviews were held with the program’s current and former managers and its technical staff to obtain information and learn their impression of the managerial and administrative process followed by Prometrópole. Hiring a project management company 40 to provide additional support alongside the managerial and technical nucleus of permanent government staff was considered to be an important factor in the successful progress of the work. The formation and professional and managerial upgrading of the technical team and the ongoing participation of the communities in the discussions are other characteristics that contributed to administrative efficiency. 7. On the other hand, negative points were also noted, mainly having to do with the frequent turnover of management leaders, which in some cases led to temporary breakdowns in decision-making. In addition, asymmetry between the public agencies involved in terms of their budgetary, financial, and human resources led to delays in dealing with obstacles that arose in the course of program execution. Nevertheless, the high degree of commitment and motivation exhibited by all the professionals involved in Prometrópole made for great willingness to overcome the difficulties and find better solutions, while at the same time improving the program’s managerial efficiency. 8. World Bank participation in the execution of Prometrópole was also the subject of evaluation. Its participation was extremely important, especially its technical support and the opportunity to benefit from the Bank’s knowledge gained from similar projects around the world. This exchange of information was a rich process and an important learning experience about systematic and integrated intervention in low- income areas. The Bank, in addition to serving as the program’s funding agent, definitely contributed to the success of Prometrópole by sharing its expertise in improving quality of life for populations living in extreme poverty. 9. However, some observations were made regarding the World Bank’s management of the program. The main problem was the turnover of consultants and Bank managers responsible for Prometrópole. Since initial conception of the program there have been four managers and countless consultants with different approaches and philosophies. The conclusion is that these changes in Bank management made for discontinuity in the lines of action followed, leading in turn to delays in the program’s timetable. Moreover, sometimes there were long delays in situations where a timely response was needed from the Bank – for example, when approval was needed for contract amendments. 10. The evaluation of administrative efficiency included consideration of the managers’ ability to mobilize resources to make up for cuts that had been made. One of the greatest challenges in the execution of Prometrópole was dealing with funding cuts and having to adapt the interventions originally proposed to the new reality. There was a difference of opinion on how to solve this dilemma. One option was to pick only some of the projects and implement them in their entirety. The other was to finish all the projects agreed upon by redirecting the scope of the program and getting the needed funds from other sources. The latter approach was chosen, and it was decided to create a project bank. This solution turned out to be efficient, since most of the Prometrópole projects that could not be executed were incorporated into new funding lines, such as the PAC. 11. Evaluation of Sustainability. In its original conception, Prometrópole had a clear strategy for subsidizing the capital cost of providing basic infrastructure for low-income communities in the Beberibe Basin. Afterwards, long-term economic sustainability would rely on recovering the operating and maintenance costs from tariffs and taxes paid by the beneficiaries. However, the census survey revealed that income from tariffs and taxes paid by the beneficiary households would not be sufficient to sustain the program’s interventions. It turned out that a relatively high proportion (85%) of households in the program’s areas of action do not pay the local property tax [Imposto Predial e Territorial Urbano IPTU], in part because there is specific legislation exempting low-income households and retirees, as well as properties that meet certain criteria. Hence, income from water and electricity tariffs alone would not be sufficient to guarantee sustainability of the rest of the projects undertaken by the program. In the case of electricity, the tariffs paid by households are low, which means that they would be exempt from paying