Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7092 PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES MANILA URBAN DEVELOPMENT PROJECT (LOANS 1272-PH/1282-PH) SECOND URBAN DEVELOPMENT PROJECT (LOAN 1647-PH) January 19, 1988 Operations Evaluation Department This document has a resfticted distribution and may be used by recipiens only in the perfonnance of their official duties. Its contents may not otherwise be disclosed without World Bank authoriztion. tg~~ woinw -F GaCtIAiL 1W OPNLY tHF WORD &A Wasbinton. D.C. 20433 USA. o. e d Oi.inew January 19, 1988 MEMORANDUM TO THE EXECUTIVE DIRECf'ORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on the Philippines - Hanila Urban Development Project (Loans 1272 PH/1282-PH) and Second Urban Development Project (Loan 1647-PH) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on the Philippines - Manila Urban Development Project (Loans 1272-PH/1282-PH) and Second Urban Development Project (Loan 1647-PH)" prepared by the Operations Evaluation Department. Attachment This document tm a mkmetd diuibuio and may be ON by _mas _a In iM I . ftcisW dut_s It content may amot wis be dbi wtot Woo t a n ABBRIVIATIONS AND ACRONMS USED APD Areas for Priority Development BDC Bliss Development Corporation CI Coomarcial/lndustrial Lots COA Comission on Audit DBP Development Bank of the Philippine. DR Department of Health DPH Department of Public Highways DPWTC Department of Public Works, Transportation and Coaunications GOP Government of the Philippines GSIS Government Insurance System HSDC Human Settlements Development Corporation HSRC Human Settlements Regulatory Coission ICB International Comptitive Bidding LCB Local Competitive Bidding MASICAP Medium and Small Industries Coordinated Action Program MERALCO Manila Electrical Authority MHS Ministry of Human Settlements MINFIN Ministry of Finmnce MMA Manila Metropolitan Area HMC Metropolitan Manila Commission MPH Ministry of Public Highways MPW Ministry of Public Works MPWH Ministry of Public Works and Highways MPWTC Ministry of Public Works, Transportation and Comunications MSSD Ministry of Social Services Development MWSS Metropolitan Waterworks and Sewerae System NCSO National Census and Statistical Office NEDA National Economic Development Authority NHA National Housing Authority NHC National Housing Corporation NHMFC National Home Mortgage Finance Corporation NNC National Nutrition Council OED Operations Evaluation Department PCR Project Completion Report PNB Philippine National Bank PPAH Project Performance Audit Memorandum PPAR Project Performance Audit Report RCDP Regional Cities Development Project SAR Staff Appraisal Report SBL Small Business Loan SIR Slum Improvement and Resettlement Program SSS Social Security System TFDDO Tondo Foresbore Dagat-Dagatan Development Office UNICEF United Nations Childron's Emergoncy Fund ZIP Zonal Improvement Program ,COUNTRY ECHNGE RATI Nm of Currency (Abbreviation) Philippine Peso (P) Yoar Exchange Rat. la First Project Apprais4l Year lb (1975) US$1 - 8.8000 Intervening Year (1976) US$1 - 8.5899 Second Project Appraisal Year Le (1977) US$1 - 8.6429 InterveninS Years (1978) US$1 - 9.2219 (1979) Us$1 - 7.3780 (1980) US$1 - 7.5110 (1981) US$1 7 7.9000 (1982) US$1 - 8.5400 (1983) US$1 - 11.1130 (1984) US$1 - 16.6990 Completion Toar of Second Project (1985) US$1 - 18.6070 Completion Year of First Project (1986) US$1 - 20.2650 Ld Sources DIN, International Financial Statistics Yearbook, Washington, D.C., 1986 la Period Averages, at market rates. /b The SAR used the official exchange rate of US$1 - 7.50 instead of the market rate shown above. L_ The SAR used tho official exchbnge rate of US$1 - 7.40 instead of the market rate shown above. /d July, 1986. WEIGHTS AND HEASURES 1 meter (a) - 39.37 inches (in) 1 square meter (sq a) 10.80 square feet (sq ft) 1 cubic meter (cu m) - 35.3 cubic feet (cu ft) 1 kilometer (km) - 0.62 mile (ml) 1 square kilometer (sq km) - 0.386 square mile (sq ml) 1 centimeter (cm) 0.39 inch (in) 1 square centimeter (sq cm) - 0.155 square Inch (sq in) 1 hectare (ha) - 10,000 square meters (sq m) or 2.47 acres (ac) 1 litre (1) - 1.057 quarts liquid or 0.26 US gallon (gal) or 0.908 quart dry (qt) 1 liter per capita per day (lpcd)- 0.26 US gallon per capita per day (ged) PROJECT PERFORMANCE AUDIT REPORT PHILIPPINS MANILA URBAN DEVELOPMENT PROJECT (LOANS 1272-PH/1282-PH) SECOND URBAN DEVELOPMENT PROJECT (LOAN 1647-PH) TABLE OF CONTENTS Pate No. Preface ......e ................................................... 1 Basic Date Sheet ............. ..................................... ii Evaluation Summary .......................................................... ir PROJECT PERFORMANCE AUDIT NDEORANDUN I. Background .......... ... .... ........... 1 II. Project Objectives and Design ........................... 7 III. Project Implementation .................................. 13 IV. Project Results ........ ...... .................... ....... 22 Figure 1 - Disbursement Profile ..... ............................ 21 Attachments A and B - Comments from the Borrower .................. 34 EAST ASIA & PACFIC REGIONAL OFFICE PCR .............. .............. 37 Tables I/II - rV and Chart V * ................................... 44 SECOND URBAN DEVELOPMENT PROJECT COMPLETION REPORT (Ln. 1647-PH) .. 48 Basic Data Sheet *...e......*@.... ..ee .............. 49 Highlights ................eeeeeeeee. .................e...e...... 51 I. CONDITIONS LETDING TO THE PROJECT ...............o........ 54 II. THE PROJECT *e.eee....e .........e. ..e.ee.ee...e..e... 55 III. MAJOR RESULTS *......* ................................... 57 rV. PROJECT IMPLEHENTATION AND COST ......................... 61 V. FINANCIAL PERFORMANCE *........ ... ......... ... *. 68 VI. INSTITUTIONAL PERFORIANCE .................. .............. 81 VII. ECONOMIC RE-EVALUATION ............... . ............. 89 VIII. CONCLUSIONS ....................... . .............c. 91 ANNEXES - I - rV 95o...........e..ee..e......o ........eee 95 Table of Contents (Contined) MANILA URBM DZVK DPHE PROJECT CCNPLtTION REPORT (Lns. 1272T/1282-1-PR) Peaz No. I* INTRODUCTION ........................................... 103 II. PROJECT IDENTIFICATION, MREPARATION AND APPRAISAL ...... 104 III. MAJOR RELTS .................., 109 Iv. PROJWCT DIFLZ1ZITATION AND COST ...... ....... ........... 116 V. FINANCIAL ASPECTS .................................... . 122 VI. INSTITUTIONAL P?RFORHANCE AND DEEYLOPMENT .............. 132 VII. ECOPOKIC R-EVALUATION ....... . 141 VIII. CO*CLUSIONS ............................................ 142 ANNEXES A. Principal Report Sources 148 B. Comparison of Appraisal and Current Project Costs 150 C Schedule of Disbursements . ...................... 153 D. CoWpliance vith Loan Covenants .......154 E. Traffic Mnagement Components 155 F. Comput-rized Traffic Signals in Bangkok and Manila .... 157 Go Studies - 158 H. Housing Agencies in the Philippines .................... 159 I. NHA Organisation 1975-1980 ............................. 161 J. Development Cost/Monthly Charge 164 K. Cost Allocation ......... . 166 L. Turnover (Resales) Among Tondo Residents . 169 H. Borrower's Couents . . . 170 PROJECT PERPONICE AIUDIT RIPORT PRILIPPMINS MANIIA URBN DEVELOPNT PROJECT (LOANS 1272-PHI/1282-PH) SECOND URm DV OpHT PROJECT (LOAN 1647-PB) PREIACE This Project Performanc Audit Report (PPAR) represents a performance audit of two Philippie projectst the Manll Urban Development Project and the Second Urban Developmnt Project. For the, first project, loans 1282-PH and 1272-PH for $32.0 million equivalent were approved on May 27, 1976, including a $10.0 million Third Window Intermediate Term Loan (No. 1272-0-PH). A supplementary loan (1282-1-PH) for $10.5 million equivalent was approved in February 1984 under the Bank's Program of Special Assistance to Member Countries 1983-84. Bank Group financing for the second project was provided through Loan 1647-PH for $26.7 million equivalent, approved on December 21, 1978. The final loan disbursement on the first project was made in July 1986 and the undisbursed balance of $3.2 million equivalent was cancelled. For the second project, the final disbursement was made July 24, 1985 and the undisbursed balance of $1.4 million equivalent was cancelled. The PPAR consists of the following: (a) a Project Completion Report (PCR) on the first project, dated April 1986, prepared by the East Asia and Pacific Regional Office; (b) a PCR on the second project, dated December 1986, prepared by the Borrower and supplemented by the East Asia and Pacific Region memorandum of December 31, 1986; and (c) a Project Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OED). The PPAM is based on the PCRs, Appraisal and President's Reports, discussions with Bank staff directly involvod with the projects, a review of the projects' files and records, and the transcripts of the Executive Directors' meetings which considered the project. The Audit finds that, in most respects, the PCRs accurately describe the projects' background and experience and provide adequate factual support for the conclusions reached. The PPAM supplements the discussion of that experience with a review of the economic and sector contexts as they affect project outcomes. Otherwise, the PPM largely sumnarizes and attempts not to repeat the details provided in the PCRs. Copies of the draft PPAR were sent to the Borrower for their com- ments. These comments have been reproduced as Attachments A and B to the report. PROJECT PEIPMANCI AUDIT aSIC KU S T PIIILIPPINEs MANILA URIDN DEVELOPMNT PROJgCT LOA_S 1272-P!!1282-P!) 9Y PROJECT DATA Appraisal Actual or Expectation Curront Estisate Total Proj.