tariffs for public lighting. The tariffs that households pay for water and sewer services are also low. 41 Illegal connections to the power and water supplies continue to be found, adding to the low probability of long-term economic sustainability based on income from the payment of tariffs and taxes. Also, most of the households that are benefiting from the investments in new housing units have limited income, which places a strain on their ability to keep their homes in a good state of repair. Finally, the long-term sustainability of supra-local and other investments (e.g., primary and secondary roads, drainage, recreational amenities, etc.) is contingent on the public sector (the Pernambuco state government and the municipalities of Olinda and Recife) deciding to take on the operating and maintenance costs. 12. Lessons Learned. From the outset up to the final phases of Prometrópole, all the stakeholders were faced with problems to deal with and a learning curve throughout the process. From this experience, it has been possible to identify some of the lessons learned along the way. The continuity provided by the program’s technical team, made up largely of permanent government staff, was crucial to the smooth development of all phases of Prometrópole. The availability of institutional memory and the makeup of the team are important factors in the success of a social project like Prometrópole. Hiring a management company to provide support on the interventions proved to be another important lesson. The contractor’s agility in making decisions and performing operations, making it possible to work faster and more efficiently on the physical development of the interventions, demonstrated the importance of enlisting such a tool in similar projects. 13. The difficulties encountered in dealing with municipal regulations, especially urban development parameters applicable to the housing units that were built, are prompting both local officials and the Bank to engage in a broader and more practical dialogue about the obstacles created by building codes and environmental regulations for projects like Prometrópole. An important outcome of this lesson has been the proposal of a new legislative framework that will consider the specific case of this type of project, with a view to avoiding problems related to compliance with regulations of this kind in the construction of housing units under the program. 14. The requirement for greater rigor in the technical evaluations during the project preparation phase is another point to be considered. The lack of sufficiently detailed information in the soil studies of the areas where the housing complexes were to be built led to countless delays in the execution phase, in addition to placing a burden on the budget lines for these interventions. Several delays in the timetable could have been avoided early on if more in-depth technical studies had been done of the areas in question. 15. The Infrastructure Program in Areas of Low Income–Prometrópole has showcased a new approach to the problem of poverty and extreme poverty with interventions implemented against the backdrop of an integrated vision of urban and social infrastructure supported by a strategic alliance between the state government, local city governments, and Compesa. Still, it must be recognized that inequalities between the stakeholders involved in the operation in terms of physical, financial, and human resources affected the program’s dynamic. For example, In cases where additional budgetary funds were needed to cover contract add-ons, the negotiations were more time-consuming in the municipalities with limited financial resources, resulting in further delays in completion of the projects. 16. Finally, the most important lesson learned is that it was possible to transform areas previously excluded from the urban fabric into communities with good quality public services and decent housing for the low-income population. The support provided by the World Bank, coupled with political will on the part of state and municipal authorities, shows that an integrated planning vision such as that adopted by Prometrópole can change reality for poor areas in large urban areas. 