-t Cost (US$ million) 65.0 91.9 la Cost Overrun (2) - 41.4 Loan Amount (US$ million) 32.0 32.0 Supplementary Loan 10.5 10.5 Disbursed 32.0 39.3 Cancelled - 3.2 Outstanding - 39.3 Date Physical Components Completed 12179 07/86 lb Proportion Completed by Appraisal Completion Date (Z) 100.0 38.0 Proportion of Time Overrun (2) - 165.0 (88.0) Ie Economic Rate of Return (Z) 27.0 Id 18.0 le Residential Development 21.0 15.0 Industrial/Commercial Development 40.0 25.0 Transport Development: Road R-10 link C-1/C-2 23.0 23.0 le Road R-10 link C-2/Dagat-Dagatan 24.0 24.0 le Road R-10 link C-2 14.0 14.0 le Traffic Development If 29.0 Financial Performance Satisfactory Poor Institutional Performance Satisfactory Mixed Physical/Social Impact Satisfactory Satisfactory la Latest available estimate of December 1984, prior to completion of the major roads component. /b Date of last disbursement used due to lack of completion date for the major roads component. All other project components were completed by mid-1983. /c The lower figure refers to completion by mid-1983 of all components, except major roads. Id Weighted average (SAR, para. 6.15). le Weighted average, estimated prior to completion of major roads component. /f Not estimated at appraisal. Cuulative Igtimat*d and AJctul VIgburgements (US$ sillion) nY77 n18F$ FV? FY0 FY81 FY82 F FY84 FY8S EstS.at.d 8.3 18.2 26.6 31.3 32.0 32.0 32.0 42.5 52.5 Actual 0.3 5.1 9.3 13.7 17.1 23.1 24.5 31.5 38.8 Estimtedl Actual (2) 3.6 28.0 35.0 43.8 53.4 72.2 76.6 74.1 81.3 rni FY87 EstSaat.d 42.5 42.5 Actual 38.8 39.3 Es tiatedI Actual (Z) 91.3 92.4 Date of Last Disbursement July 17, 1986 OTHER PROJECT DATA Item Original Actual or Reestimated First mention In files 08172 Govornment Application 05173 Appraisal 04/75 08/75 Nogotiations 12175 04176 Board Approval 05176 05/76 Loan Agreement 06/76 06/76 Effectiveness 08/76 12/76 Closing Date 09181 07/86 Borrowers Government of the Republic of the Philippines Executing Agencies: National Housing Authority (NHfA) Metropolitan Watewrks and Sewerage System (MWSS) Ministry of Public Works and Highways (NPWH) FY of the Borrowers January 1 to December 31 (as of 1977) Follow-on Project: Second Urban Development (Loan 1647-PH) Amounts US$32.0 million equivalent Loan Agreements 01/79 K ~~~~~~~~~~~~iv (Staff-woks) nn rF74 FY75 nY76 FY77 FY78 FY79 =FY n ln Preparation .9 32.0 91.4 6.2 Appraisal 143.4 Negotiations 9.8 Supervision 4.4 103.3 51.9 25.9 22.0 21.8 7.8 n3n84 nY86 Total Preparation 130.5 Appraisal 143.4 Negotiations 9.8 Supervision 11.8 15.4 16.6 2.3 _283.2 Total 566.9 FIELD MISSION DATA Item Month/ No. of No. of Staff Report Year Weeks Persons Weeks ^tle Identification 08173 3.0 5 15.0 08173 Preparation 11173 2.5 7 17.5 - Preparation 03/74 1.0 3 3.0 05174 Preparation 05174 1.5 4 6.0 05174 Preparation 05174 3.0 3 9.0 06/74 Preparation 09/74 3.0 5 15.0 12/74 Preparation 02/74 2.0 8 16.0 03175 Pre-appraisal 06/75 6.0 4 24.0 - Appraisal 08175 3.5 9 31.5 09/75 La Post Appraisal 11/75 2.0 6 12.0 12/75 Lb Supervision I 07/76 2.0 5 10.0 10/76 Supervision II 11/76 2.0 5 10.0 01/77 Supervision III 03/77 2.0 4 8.0 05/77 Supervislon IV 08/77 2.5 5 12.5 09/77 Supervision V le 11/77 4.0 3 12.0 12/77 Supervision VI Id 03/79 4.0 6 24.0 04/79 Supervision VII Id 09/79 2.0 4 8.0 10/79 Supervision VIII 14 07/80 3.0 5 15.0 08/80 Supervision IX Id 11/80 3.0 3 9.0 12180 Supervision X /d 07181 1.5 3 4.5 09/81 Supervilsion XI l- 10/82 2.0 5 10.0 11/82 Supervision XII /e 09/83 2.0 4 8.0 10/83 v ?I3W MII81ON bTA (cont'd.) Sup.rvlelon XIII La 03184 2.0 4 8.0 03184 SupervisLon XIV Li 11184 2.0 4 8.0 12184 Supervoi1@ XV ,& 03185 2.0 3 6.0 06185 Sup.rvision XV! Li 12185 2.0 2 4.0 12185 I& Issue Papor. lb Updated Decision Mmorandum. Ic Joint supervision sad identific.tion mission. 1d Joint supervision with Loan 1647-PS. Ia Joint supervision with Loans 1647-PH and 1821-PH. DAMJECT PoE mANcE AUDIT WAIC DATA MET SECOND 1135N DEVEWPMHNT PlOJECT (Loga 1647-PR) 1E OJACT DATA Appraial Actual or Ixyota.Ion Current Estimate Total Projcet Cost (US$ million) 69.9 67.7 Cost (Underrun) (2) - (3.1) Loan Amunt (US$ million) 32.0 32.0 Disbursed 32.0 30.5 Cancelled - 1.5 RepLid La Outstanding - 1La Date Physical Components Completed 12/82 07/85 Proportion Completed by Appraisal Complotion Date (2) 100.0 65.0 Proportion of Tim Overrun (2) - 65.0 Economic Rate of Return (Z) 34.0 19.0 /b Financial Performance Satisfactory Poor Institutional Performance Satisfactory Mixed Phys cal/Social Impact Satisfactory Satisfactory /a The revised loan amortisation schedul- established payments to be made twice a year in the amount of US$1.0 million equivalent, begin- ning 09/15/86 and ending on 03115/99 but curreat repsyment figures were not available to the Audit. lb Estimated by NUA and considered optimistic by the Audit, because the market values used by NBA for comercial and industrial lots are probably too high. The Audit was unable to roestimate th ERR due to lack of adequate data. Cumulative Estimated and Actual Disbursements (US$ million) PY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 Appraisal 1.5 5.7 18.8 29.3 32.0 - - - Actual 0.0 2.6 6.5 12.6 18.2 25.7 30.2 30.5 Actual as 2 of Appraisal 0.0 45.6 34.6 43.0 56.9 80.3 94.4 95.3 Date of Last Disbursements July 24, 1985 vii OTHER PROJECT DATA Actual or Item Original R.estimted First mention in files 03177 Appraisal 09177 10178 Negotiations 11178 11/78 Board Approval 12/78 12178 Loan Agreement 01179 01179 Effectiveness 03179 04179 Closing Date 06184 06185 Borrowert Government of the Republic of the Phillppine Executing Agencies: National Housing Authority (NHA) Ministry of Public Works, Trnsportation and Coimnmnications (MPWTC) National Economic Development Authority (NEDA) Ministry of Human Settlements (MB8) Metropolitan Manila Water and Sewerage Systems (MWSS) FY of the Borrowers January 1 to December 31 Follow-on Projects Third Urban Project (Loan 1821-PH) Amount: US$72.0 million equivalent Loan Agreement: 06180 Staff Input (staff-weeks) FY77 FY78 FY79 FY80 FYn8 FY82 FY83 FY84 Preparation 19.3 34.5 Appraisal 122.7 1.1 Negotiations 28 6 Supervision - 23 9 19.4 258 81 13.8 16.8 Total 19.3 157.2 53.6 19.4 25.8 8.1 13.8 16.8 FY85 FY86 Total Preparation 53.8 Appraisal 123.8 Negotiations 28.6 Supervision 15.9 10.6 134.3 Total 15.9 10.6 340.5 FIELD MISSION DATA Month/ No. of No. of Staff Report Item Year weIM. Persons weMk D t- Identification 03/77 2.6 6 15.6 03177 Preparation 07177 2.0 5 10.0 08177 Appraisal 10/77 4.0 6 24.0 11177 Supervision I La 02/79 4.4 6 26.6 04179 Supervision II la 08/79 2.6 4 10.4 10179 Supervision III La 06/80 3.0 5 15.0 08/80 Supervision IV La 11/80 2.9 3 8.7 12180 Supervision V I* 02181 0.9 1 0.9 03/81 Supervision VI jj 06/81 1.3 3 3.9 08181 Supervision VII lb 11/81 3.7 2 7.4 12/81 Supervision VIII lb 09182 2.0 4 8.0 11/82 Supervision IX /b 08183 2.0 4 8.0 10183 Supervision X Lb 02184 2.1 4 8.4 03/84 Supervision XI /b 10/84 2.1 4 8.4 12/84 Supervision XII Lb 04/85 2.0 3 6.0 06/85 Supervision XIII jb 11/85 3.6 2 7.2 la Joint Supervision with Loan 1282-PH. Lb Joint Supervision with Loans 1282-PH and 1821-PH. ix PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES MANILA URBAN DEVELOPMENT PROJECT (LOANS 1282-PHI 1272-PH) SECOND URBAN DEVELOPMENT PROJECT (LOAN 1647-PH) EVALUATION SMMIARY Introduction The first and second urban development projects financed by the Bank in the Philippines followed each other in rapid succession. Loans 1282-PH and 1272-PH, approved in May 1976 for the first project, also financed the prepar- ation of the second. Before the first project was 25 percent complete, in December 1978 Loan 1647-PH for the second project was approved. This circum- stance made lt dlfflcult for the Borrower and the Bank to use the lessons learned from the first project in the design of the second, and consequently, both projects share many of the am* implementation dLfficultLes. This is in part the reason why the two projects are being audited togethor. Objectives Both projects aimed at establishing cost-effective models for the provision of housing services to lower-income urban families through in situ upgrading of slum areas and the development of new serviced sltes for housing construction. The first project focused exclusively on Manila, while the second combined the investments in the capital with the expansion of the shelter program to three regional cities (Cacayan do Oro, Cebu and Davao). Through technical assistance the projects were expected to strengthen the recently created National Housing Authority (NH) -- an autonomous public corporstion dedicated to the production of housing and the principal implemen- tation agency for both projects. In the execution of the projects, NRA was expected to coordinate the activities of several other particLpE Ing agencies, including the Metropolltan Waterworks and Sewerage System (kwiS) and the Manila Electrical Authority (HERALCO) for water supply and power in Manila, and the Ministry of Public Works, Transport, and Communications (MPWTC) for toads, drainage, and community facilities both in Manila and in regional cities. In addition, under the second project, NUA was to function as a mnuiLipal lending bank to local governments, financng project Implementation in regional cities. Shelter objectLves were combined with employment promo- tion through the development of commercial and industrial lots at the housing sites and a line of credit to assist small businesses. This line of credit was to be administered by the Philipplne National Bank (PNB) with assistance from other Government-owned banks and agencies under NHA's general coordina- tion. Other main project objectives comprised supporting the MPWTC program of I traffic improvement and urban transport infrastructur expansion in Manila, under the first project, and its flood control program, also in Manila, under th, second project (PPAM, par&s. 18, 32 and 33). Imolem_ntation Experience Both projects were largo in terms of the scope of the physical works involved, complex in their institutional arrangements, and novel in so far as sites and services and upgrading components were being Introduced for the first tim In the Philippines under the auspices of a new agency. The combi- nation of these design features did not facilitate implementation nor did the process of project preparation. Project preparation was deficient in the case of the first project due to the inexperience of NHA and the Bank In the kinds of work proposed. It wa also deficient in the second project, as design wa completed before clear lessons could be drawn from the first project. Both projeets took significantly more time to complete then expected at appraisal. The completion of the first project wa delayed by approxinately six years and that of the second by two years. Delays were caused by slow start-up mainly due to 'land acquisition problems, recurrent technical and administrative difficulties, and increasing financial problems associated with the downturn of the Philippine economy as of the oarly 1980. (PPAM, paras. 44-53). The first project was significantly modified during implementation, partly to overcome design shortcomings, but also in response to political pressures of various sorts (PPAM, paras. 54-58). A more than double increase in the scope of work of the major roads components made necessary the approval of a supplemental Bank loan in February 1984, for the equivalent of $10.5 million, to permit completion of that component. Cost overruns for the proj- ect as a whole ascended to approximately forty percent. Modification made tn the second project during implementation consisted In reductions in the scope of certain components in order to offset cost overruns In others. These adjustments, together with the rapid devaluation of the Philippine peso, kept project costs within the original budget, without unduly distorting its basic design (PPAM, paras. 59-64). Remedial action afforded by the Bank's 1983-84 Special Assistance Program (SAP) has proven essential for the completion of the projects (PPAM, para. 53). Bank loan disbursements on both projects accelerated after FY83 due to the positive Impact of the SAP but, in general, lagged considerably behind the over ambitious appraisal estimates (PPAM, par"s. 65-67). Results The Audit would have profited from a field inspection of the proj- ects and the lack thereof has caused certain uncertainties as to the project results to remain unclarified. In general, the physical results of both projects were satisfactory. In shelter, the first project succeeded in upgrading the Tondo Foreshore Area comprising approximately 15,000 housing structures, and in developing an additional 2,000 serviced lots in adjacent Dagat-Dagatan to relocate families dislodged as the Tondo underwent rehabilitation. The second project extended xi the program to three regional cities, where so"e 7,500 housing structure were upgraded and approximately 2,700 serviced sites were provided, and also expanded the Dagat-Dagatan development by an additional 8,300 residential lots. Urban transport investments under the first project increased the road carrying capacity of Manila by over 20 km and introduced several improvements in the traffic system. Modest but also satisfactory outeomes resulted from the supply of equipment *nd technical assistance in support of the Manila Flood Control Program, although the solution for the flooding problems of the city would require substantial additional investments (PPM, paras. 68-75 and 79-86). Notwithstanding the satisfactory physical accomplishments of the projects, the longevity of the infrastructure "ssots built is still dependent upon maintenance arrangements which have not yet been completed. The turnover of assets!to the appropriate line agencies responsible for operations and maintenance has not been clearly defined in the project documents and had ultimately to be specified by a Cabinet and Presidential directive signed in December 1986. Though the matter is In principle resolved, the directive has yet to be implemented (PPM, para. 103). Furthermore, the indirect environ- mental impact of the second project upon the built-up communities surrounding the Dagat-Dagatan housing development has been adverse, since site in-filling contributed to worsen flooding problems in the lower-lying areas (PPAM, para. 83). The two projects were less successful in several important aspects. In the area of urban employment, project performance was disappointing because the expected demand for commercial and industrial lots and for small business credit did not materialize. While this outcome resulted primarily from the downturn of the Philippine economy, the Investment risks could have been minimized if demand studies and marketing plans were completed during project preparation (PPAM, paras. 76-77 and 87-88). Secondly, both projects were unable to meet financial appraisal objectives on account of: (a) the introduc- tion of subsidies in the prices of residential lots at higher levels than deemed appropriate at appraisal; (b) the erosion of the potential profits from the sale of commercial and Industrial lots by lack of demand, thus invalidat- ing the cross-subsidy assumption made at appraisall and (c) the general short- coming in collections of project accounts receivable by NHA and local govern- ments. The combined impact of these factors resulted in extremely poor cost recovery performance under both projects. This outcome has unnecessarily increased the already heavy external debt burden of the Philippine Government and seriously curtailed NRA's financial sustainability (PPAM, paras. 92-98). The institutional development achievements of the projects were less than fully satisfactory because NHA was unable to attain financial stability and serious problems of inter-agency coordination porsist. Moreover, the projects assumed a narrow institutional development approach, which appears unrealistic in retrospect, since it focused primarily on the internal admini- stration of NRA and failed to take into account the policy environment within which the Authority operated. This approach isolated the projects from a major sector reorganization which took place in 1978, when the Ministry of Human Settlements (MHS) and six other shelter agencies were created, and clearly reduced the potential for establishing long-term sector programs based xit on the alternative housing development *odel- which both projects aimd to demonstkate (PPAM, paras. 99-102). Conclusions and Lessons Learned Project Investments met real needs in the shelter and urban trans- port sectors in the Philippines, fmiliarixed the Borrower with now techniques for housing development, and reduced (although not replaced) the Government's reliance on th. slum demolition and resettlement approach to the problem of landless squatters by introducing a more humane and efficient alternative. Although still facing serious problems at present, NHA evolved through the implementation of these two projects into an agency capablo of delivering significant quantities of shelter to low-income Filipinos (PPAM, pare. 99). Estimated returns on project investments approached appraisal esti- mates in urban transport but wero lower than anticipated In shelter and employment promotion. In the case of shelter, losses ensued from the deferral of benefits associated with overdrawn project implementation and from large unexpected increases in development costs. Yields on commercial and indus- trial estato Investments were reduced under the first project by approximately 60 percent compared to appraisal projections, and remain negativo under the second project. These investments were the most directly affected by the sharp retraction in business activity which followed tho national financial and political crises of the early 1980. (PPAN, paras. 72-76 and 86). Negative financial returns represent a serious constraint to the replicability of the projects. The combined impact of minimal returns and high levels of inflation have decapitalized NHA in real terms to the extent that it is presently ill-equipped to service its external loans, including those made under the projects being audited (PPM, paras. 97-98). On the basis of the implementation experience with these two proj- ects, the following lessons surface (PPAM, pares. 105-107)5 (i) as a demonstration project, the Tondo upgrading was over ambitious and overwhelmed the implementation capacity of the Borrower and of local contractors. However, an alternative smaller target area would probably not have been politically acceptable at the time; (ii) a two year interval between the Inception of the first and the second projects proved Insufficient to consolidate the operational capacity of NHA and, at the same time, expand the program on a national scale. The repetition of certain operations errors might have been avoided by postponing the second project until substantial results were achieved under the first; 1/ 1/ The East Asia and Pacific region contends that the advantages of "aiting the results of the first project were outweighed by the need to sustain institutional developments and to retain trained staff who would have been unemployed had the second project been delayed several years. sili (i1i) implemntation proble_ faced by both projects could have been 4ninizsed If the acquisition of land had been completed prior to loan effectiveness and final engneering deigns for *11 components had baen produced during the project preparation pbhe. Sfmilarly, Insufficient attention appears to havo boen given to non-clivl works elmenta of the projects during preparation and execution of both projects; (iv) loss than successful Institutional arrangements for project lnple- mntation revealed unrealistic appraisal expectations as to the Inter-agency coordination capacity of NHA and the level of financlal management and policy coherence required by tho projects at the national and local levels. Improv m nts of both financial and Institutional arrang_ments ar- presently being explored sectorvide, but satisfactory solutions have yet to be devised. At the sector level, the Audit attributes the modest policy gains achieved by the projects largely to the lack of an effective dialogue between the Borrower and the Bank on several lmportant sector issues such as bousng finance and relocation of squatter settlements. It is conceivable that by slowaing down the pace of sector lending the Bank might hve better a aisted the Borrower in Instituting needed sector policy reforms. Fortunately, how- ever, both the Region and the Borrower report the lessos learned from the experience of these first two projects are presently being applied to the restructuring of on-going operations and aro likely to improve future project performance (PCR - Urban I, paras. 8.10-8.11 and PCR - Urban II, pars. 5.28- 5.32). PROJtCT PERFORMANCS AUDIT HUDNRANDUM PHILIPPINES MANILA URBAN DEVELOPMENT PROJECT (LOANS 1272-PH/1282-PH) SECOND URIAN DEVELOPMENT PROJECT (LOAN 1647-PH) I. BACKGROUND Economc Context 1/ 1. The first and s*cond Bank-assisted urban development projects In the Philippines followed a decade of rapid economic growth in the 1960., when GNP increasod in real torms at an average rate of 5.5 poreent per annum (or approximately 2.2 percent per person per year). Unfortunately these economic gain had been unevenly distributed among regions and income clases, and considerable poverty persisted at the end of the decade. Low levels of tau- tion reulted in inadequate public expenditures for necessary Infrastructure and social services. Rapid population growth (at an average rate of 3 percent per annum between 1960 and 1970) placed an additional burden upon the avail- able supplies. In response to these needs, the aim of both projects was to increase the provision and correct the inequalities in the distribution of urban services and shelter by supporting public sector investments for the benefit of the urban poor. 2. When the first project was prepared In the early 1970-, economic growth In the Philippines accelerated to an average annual rate of 6.2 percent end the level of public investment increased significantly. Expanded agricul- tural output led to self-sufficiency in food grains and yielded exportable surpluses. The population growth rate slowed to 2.5 percent per year. Higher prices for agricultural export counodities (particularly sugar) temporarily sheltered the economy from the world recession. In general, economic condi- tions appeared promising and the time propitious for launching a major effort to expend public services in urban areas. Tho manner and form in which these services would be supplied wa expected to have a significant bearing on employment, income distribution and on reducing urban poverty. 