42 Annex 7: List of Supporting Documents Project Appraisal Document (Report No: 23331-BR) – April 1, 2003 Loan Agreement (Loan Number 4690-BR) – July 23, 2003 Olinda Project Agreement (Loan Number 4690-BR) – July 23, 2003 Recife Project Agreement (Loan Number 4690-BR) – July 23, 2003 Restructuring Paper (Report No: 55083v1-BR) – June 10, 2010 First Amendment to the Loan Agreement – January 23, 2004 Second Amendment to the Loan Agreement – October 13, 2006 Third Amendment to the Loan Agreement – March 6, 2009 Fourth Amendment to the Loan Agreement – June 23, 2010 Aide Memoire – July 26-28, 2010 MTR Aide Memoire – April 25-28, 2006 Aide Memoire – February 22-25, 2011 Aide Memoire – June 27-28, 2011 Aide Memoire – November 21-22, 2011 Resettlement Mission BTOR – May, 2011 Aide Memoire – April 25-27, 2012 BTOR – June 27-28, 2012 Implementation Status Report 1 – June 2003 Implementation Status Report 2 – December 2003 Implementation Status Report 3 – June 2004 Implementation Status Report 4 – December 2004 Implementation Status Report 5 – May 2005 Implementation Status Report 6 – January 2006 Implementation Status Report 7 – May 2006 Implementation Status Report 8 – December 2006 Implementation Status Report 9 – March 2007 Implementation Status Report 10 – June 2007 Implementation Status Report 11 – December 2007 Implementation Status Report 12 – June 2008 Implementation Status Report 13 – December 2008 Implementation Status Report 14 – June 2009 Implementation Status Report 15 – November 2009 Implementation Status Report 16 – May 2010 Implementation Status Report 17 – September, 2010 Project Progress Report (3rd Quarter 2010 / October 1, 2010) Project Progress Report (4th Quarter 2010 / January 3, 2011) Project Progress Report (1st Quarter 2011 / March 3, 2011) Project Progress Report (2nd Quarter 2011 / July 1, 2011) Project Progress Report (3rd Quarter 2011 / October 1, 2011) Project Progress Report (4th Quarter 2011 / January 2, 2012) Project Progress Report (June 30, 2012) Borrower’s Project Assessment Report, March 2012 (Datamétrica) 43 Annex 8: Project’s Revised Components Part A: Beberibe River Basin Investments Carrying out of infrastructure and environmental projects for the overall improvement of the built and natural environment of the Beberibe River Basin, consisting of construction or rehabilitation of roads and accesses, including improvements in the general flow of traffic. Part B: Water Distribution 1. Installation of pipes and carrying out of the associated works to improve water supply and distribution within the UEs. 2. Acquisition of micro meters for household installation in the UEs. 3. Installation of macro meters and the carrying out of associated works required to isolate the water distribution network within the UEs. Part C: Low-Income Investments 1. Carrying out of urbanization projects in groups of low-income settlements located in the UEs 03 (R), 20 (R), 21 (R) and 23 (R), which are in Recife, and 07 (O) and 17 (O), which are in Olinda, consisting of water supply and sanitation improvements and household connections, secondary road improvements, micro drainage works, installation of public lighting, creation or rehabilitation of public parks and open spaces, resettlement of population from risk prone areas, and other related investments. 2. Carrying out of environmental projects for the overall improvement of the built and natural environment of the Beberibe River Basin, consisting of: (i) creation or rehabilitation of parks, including the ecological park along the margins of the Beberibe River; (ii) pocket parks within each of the UEs; and (iii) the Matadouro de Peixinhos complex of historic buildings. Part D: Complementary Urbanization Activities Carrying out of activities complementary to the activities under the preceding Parts of the Project, consisting of: (a) engineering design services for the Subprojects and for UEs 03, 04, 08, 17, 19, 20, 21 and 23 in Recife, and 07, 11, 12, 13, 15 and 17 in Olinda; (b) development of social outreach and community participation activities; (c) provision of environmental education to the communities; (d) feasibility studies for a sites and services project; (e) studies for expanding to other areas in the RMR the investments made under the Project; (f) Project monitoring and evaluation activities; and (g) formulation of the EMP. Part E: Project Management Provision of technical assistance to CONDEPE/FIDEM and the Municipalities in the management, implementation and coordination of the Project and the acquisition of goods associated with Project management. 44 Annex 9: Project’s Results Framework Latest Values Outcome Indicators Baseline Target Values (December 2011 - June Additional Information 2012) Increase of US$2,963 per Increase of US$ 7,349.92 Data are available for property in slum per property in slum values of both land and upgrading areas and upgrading areas and Increase in property houses, in both titled and increase of US$3,774 per increase of US$ MET values by end of Project untitled areas, from a property in resettlement 24,277.64 per property in Cities Alliance funded areas. resettlement areas. study. 