1/ This section is based on World Bank studies of the Philippines' economy covering the 1970/85 period, especially the followings "Philippines, Priorities and Prospects for Development, Basic Economic Report, "1976 (Report No. 1095a-PH); *The Philippines, Domestic and External Resources for Development," 1979 (Report No. 2674-PH); and "Philippines, Country Economic Reports A Framework for Economic Recovery,' 1986 (Report No. 6350-PH). 2 3. Notwithstsading this positive outlook, the Philippine economy still had s*veral underlying webknessea that were already apparent in the mid-seven- ties. Industrialization was slower than dsaired, exhibited high lncremental capital-output ratios, and generated only modest employment growth. Government policy was inappropriate in many areas (Including trade, financial and exchange rate policies), lXading to Investments in sectors where the Philippines did not have a clear comparative advantage and artificially inflating the value of the peso. External borrowing and imports expanded rapidly but export receipts and domestic resource mobilization lagged. 2/ 4. With the collapse of sugar prices In late 1975 and the second sharp rise in oil priccs, the country's terms of trade dropped by 23 percent and growing import requirments were not met with commensurately higher export receipts. In consequence, the current accounts deteriorated from an average surplus of 0.7 percent of GNP In tho 1970174 period to a deficit of 4.6 per- cent of GNP in 1975180. 5. To sustain a high rate of growth under these circmstances, the Government borrowed extensively to finance the deficit and continue to stimu- late exports. Unfortunately, much of the gains expected from an expansion in export volume were offset by further deterioration in the terms of trado. As most of the deficit was financed with foreign loans, total external debt increased from US$4.9 billion at the end of 1975 to US$13.4 billion at the end of 1979, an average rate of growth of 28.6 percent per year during that period. 6. The performance of the economy deteriorated sharply in the early 1980s under rising interest rates, culminating in an open crisis situation in 1983. When a political crisis developed following Aquino's assassination in 1983, and the international banks stopped lending to the Philippines, the country's critical balance of payments position was exposed. The situation was further exacerbated by a flight of capital out of the country, and a cessation of short-term lines of credit from the commercial banks. 7. To overcome the crisis, the Philippine Government negotiated, in 1984 and 1985, a new repayment schedule for the country's 'external debt and introduced stern internal financial stabilization measures, including drastic cuts in public expendituroe and restrictive monetary policies. Stabilization measures succeeded in eliminating the current accounts deficit and in reducing domestic inflation from a peak of 50 percent in 1984 to 2 percent in the first quarter of 1986, but at the expense of a sharp drop in economic activity. During 1984185, GDP growth was negative, unemployment increased, income per capita declined to 1975 levels, capacity utilization dropped further and real 2/ While a large portion of the public sector investment helped reduce petroleum imports by developing indigenous energy sources, much of the investment financed during this period was ill-conceived, overpriced, or proven unprofitable under a rapidly changing international onvironment. Similarly, much of the private investment was made with guarantees or loans from government-owned banks and became a public liability when proven unprofitable. 3 interest rates reached unpracedented high levels. The new administration of President Coreson Aquino is attempting to increase output, employment and living standards, but In an environsmnt of constrained domestic and interna- tional resources, prospects of recovery are not encouraging. 8. Project Implementation in all sectors suffered from the downturn of the Philippine economy and from the counterproductive policies that preceded the crisis. The first and second urban projects were no exceptions. Direct sector rejarcussions of macro conditions were apparent in the declining per- formance of the construction industry under accelerating inflation and tighter credit; in delayed transfers of funds for project investments; in the depressed demand for coc- rcial and industrial lots; in greater difficulties of recovering project costs as urban unemployment and political interference increased; and in less efficient project administration. The attention of public sector managers became increasingly focused on more pressing political issues to the detriment of project performance. These circumstances partly explain why the projects being audited achieved less than fully satisfactory results. While both projects focused upon high priority urban development investments, their chancos of success were reduced due to the prevailing coun- try conditions. Urban Trends, Institutions and Policy 31 9. Although the Philippines was still a predominantly rural economy in the mid-1970s, the urban population had expanded rapidly during the previous decade, reaching approximattly 12.3 million (or thirty percent of the total) in 1975. Urban growth had intensified the dominance of Manila in the coun- try's urban structure. With an estimated 5.6 million inhabitants in 1975, Manila Metropolitan Area (MMA) comprised roughly 13 percent of the national population and accounted for nearly one-half of the gross domestic product. 10. Much of the urban population growth resulted from economically motivated internal migration.4/ Avorage incomes in the cities were signifi- cantly higher than In rural areas and urban employment was expanding at 31 This section draws from several studies on the urbanization of the Philippines. Complete references are provided in Annex I to this PPAM. 41 For a discussion of migration trends in the Philippines, see KLL. Zacharish, 'Migration in the Philippines with Particular Reference to Less Developed Regions,, The World Bank, Internal Memorandum dated July 2, 1975. 4 approximately th same rate as the labor force. .I The growth of urban employment during the 1965175 period, almost wholly In the formal and informal service sector (estimted to account for 67 percent of all urban jobs in 1973), prevented open urban unemployment from rising above 12 porcent.1I Howver, sustained labor forc expansion and low labor absorption In manufac- turing led to increasingly higher rates of underemployment, as seasured by the incidence of part-time and poorly paid workers. 11. Depite higher average income levels, the distribution of income and public services in urban areas since World War II was very skewed, making for wide disparities in standards of living, a high Incidence of poverty and rapid expansion of slum and squatter settlements where environental conditions were appalling. Typically, the urban poor lacked regular employment, occupied crowded and structurally Inadequate housing (mostly as renters or squatters) and had limited or no access to water supply, sanitation, electricity and transportation. 71 By the mid-1970s, an stimated 40 percont of the urban population (*ad 80 percent of the slum population of major cities) had insuf- ficient means to obtain an adequate diet and satisfy other essential needs. Such were the prevailing conditions of the 180,000 inhabitants of the Tondo Foreshore slum in banila -- the largest and most visible settlement of its kind in the country, which became the focus of the first urban project. 12. Informal sector production accounted for the bulk of tho housing supply in urban areas during the 1960.. Formal sourcos of finance for housing were limited and accessible only to persons at the highest levels of the urban 5/ It is estimated that total family income in the Philippines grew by about 4.7 percent per annum in real terms during 1961171 and that urban incomes Increased at about 5.5 percent a year during that period. The rurallur- ban income gap persisted throughout the 1970.. By 1982, the median family Income in Metro Manila was estimated at P 1,750 or more than twice the national level. See, BCS, Family Income and Expenditures SurvjsL (of 1957, 1961, 1965 and 1971). See also Coleman et al., "Shelter Sector Assessments An Update on the Phlilppines,' paper prepared for the United States Agency for International Development, Washington, D.C., June 1984. 6| Between 1965 and 1975 measured unemployment remained at approximately 5 to 7 percent for the Philippines as a whole and declined in the MM from 16 to 11 percent. Although systematically higher than the national average, much of the unemployment in the MM and in other urban areas wa of temporary nature, reflecting the fact that migrants often spent their initial weeks or months in the cities *earching for jobs. See The World Bank, 'The Philippines -- Priorities and Prospects for Devolopment, Basic Economic Report,, (Report No. 1095a/PH), datod May 5, 1976. 71 At the end of the 1960o, only half of all urban dwellings had piped water, a fifth had electricity, and a tenth had a flush toilet. Se8 World Bank, SAR, NManila Urban Development Project" (Report No. 1032l- PH), dated May S, 1976. income distribution scale. Contrary to what has ben obsorved in several other developing countries (where iaitially low-standard bouaing units are gradually upgraded), the quality gap between formally and informally financed housing in the Philippines has tended to perdsit over time.8i Thus, although housing production expanded at an average rate of 4.5 percent por year (or approxi- mately as fast as the urban population), most of the new units built were of low quality. Therefore, the houaing problem of the mid-1970s centered upon the rapidly deteriorating quality of the housing stock and its lack of ade- quate infrastructure services. 13. The size of the housing problem in the 1MA w"s of particular concern to the Government. A large concentration of poorly housed people In the capital was seen by the Government not only as contrary to the Image It wanted to project of Manila (as a modern city capable of attracting new businesses and tourism), but also as a potential political threat to the regime. 9( To address this problem, the Government financed partial demolitions of squatter settlements and forced relocation of an estimated 16,000 squatter families from the MA( to the periphery of the city. The program proved to be ill- conceived and wasteful. By 1973 most families had already abandoned their resettlement sites and returned to Manila. 14. In general, Government programs had not been concerned specifically with the spatial distribution of development or with programs to aid the urban poor. Confronted with armed insurrection in the provinces, both the private sector and the Government concentrated investments in Manila, constraining economic growth outside Central Luzon.101 In the capital, public investments during the early 1970. were directed to the expansion of water supply and electricity systems, but housing, sewerage disposal and transportation ser- vices were neglected. 15. It was only during the 1975180 period that the Philippine public housing system began to expand, both institutionally and in terms of actual investments. In 1975, coinciding with the approval of the first urban proj- ect, the Government created the National Housing Authority (NHA) and the Manila Metropolitan Comission (MMC). The NHA had a mandate to provide shel- ter to the urban poor throughout the country (primarily through slum-upgrading and development of serviced plots) and was chosen to act as principal agency for the implementation of the first urban project assisted by the Bank. The 8/ See Struyk, R. and Turner, M., Housing Finance and Housing Quality in the Philippines and Korea, The Urban Institute, Project No. 3312, March 1984. 91 In a 1972 study, the Presidential Assistance on Housing and Resettlement Agency (PAHRA) concluded that about 50 percent of the national housing shortage was concentrated in the HMA. 10/ Moderate efforts to promote urban decentralization were Initiated by the Government in the mid-1970s. These consisted of incentives for the location of small and medium scale industries outside the NMA and of investments in water supply programs in several regional cities. 6 MMC wa expected to coordinato the activities of the 17 local governmants operating in the Metro Menila and play only an incidental role In project execution. 16. On the occasion of the approval of the second urban project other institutional changes were made. In 1978, the Ministry of Human Settlements (MHS) was created and shortly thereafter three new agencies were established to operate a secondary housing mortgage finance facility. Under this newly organized system, public housing financo was highly subsidized and conceived as an incentive to the private sector housing industry. In practice it served primarily the more affluent segments of the urban population. 11/ As a result of these various initiatives, by 1980 the MHS had consolidated under its jurisdiction the various functions of public housing production, finane-, regulation, and policy formulation; program implementation became the respon- sibility of seven semi-autonomous public corporations, theoretically coordina- ted by the MHS. 121 Notwithstanding the financial subsidies and the expanded state presence in the housing sector, both public and private formal housing finance combined served only an estimated 20 percont of the Market in the early 19809. 17. In sum, throughout the 1970s and early 80s, the urban sector in the Philippines underwent many institutional cnanges, but the prevailing govern- mental policies and practices showed little support for distribution objec- tives. The NHA's mandate to serve the urban poor remained isolated and unstable in this context, sustained primarily by the support received from the World Bank. Lank sector policy, informed by several urban sector missions which visited the Philippines during the 19709, focused on the need to reduce the great disparities in access to basic public infrastructure within urban areas (particularly in Manila) and on the need to supply regional cities with improved public services, fostering decentralization. Despite the apparent 111 For a detailed analysis of the Philippines housing finance system see The World Bank, 'The Philippines Housing Finance,, (Report No. 3732-PH), dated April 12, 1982; and Raymond J. Struyk and Margery A. Turner (1985). 12/ In 1980 the public shelter system in the Philippines comprised: (i) housing finance agencies including the Home Development Mutual Fund (HDMF), the National Home Mortgage Finance Corporatwon (NHMFC), and the Home Financing Corporation (HFC); (ii) housing production agencies, including the National Housing Authority (NHA) and the Human Settlements Development Corporation (HSDC) with its subsidiaries BLISS Development Corporation and National Housing Corporation; and (iiI) a regulatory agency denominated Human Settlements Regulatory Cosmission (HSRC). For a brief description of these seven housing agencies, see Annex H to the PCR of the Manila Urban Development Project. Also, a very useful analysis of the institutional development of the Philippines' housing sector, with particular emphasis on the NHA, is provided by Kearns, J.M. The Philippines National Housing Authority: Lessons from Ten Years of Institutional Development,' The World Bank, Internal Discussion Paper, dated December 1985. 7 lack of Governmental support for these prioritios, they became central objec- tives of Blnk lending in the Phllippine through the firat and second urban projects. The joint audit of these projects offers an opportunity to examine the extent to which discrete projoet objectives are sustainable In an advers, policy environment. II. PROJECT OBJECTIVES AND DESIGN Manila Urban Development Proiect 131 18. This wa the first shelter project financed by the Bank in the Philippines and also the first attempt to demonstrate a more efficient alter- native to the Government's large scale slum demolition and rosettlement pro- gram in Manila. The project would provide improved onvironmental sanitation, infrastructure and social services for the Tondo Foreshorc, one of Manila's worst slums, develop a first phase of sites and services program at nearby Dagat-Dagatan, and build and Improve trunk infrastructure and roads In the Tondo Foreshore area. It would also support traffic Improvements in the MM and provide technical assistance to key sector institutions. 19. The project was appraised in August 1975. Total costs were estima- ted at $65.0 million equivalent (P 487.8 million), 49 percent of which was in foreign exchange. Bank loans 1272/1282-PH for $32.0 million, approved in May 1976, were to finance the entire foreign exchange component; the Borrower, which was the Government of the Philippines, would finance the remaining local project costs. 20. The new NHA would be responsible for overall project coordination, execution of the Tondo upgrading and Dagat-Dagatan sites and services compo- nents, and the administration of various advisory services. 21. The other agencies involved in project implementation were: (i) the Metropolitan Waterworks and Sewerage System (MWSS) for tho trunk sewer and water supply components; (it) the Department of Public Highways (DPH) for major roads, bridges and traffic components; (Ili) the Developmont Bank of the Philippines (DBP) for the administration of a line of credit for materials for home improvement and small businesses development at the housing sitesl (iv) the Department of Public Works, Transportation and Communications (DPWTC) for a land use and transportation study, a two-year program of cadstrum surveys, and for additional studies intended to strengthen the newly created MHC; (v) technical assistance to small businesses at the housing sites was to be pro- vided by NRA through the Medium and Small Industries Coordinated Action Program (MASICAP); and (vi) the Department of Health (DH) and the National 13/ See Staff Appraisal Report (SAR) No. 1032a-PH dated May 5, 1976, President's Report No. P-1763-PH dated May 17, 1976, and Loan Agreements 1282-PH and 1272-T-PH dated June 9, 1976. 8 Nutrition Council (NNC), undor the supervision of NIA, would adminlster the health progrms eomwn to Tondo and Dagat-Dagatan. 141 22. The effectiveness of the Bank loans was conditional on tho signature by the Borrower of Subsidiary Loan Agreements with NRA and HWSS. Other spe- cial conditions of loan offectiveness included employment of engineering consultants by NHA to assist In project execution and manageoent development, and thc issuance of a Prosidential Letter of Instruction (LOI) specifying the responsibilities of each agency involved in the project. 23. Project components (described in the PCR, par&. 2.06), with estima- ted costs with contingencies, comprised: (i) Tondo Foreshore U,gradint (180 ha), serving approximately 15,000 housing structures as well as che. development of 1,000 new residential lots and 15 ha of Industriallcommercial estate, established at $27.1 million equivae10r (or 41.7 percent of total cost); (ii) Dazat-Daaatan sites and services (40 ha), including development of some 2,000 residential lots and 3.5 ha of industrial/commercial estates, estimated at $7.7 million equivalent (or 11.8 percent of total cost); (iii) Offeite Sever and Water SuPPlY Trunk Infrastructure serving the Tondo and Dagat-Dagatan sites, estimated at $1.5 million equivalent (or 2.3 percent of total cost); (iv) Major roads (7.5 ktn) and two bridges, estimated at $13.6 million equivalent (or 20.9 percent of total cost); (v) Traffic im,rovements, including procurement of traffic signal and radio equipment and construction of intersection works, road mar- kings and bus shelters, estimated at $6.9. million equivalent (or 10.7 percent of total cost); and (vi) Technical Assistance for detailed engineering (NHA, MWSS, DPH), man- agement support (NHA, DPH), preparation of future projects (NHA), special studies (MPWTC and HPWTC on behalf of MHA), training and fellowships, estimated at $8.2 million (or 12.6 percent of total cost). 24. The housing components were conceived as including a wide rang of comunity services. Settlers at Tondo and Dagat-Dagatan would be provided with secure tenure to residential lots served by roads, footpaths, surface drainage, water supply, sewerage and street lighting, as well an complementary health, nutrition, primary and secondary education and other social services, 14/ In addition, the NRA had the option of engaging the Philippines Business for Social Progress Foundation (PBSP) and/or the Philippines Notional Bank (PNB) to assist DBP in the provision of credit to small busin$esse. 9 and would have accses to building matorials lo s for hom Improvements. In addition, the 3,000 resettlement lots would contain sanitary core units to facilitate housing consolidation. Employment creation would be prooted at the sites through 18.5 ha of serviced Industrial and comercial estates, and a supervised line of credit for small scale businesses. Land at both sites would be transferred to settlers through a twenty-five y-ar leso instrument with option to purchase after five years. 