1.19 (UEs 20-21) and Avg# of floors: 1.18 (UE - 1.08 (UE 17) 20-21), 1.09 (UE 17) Walls not made of Measure of the quality of 1% (UE 20-21), 0% (UE concrete: 9% (UE 20-21), - MET housing pre and post- 17) 4.5% (UE 17) Results of the Borrower’s project (a measure of Final Assessment Report private investment) Floors not covered with show an overall Data for UEs 20-21 and 6% (UE20-21), 5% (UE concrete or tiles: 8%(UE - improvement of housing 17 17) 20-21), 15% (UE 17) quality post-Project. Ext. walls w/o plaster: 20% (UE 20-21), 13% 20% (UE 20-21), 30% - (UE 17) (UE 17) Positive evaluation on: Pavement quality: 31% 59% (UE 20-21), 15% - (UE 17), 31% (UE 20- (UE 17) 21) Local access routes: 31% 32% (UE 17), 55% (UE MET - (UE 17), 47% (UE 23) 23) Results of the Borrower’s Measure of the Final Assessment Report perceptions of the quality Public lighting: 59% (UE show an overall of life of project - 66% (UE 20-21) 20-21) improvement of beneficiaries perception of quality of Water supply: 39% (UE 90% (UE 20-21), life by Project - 20-21), 65% (UE 23) 93%(UE 23) beneficiaries 16. Sewer system: 17% (UE 58% (UE20-21), 45% - 20-21), 17% (UE 23) (UE 23) 16 Although several pieces of this indicator were not met, there was an overall improvement (on average, each piece had a positive 7.7% increase) in the perception of quality of life. In addition, the results framework does not capture the full extent to which quality of life improved; as discussed in the main text, the Datametrica survey offered a far more comprehensive analysis of this indicator. For instance, the Results Framework does not include results from all the UEs where housing improvements took place. Also, most of the pieces of the indicator that worsened (such as trash collection, public security, educational system) were not directly tied to the Project’s interventions. 45 Garbage collection: 39% 66% (UE 17), 80% (UE (UE 17), 85% (UE 20- NA 20-21), 74% (UE 23) 21), 70% (UE 23) Public safety: 14% (UE 12% (UE17); 25% (UE 17), 4% (UE 20-21), 5% NA 20-21), 17% (UE 23) (UE 23) Public transportation: 39% (UE 20-21), 24% 36% (UE 20-21), 53% NA (UE 23) (UE 23) Leisure areas: 3% (UE 17% (UE 20-21), 12% NA 20-21), 7% (UE 23) (UE 23) Educational system: 14% 35% (UE 17); 50% (UE (UE 17); 27% (UE 20- NA 20-21); 42% (UE 23) 21); 53% (UE 23) Public health system: 43% (UE 20-21), 21% 20% (UE 20-21), 34% NA (UE 23) (UE 23) The sum of US$23.91 was chosen as the The marginal increase in average water bill, based the receipts of the water on the findings of the distribution company census survey. 7,202 (COMPESA) was used to connections were estimate the economic Measure of installed in the benefits of the program. improvements in cost - - geographic area served recovery of OM&R of by Prometrópole. With *COMPESA was investments an average payment rate originally supposed to of 84 percent, the annual supply this information. economic benefit from The data from water distribution was Datametrica was used as estimated at a substitute. US$1,735,452. State: 47.40% (plus .60% Bank fees) Program decentralization Spending by the end of structure (measure of Olinda: 18.90% Project: 43% executed by decentralization of - MET 17 State, 57% executed by control to local levels and Recife: 33.10% municipalities community) Overall spending by municipalities: 52% 17 Although the Project did not exactly meet the decentralization indicator, the numbers came very close (within 5% of each percentage). The ICR team considers the indicator met given that the percentages were meant to serve as guidelines; it would have been very difficult and beyond the objective of the Project to meet these exactly. 46 Latest Values Output Indicators Baseline Target Values (December 2011 - June Additional Information 2012) Water supply systems: Total # contracted m 107,878.60, # executed m 103,764.30 (96%). SUBSTANTIALLY Number of new/legalized 100% of works Household water MET water supply and sewer - contracted have been connections: Total # Remaining works are facilities and connections executed contracted m 8,528, # being carried out. executed 7,434.00 (87%); Reservoirs: Total # contracted 1, executed 1 (100%). SUBSTANTIALLY Number of kms of 100% of works Contracted m 11,292.1, MET drainage constructed or - contracted have been executed m 10,335.1 Remaining works are rehabilitated executed (92%). being carried out. Contracted 940m, Increase in length of 100% of works executed 0 (0%). Will be thoroughfares with public - contracted have been executed by CELPE: MET