25. Upgrading works at the Tondo were expected to displace approximately 1,500 families, one-third of which would be relocated to Dagat-Dagstan and the remainder accomodated at the new lots developed on site. However, because other public works In the Tondo area would di lodge some eight to nine thou- sand additional families, relocation needs far exceeded the property's capa- city to provide new serviced sites.151 26. Cost recovery provisions for the housing components wer estimated to yield approximately $30 million *quivalent, or 82.7 percent of the combined costs of site development and off-site infrastructure. The national govern- ment would absorb 53.6 percent of total project costs, including the remaining housing related costs and the full cost of technical assistance, major roads and traffic components (see PCR, Annex K). The costs allocated to the housing beneficiaris would be reduced, by approximately sixty percent, through cross- subsidies afforded by NHA on the basis of expected profits from the sale of commercial and industrial properties on site and revenues from other public ag*ncies. 27. Procurement conditions provided for international competitive bid- ding (ICB) for civil works contracts valued at more than $1.0 million equiva- lent each. Civil works contracts of lesser value would be procured on the bais of local competitive bidding (LCB). Contracts for equipment and mater- ials valued at over $100,000 equivalent would be procured through ICB and those of lesser value through LCB, excopt for equipment under the small busi- ness credit program, which would be procured according to DBP procedures. An estimated 46 staff-years of consulting services would also be contracted for the project. 28. The project was justified by its expected employment impact, direct welfare gains for approximately 180,000 persons living at the Tondo, and Increased traffic efficiency. The weighted overage economic rate of return on Investments at the Tondo Foreshore, Dagat-Dagatan and major roads components of the project was estimated at 27 percent (SAR, para. 6.15). 15/ The resettlement of approximately 9,200 families was associated with the following public workss Tondo upgrading (1,500 families, of which 1,000 would be resettled on site), two roads financed under the project (700 families), the international port complex (3,800 families), and a fisheries port at Navotas (3,200). During Negotiations the Government agreed to transfer to NUA the funds necessary to pay for the incremental cost of providing serviced lots for the families to be resettled as a result of the port expansion and highway construction works. 10 29. The project was expected to be completed in four years, by March 31, 1981. The loan closing date was scheduled for six months later, on Septembor 30, 1981. 30. At the Board meeting which considered the project, Bank Executive Directors discussed the risks arising from assigning major project Implemen- tation responsibilities to NHA, in view of the obvious Inexperience of that agency in the type of work proposed. Additional questions were raised as to the level of Government commitment in support of NHA and the project; the risks of speculation associated with the process of land lease transfors at the Tondo; and the relatively narrow project approach to the large urban problems of the Philippines, which called for a broader urbanization strategy. Considering these various risks, Board members requested to be periodically informed of the progress of the project and expressed interest in visiting the Tondo on occasion of the following Annual Meeting, scheduled to tak. place in Manila in 1977.16/ Second Urban Devlovment Prolect 17, 31. The second project was appraised in October 1977, before the perfor- mance of the first project could be assessed. Inevitably, therefore, both projects shared the relatively high risks of experimenting with new programs and institutions.181 32. Although similar in concept to the first project, the objectives of the second project were more specifically focused on the housing sector, where emphasis would be given to the expansion of the Government's upgrading and sites and services program to three regional citiess Cebu, Davao and Cagayan de Oro. In Manila, sites and services development was expected to represent a net addition to the city's housing stock rather than simply provide for relo- cation needs. 33. The objectives of both projects convergod In the areas of employment creation (through provision of commercial and industrial estates and small business loans) and Institutional development (through technical assistance to 16/ See Minutes of World Bank Board Meeting of 05127176. 171 See Staff Appraisal Report No. 2048a-PH dated December 1, 1978, President's Report No. P-2431-PH dated December 7, 1978, and Loan Agreement (Loan 1647-PH) dated January 26, 1979. 18/ The SAR (paras. 1.21-1.26) refers to the experience gained at the time under the first project as satisfactory, noting that the NHA staffing problems, which had caused delays, had been resolved. Nonetheless, the SAR acknowledged several project risks, including the inexperience of the local units formed to implement the project at regional cities; addi- tional staffing problems at NHA, which were likely to arise from the rapid expansion of its work program; and financial uncertainties due to the untested ability of NHA and the participating local authorities to collect payments from project beneficiaries. NHA and the participating loc-al governments). Technical assistance under the second project would also be extended to the National Kconomic Development Authority (NEDA), to the HC and to the newly created MRS. In addition, the project would support the Government's drainage and flood control program in Manila through procurement of flood control equipment and technical ssis- tance. 34. Total project cost, estimated at $69.9 million (P 517.7 m), included 46 percent in foreign exchange. Bank loan 1647-PH for $32.0 million equiva- lent, approved in December 1978, would finance the foreign exchange component of the project. The Government, which was the Borrower, would finance the remaining local costs. 35. The project's financial plsn allocated $26.2 million (81.9 percent) of the proceeds of the Bank loan to be re-lent by the Borrower to NRA (under a Subsidiary Loan Agreement to be entered into between the Borrower and NRA) on terms and conditions approved by the Bank. The NHA was expected to relend to local governments the necessary amounts for the implementation of the project in regional cities. Of the remaining loan proceedings, $4.6 million (14.4 percent) would be made available by the Borrower to the Ministry of Public Works, Transportation and Communications (MPWTC) and $1.2 million (3.8 per- cent) to NEDA, MWC and MRS. The foreign exchange risk on the Bank l.oan would be borne exclusively by the Borrower. 36. Loan effectiveness, planned for March 1979, would be conditional upon the signing of the Subsidiary Loan Agreement between the Borrower and NHA. The Borrower was expected to satisfy other special loan conditions Including provisions for: additional NHA staff; periodical reports on the technical, financial, and personnel aspects of project implementation; yearly audited project accounts; establishment of satisfactory procedures for the selection of housing beneficiaries under the project; maintenance of adequate project implementation offices at regional cities; and timely information about any major public expenditure (other than the project's) planned in the project areas. 37. Project components (described in the PCR, para. 2.05), with respec- tive costs and contingencies, comprised: (i) Dazat-Daaatan sities and services (90 ha) for about 8,600 families in Metro Manila, at an estimated cost of $33.6 million equivalent (or 48.1 percent of total project cost), including land acquisition, provision of water supply, drainage, sewerage paved footpaths and streets, schools, clinics, and material loans for house improvement. On-lot development would range from single service connections to a shell house; (ii) Sites and services (3.800 families) and uggradina of about 74 ha (7.600 families) in the regional cities of Cebu, Davao and Cagayan de Oro, at an estimated cost of $15.8 million equivalent (or 22.6 percent of total cost), including land acquistion and reclamation, and the provision of water supply, waste disposal, surface drainage, sea defense works, footpaths and streets, health, education and 12 social services, and building saterials loans for hoam Improvement. Development standards at sites and service areas would provide for lot eonnections to utility networks, while only public water tape and septic tanks (or pit privies) would bo provided at upgrading sitesg (iII) Industrial and co.m.rcial estates at Da-at-Dta-tan and two reaional cities, involving development of approximately 56 ha (40 ha in Dagat-Dagatan and 16 ha in rogional cities) at an estimted cost of $9.4 million (or approximately 13.4 percent of total cost); (iv) Small business development In Dagat-Dagatan and regional cities including technical assistance and a line of credit stimated at $2.2 million equivalent (or 3.1 percent of total cost); (v) Flood control equigment for Metro Manila in support of the Government's flood control program lnitiated In 1972, at an estima. ted cost of $3.0 million equivalent (or 4.3 percent of total cost); and (vi) Technical assistance and traininat at an estimated cost of $5.9 million equivalent (cv 8.5 percent of total cost), including foreign and local advisory services to bo allocated to NHA, NEDA, MMC, MHS and MPWTC for institutional development, special studies and support for project implementation in Manila and at regional cities. 38. The NHA would have the largest share of project implementation responsibility, including all elements of the Dagat-Dagatan sites and services component (except main drainage); all elements of the slum upgrading and sites and services components at regional cities (except sea defense works); all elements of the employment components; and the administration of various technical assistance services. 39. Other agencies involved in the Implementation of the project included MPWTC for construction of main drainage and sea defense works at the housing sites, procurement of flood control equipment, and preparation of an updated Metro Manila Flood Control Master Plan; and NEDA, MA, and NMS, responsible for studies and training activities financed under the technical assistance component. 40. All housing beneficiaries under the project would sign a 25-year lease contract with NHA, which contained an option to purchase the freehold interest after a qualifying period of 5 years or less, to be determined in each city according to local conditions. As in the first project, the cost recovery approach in this second project aimed at recovering from project beneficiaries (through the lease cum mortgage and the sale of C/I lots) all investment costs other than those normally financed from general tax sources and utility charges. Beneficiaries of residential lots would be charged 75 percent of development costs and buyers of industrial and commercial proper- ties would pay approximately 110 percent of development costs. The projected profits from the sale of industrial and commercial areas would be used to subsidize the price of residential lots. 13 41. Procurement procedures, virtually identical to those of the first project, established ICS procedures for civil works contracts valued at more than $1.0 million equivalent and LCB for those of lesser value. Except for materials for house omprovement, all contracts for goods and equipment valued at over $100,000 would be awarded on the basis of ICB. For purposes of bid comparison, local manufacturers would receive a preferenco of 15 percent of c.i.f. cost (or the level of customs duties, whichever was lower). Tne remaining contracts for goods and the procureoent of materials for house Improvement would be tendered by LCD. Approximately 84 staff-years of con- sulting services werc planned for technical assistance, over half of which was to be supplied by local consultants. 42. The justification for the project rested on Its expected poverty alleviation impact, realised through the distribution of housing and commuity services (estimated to benefit approximately 20,00C lower Income families in Manila and regional cities), and the creation of some 25,000 jobs in the vicinity of the housing sites. Of total project cost, some 72 percent was estimated to benefit the urban poverty group directly. The weighted average economic rate of return for the housing and employment components, which represented 85 percent of the project cost, was estimated at 34 percent. 43. The project was expected to be completed in five years, by December 31, 1983, and the Bank loan closing date was scheduled for June 30, 1984. III. PROJECT IMPLEMENTATION Implementation Schedule 44. Both projects experienced considerablo implementation delays. The first project took ten years to complete rather than four, and the second took six and a half years Instead of the five years projected at appraisal. The PCRs give detailed accounts of the factors leading to the unsatisfactory performaace and clearly show that appraisal schedules were over-optimistic given the innovative nature of several components, the inexperience of the participating agencies, and the difficult political climate surrounding the project. 191 45. The start of the two projects was delayed due to inadequate prepara- tion. Rights-of-way and detailed engineering designs were lacking for most components. The topographic surveys and soil tests, conducted during the preparation period, were inaccurate and had to be redone. Under the first 19/ In the case of the first project see PCR, paras. 4.01-4.06. For the second project see PCR, pars. 4.01-4.13. In general, the implementation schedules prepared at appraisal made insufficient or no provisions for the completion of detailed engineering, land acquisition, execution of financial and social elements of the projects; and did not account for the learning time required by new units and project agencies. project, construction of the sewer outfall and the major roadt components began only four years after loan approval. 46. Most of these initial problems could have been avoided. Bank exper- ience at that time had already demonstrated the iWportance of acquiring land and completing detailed engineering a part of the project preparation pro- cess, as for example in the case of the Lusaka Urban Project and the Senegal Sites and Services Project, both approved in 1972. Unfortunately, these lessons appear not to have been applied to the first and second urban projects in the Philippines. 47. Delays continued to accumulate during project execution due to inefficiencies in contract administration and procurement problem at times involving unsatisfactory quality of work by contractors and consultants. These difficulties were due to the inexperience of NRA compounded by other factors, such as unanticipated physical contingencies, lack of coordination among implementing units and agencies, and changes in project design during execution. 48. In the case of the first project, technical difficulties were aug- mented by the sheer size of the Tondo upgrading component, the complexity of which neither the Bank nor NRA had predicted. The process of reblocking, whereby plots were realigned and houses were moved within each block, for example, resulted in the relocation of approximtely 14,000 structures (or almost the entire community) instead of only the few layout adjustments planned at appraisal. Operations on site were seriously disrupted by politi- cal problems arising from community groups competing for power and trying to use the Bank to discredit the Government, whose policies toward squatters outside the project area were inconsistent with policies adopted under the project. 49. Implementation difficulties were carried over from the first to the second project. Over a year elapsed between loan approval and the first significant allocation of local funds for that project. The coordination of activities by different agencies clearly remained beyond NHA control. 50. In Manila, start-up problems at Dagat-Dagatan (which corresponded to 65 percent of the project) were surmounted only In the third year of implemen- tation. Land reclamation on site posed special difficulties, the original drainage built by MPWTC proved defective, and the necessary corrections were postponed by tWo years until interagency differences regarding the works were resolved. Construction standards for the water supply and sewer systems were upgraded by MSWW mid-way during implement-tion. As a result, construction was delayed pending completion of new engineering designs and contract revisions. The quality of construction was often wanting. A main sewer was found to have serious structural defects and the corresponding contract had to be reten- dered. 51. In regional cities (whose components represented 24 percent of the project), works began more than a year late on account of land acquisition problems, difficulties in securing community cooperation, and unwillingness of local governments to commit funds in advance of Bank loan effectiveness. 1S Throughout lmplementation, project administration suffored from the Inexperi- *nc- of local goverAneuts In the kinds of work Involved. 52. The growing scarcity of local financial rsources In the 1980s created additional implementation problem_ for both projects. From that time on, counterpart funds were disbursed late and in insufficient amounts. This circumstanc- forced NHA to use cxpnsive short-tenc bock loan to maintain project momentum. High Inflation rat" In 1983 and 1984 led to construction price escalations in excess of the contingency sargins built into most civil works contracts, requiring the cancellation of seven contracts In Dagat- Dagatan and additional time for retenderlng. 53. Remedial action afforded by the Bank's 1983-1984 Special Assistance Program (SAP), although somewhat late, proved essential for the completion of both projects. Under the SAP, a supplemental loan of US$10.5 sillion, made in March 1984, helped finance cost overruns for the major roads and bridges component under the first project. For both projects, a revolving fund was established for the disbursement of loan proceeds and disbursement percentage on civil works were raised from 43 to 85 percent. Th e measures accelerated disbursements and ultimately permitted tho belated completion of the two projects. Modification in Proiect Design 54. Physical and administrative changes were made to the projects during implementation. In the first irolect, the main changes in physical scope reflect inadequate pre-appraisal preparation and includes (i) in the Tondo uparading component land reclamation works exceeded estimated quantities by over six times due to unforeseen subsoil conditions and obstructions, and the total upgraded area was reduced by 24 percent (from 180 ha to 137 ha) causing a 14 percent decline In the number of housing structures bmproved (from 17,500 to 15,000) and increasing by four times (from 500 to 2,500) the number of structures that had to be demolished; (ii) in the sewer outfall component, design specifications for pipelines were raised (from 48" to 72" diameter), a wator main was built across the C-2 highway, and the scope of work of the ewer inter- ceptor was increased; (iii) in the ma1or roads and bridges component, additional works under- taken included filling the Fisherman's Channel for alignment of highway C-2, construction of a bridge over the Navota River, and extension of highway C-2 through the Dagat- Dagatan industrial and commercial area. The volume of civil works under this component is estimated by MPWH to have increased by 110 percont over SAR esti- mates. To help finance cost overruns of this component, a supple- mental loan (No. 1282-1-PH) for $10.5 million equivalent was approved in February 1984; 16 (iv) in the traffic imarovement component, civil work, valued at approxi- mately $2.4 million equivalent were deleted. comprising mrosuDent of intersections, construction of bus shelters, road mArkings end footpaths. 55. In the second project, approximately $4.9 million equivalent in civil works were deleted to maintain overall costs within the appraisal bud- get. The elements deleted compriset (i) the Carbon upgrading site (1 ha) at Davao, originally budgeted at approximately $0.1 million equivalent; (ii) a water main at Dagat-Dagatan, estimated at appraisal at approxi- mately $0.6 million equivalent, which was built under the regular MWSS capital investment program; (iii) three of the five elementary schools at Dagat-Dagatan, estimated at appraisal at approximately $2.1 million equivalent; and (iv) on-lot development at Dagat-Dagatan, including service connections, fire walls, and housing options, estimated at appraisal at approxi- mately $2.1 million equivalent. 56. In addition, the scope of the industrial and coummercial lots compo- nent was increased by approximately 46 percent in terms of area developed. The increment occurred in regional cities (Davao and Cagayan de Oro) and neither the PCR nor other project documents provide a clear Justification for this particular change. 57. Both projects introduced administrative changes during implementa- tion, of which the most important are the following: (i) transfer of the responsibility for administering the small businesses and materials loan Programs from local banks (DBP and PNB) to NHA; (Ui) modification in the formula for recovering costs from project bene- ficiaries (which resulted in much higher subsidy levels) and in the form of tenure under which land would be transferred (changed from leasehold with option to buy to freehold tenure). 