lighting coverage executed 8,286 m contracted and executed (100%) SUBSTANTIALLY Number of families MET resettled due to Contracted 1,674, 100% of works Remaining housing geotechnical risk and/or executed 1,385 (83%). - contracted have been development works are encroachment into The remaining 27% are executed being carried out. There environmentally sensitive in execution. are 289 families still to areas be resettled. MET Number of community- The difference between based groups integrated 5 groups (Local target value and achieved 7 groups (local into the slum upgrading - Development Councils) value refers to the fact development councils) process in the Beberibe with 152 members. that the # of UEs was River Basin reduced in Project restructuring Number of kms of SUBSTANTIALLY 100% of works Contracted m 37,813.50, pedestrian paths and MET - contracted have been executed 37,039.50 communal space Remaining works are executed (98%). improved and/or built being carried out. Number of kms of new SUBSTANTIALLY 100% of works Contracted m 27,875.10, and/or rehabilitated MET - contracted have been executed 26,887.60 primary and secondary Remaining works are executed (96%) access roads being carried out. 47 Increase in access to recreational activities SUBSTANTIALLY (pocket parks), including 100% of works MET rehabilitation of - contracted have been Contracted 6, executed 4 Remaining works are historical infrastructure executed being carried out. “Matadouros de Peixinhos� SUBSTANTIALLY Number of new property MET - 1,674 1,385 titles issued Remaining titling is being carried out. Direct employment of (In all the works): 100% of works with community member is Contracts at the period: - community members MET execution of Project 46; Accumulated employed works contracts: 2,120. Supra-local Program Cost structure, Supra-local infrastructure: 12.5%; defined as the proportion infrastructure: 12%; Local infrastructure in of total project costs Local infrastructure in poor areas: 41.5%; spent on overhead, poor areas: 46%; Housing units: 24.6%; surveying and planning, - Housing units: 21%; MET 18 Social outreach: 4.3%; infrastructure Social outreach: 3%; Engineering designs: investments, titling, Engineering designs: 5%; 4.6%; Project social services, Project management: management: 11.8%Bank resettlement etc. 13%. fees: 0.6% 18 Although the Project did not exactly meet project cost indicators, the numbers came very close. The ICR team considers this indicator met given that the percentages were meant to serve as guidelines; it would have been very difficult and beyond the objective of the Project to meet these exactly. 48 Annex 10: Project Intervention Map Survey Areas, Urban Upgrading Intervention Areas (Part C), Direct beneficiaries count PROMETROPOLE Original Intervention Areas, Indirect beneficiaries count Water Distribution / Sectorization Intervention Areas (Part B), COMPESA/Prometropole, Indirect beneficiaries count* Ecological Reserve Cultural Center Matadouro de Peixinhos Municipal Limit Water Beberibe River Basin UE Limit * Only the beneficiaries from Part B within the original intervention area were counted as indirect beneficiaries. 49 To Ciudad Guayana 70°W 60°W 50°W 40°W R.B. DE GUYANA VENEZUELA SURINAME French Guiana BRAZIL r i n oc (Fr.) AT LA NT IC O o COLOMBIA Boa Vista O CEA N AMAP� STATE CAPITALS RORAIMA NATIONAL CAPITAL Macapá 0° 0° RIVERS Ne zon gro Ama MAIN ROADS A m a z o n Belém São Luís Manaus RAILROADS Am az o jó s n Fortaleza STATE BOUNDARIES pa a Ta eir ad B a s i n Teresina RIO GRANDE M PA R � MARANHÃO INTERNATIONAL BOUNDARIES CEAR� DO NORTE AMAZONAS Natal u s Pur u Xing PARA�BA João Tele PIAU� Pessoa s Porto Velho PERNA Tocantins Pir es MBUC O Recife ACRE Maceió ia Rio Branco o Palma sc Aragua 10°S RONDÔNIA Juruena ALAGOAS ci an Aracaju São Fr TOCANTINS BAHIA MATO SERGIPE PERU GROSSO Mato Grosso B r a z i l i a n Salvador Plateau Cuiabá F.D. BRAS�LIA BOLIVIA GOI�S Goiânia MINAS GERAIS PAC IF IC To H i g h l a n d s Santa Cruz MATO GROSSO Belo ESP�RITO OC E AN CHILE DO SUL Grande Horizonte SANTO Paragu y 20°S Campo 20°S Grande Vitória To a Santa Cruz SÃO PAULO á an RIO DE r Pa PARAGUAY São Paulo Rio de JANEIRO Janeiro PARAN� AT L AN T I C Curitiba BRAZIL STA CATARINA OCE AN Florianópolis RIO GRANDE DO SUL Porto Alegre 0 200 400 600 Kilometers 30°S 30°S ARGEN TINA ARGENTINA SEPTEMBER 2009 IBRD 33377R This map was produced by the Map Design Unit of The World Bank. To Buenos The boundaries, colors, denominations and any other information Aires URUGUAY 0 200 400 Miles shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. 70°W 60°W 50°W 40°W To Montevideo