58. The closing date for the Bank loans was extended for the first project by four-and-a-half years from September 1981 to December 1985, and for the second project by one year from June 1984 to June 1985. Project Costs and Disbursements (a) Costs 59. The Audit is limited in its analysis of project costs by insuf- ficient data and inadequate appraisal cost estimates. The PCRs for both projects provide incomplete actual expenditures (for the second project actual 17 costs of flood control and technlcal assistance components are missing), show these data In peso amounts aggregated by component rather than on a yearly basis, and offer only averag* exchange rates over the life of the projects as the basis for converting total project cost into US dollar equivolents.201 This is unfortunate since currency conversions are likely to be significantly distorted if made on the basis of Imprecise, aggregated data, particularly under highly variable inflation rates as in the Philippines during the first half of the 1980.. 60. Comparison between actual and appraisal cost estimates are hampered by the various modifications made to the scope of work of both projects during I'plementations as well as by the deficiencies of some of the appraisal esti- mates based on preliminary engineering. 61. These uncertainties notwithstanding, cost performance in US dollar terms seems to have improved somewhat from the first to the second project.21/ At an estimated total cost of $91.86 million equivalent, the first project exceeded the $65 million appraisal figure by over 40 percent, while the cost of the second project, at $67.75 million equivalent, resulted 3.1 percent lower than the appraisal estimate. Notwithstanding the apparent improvement, cost performance under the second project remains uncertain because the actual impact of currency devaluations cannot be measured with the available dats.22/ 62. The estimated cost overruns under the first project reflect substan- tial increases in the actual costs of major roads and bridges, sewer outfall, and Tondo upgrading components, which together accounted for 65 percent of project costs at appraisal. The increases may be attributed to extensive start-up delays, large increases in the price of land, additional expenses associated with deficient works requiring replacement or extensive repairs, and generally higher labor and materials costs resulting from the second major increase in oil prices in 1978/79. 231 In addition, cost increases reflect the underestimation of work quantities at appraisal. The deletion of the civil works from the traffic component, combined with minor differences in the cost of technical assistance, kept overall project costs from rising further. 20/ Project costs are reported in the PCR, paras. 4.07-4.09 and Annex B for the first project, and in the PCR, paras. 4.14-4.21 for the second proj- ect. In both cases detailed expenditures by components are shown in Philippine pesos only. 211 Total costs for the first project were converted into US dollar equiva- lent using the average exchange rate of P 10.6 - US$1.00. For the second project the average exchange rate used was P 11 - US$1.00. 22/ Using preliminary expenditure figures expressed in current pesos, the PCR estimates a 45 percent cost overrun for the second project (Table 4.1) despite reductions in project scope. 23/ See for example, Letter from MPWH dated 03/21/86 in PCR, Annex M, where the Borrower estimates at 110 percent the increase in scope of work for the major roads and bridges component under the first project. 63. In the second project, cost inctases in pesos war* asociated with similar factors. Larte varlations In the work quantities for land reclamation and sharp rlass In land prices increased considerably tho cost of shelter development, while Implemntation delays led to extonsions In th engagement perlod of xpatriate consultants, causing technical aistance cost to rise. As notod above, reductions in project *cope and the devaluation of the peso kept overall costs below appraisal estimate. 64. The following tables uarilse tho appraisal and current cost esti- mates for the two projects (Table 1) and show tho variations in cost structure which occurred during implementation (Table 2). TABLE 1: Total Pro1Oct CoEt - Estimated vs. Actual (in current US$ million) Component Urban I Urban II SAR Actual Difference SAR Actual Difference (X) (X) Shelter la 30.22 42.77 + 41.5 45.68 39.47 - 13.6 Upgrading 21.94 34.33 + 56.5 6.27 8.79 + 40.2 Sites & Services 5.39 8.13 + 50.8 37.35 30.35 - 19.1 Materials Loans 2.89 0.31 - 89.3 1.88 0.33 - 82.4 Employment lb 6.03 8.83 + 46.4 15.31 18.46 + 20.4 C/I Lots 5.59 8.67 + 55.1 13.15 18.00 +136.9 Small Business Loans 0.44 0.16 - 63.6 2.16 0.46 - 78.7 Urban Transport 20.59 34.83 + 69.2 NA NA NA Major Roads/Bridges 13.66 31.97 +134.0 NA NA NA Traffic Improvement 6.93 2.86 - 58.7 NA NA NA Technical Assistance and Trainina l_ 8.19 5.43 - 33.7 5.93 7.60 + 28.2 Flood Control Equipment le NA NA NA 3.02 2.22 - 26.5 TOTAL 65.03 91.86 + 41.3 69.94 67.75 - 3.1 /a The cost of upgrading and sites and services includes land, on-site and off-site infrastructure, community facilities, utilities and core-unit construction. In the case of upgrading the estimated cost also includes in-fill sites end services. Lb The cost of commercial and Industrial lots (C/I) was shown in the apprai- sal and PCR tables for both projects as part of the cost of upgrading or sites and services. The Audit has disaggregated these costs based on the area used for commercial and industrial lots in each site as a proportion of the total. Le The actual cost of technical assistance and flood control equipment under the second project was estimated by the Audit on the basis of disbursement figures for Bank Loan No. 1647-PH since cost data on these components are lacking in the PCR. 20 TABLE 2, Cost Structur- - Aoraissl vs. Actual (in percentages) Component Urban I Urban II SAR Actual SAR Actual Shelter 46.5 46.6 65.3 58.3 Employment 9.3 9.6 21.9 27.2 Transport 31.7 37.9 NA NA Technical Assistance 12.6 5.9 8.5 11.2 Flood Control NA NA 4.3 3.3 TOTAL 100.0 100.0 100.0 100.0 (b) Disbursements 65. Both projects failed to utilixe the full amount of the Bank loans due to exchange rate fluctuations and in spite of the increase in disbursement percentages made under the SAP. The undisbursed balance of $3.2 million equivalent on loans for the first project (7.5 percent of the total $42.5 million approved) was cancelled on the closing of project accounts in July 1986. Likewise, the undisbursed balance of $1.4 million (or 4.4 percent of the original loan amount) on loan 1647-PH was cancelled in July 1985. 66. The disbursement profile (see Figure 1) closely approaches the Bankwide average for urban projects in the second project, but lags consider- ably behind that curve in the first project. 24/ In both cases, actual dis- bursement periods significantly exceeded the optimistic appraisal projections. Disbursements on both projects accelerated after FY83, reflecting the positive impact of the measures taken under the Bank's Special Assistance Program. 67. Table 3 shows the proportion of total project cost financed by the Bank loans as compared to appraisal projections. The Bank's share doclined from 49.2 to 42.7 percent in the first project but remained only slightly below the appraisal estimate for the second project. 24/ The disbursement profile drawn by the Audit for the first project diffors from that shown in the PCR because the latter was prepared prior to the closing of loan accounts. - 21 - Figure 1 PHIUPPINES MANILA URBAN DEVELOPMENW PROJECT (LOAN 1272.PH/1282 PH) SECOND URBAN DEVELOPMENT PROJECT (LOAN 1647.PH) DXbunment ProUb Y d Totd LoWn 100 90 - 13 80 D 70 - 60 - 40: -ecr 1 2 3 4 5 6 J 8 9 10 11 30A I-7 7 9 8 1 82 8 4 8 5 8 URBAN II7 7 79 so 81 32 83 84 as 8668 PCR Chaft V sh4 grph of 'Actual Uno ir dWuasefnts shdted fownd by one year. Le. asni date of eecthes was 4/26/79, fV80 ws used as Year i bdisbursement. Wo0d 8nk404,2 22 TABLE 3t Proiect Financing (in percentages) Source of Finance Urban I Urban II SAR Actual SAR Actual Borrower 50.8 57.3 54.2 54.9 Worid Bank 49.2 42.7 45.8 45.1 TOTAL 100.0 100.0 100.0 100.0 IV. PROJECT RESULTS 68. The evaluation of both projects was based exclusively upon documents made available to the Audit and the lack of direct field experience partly explains the tentative nature of some of the conclusions presented in the following paragraphs. Physical Accomplishments of the Manila Urban Development Proiect 69. The physical accomplishments of the project were satisfactory des- pite considerable implementation delays and cost overruns. Investments in shelter, corresponding to approximately 46.6 percent of actual project cost ($42.8 million equivalent), resulted in the upgrading of approximately 15,000 housing structures at the Tondo Foreshore Area and in the development of 2,500 serviced sites for relocation in Dagat-Dagatan (2,000) and Tondo (500). Compared to appraisal targets, these achievements represent approximately 71 percent for upgrading and 83 percent for serviced lots. 70. The shelter investments combined with the provision of secure tenure led to considerable private investment in housing on both sites. In Tondo approximately 13,500 structures were improved (90 percent of the total), 12 percent of which were replaced by entirely new houses (PCR, para. 3.04). A similar build-up process was reported for Dagat-Dagatan (PCR, para. 3.08). Development costs for an average residential lot (at approximately $2,400 equivalent) amounted to about a quarter of the cost of a minimum-standard, contractor-built, completed house in Manila, demonstrating the more efficient use of public resources in upgrading and sites and services than in conven- tional housing programs. 71. The two housing sites were provided with all elements of on-site and off-site infrastructure planned and services are fully operational, although in some sections in Tondo flooding and water shortages remain a problem.25/ 251 According to the Borrower, the incidence of periodical flooding in Tondo is not due to defects in drainage works built under the project but rather to the elevation in the sea level which occurred since then. See PCR, Annex M, Letter from the MPWH dated March 6, 1986. 23 The community facilities built include 165 classrooms for primary education, 50 classrooms for secondary education, two now health centers and upgrading of five others, and 28 multipurpose community centers, all of which are operating regularly. 72. Unquestionably, the higher levels of housing-related services sup- plied by the project vastly Improved the physical environment, health, and general welfare of over 185,000 people in Tondo and Dag-t-Dagatan, compared to their previous living conditions. Based on market values for land at Tondo and Dagat-Dagstan at project completion, the Audit concurs with the PCR (pare. 7.02) in estimating the economic rate of return on the shelter investments at 15 percent (vs. 21 percent at appraisal). The lower ERR reflects higher development costs and the fact that the project took longer to complete than expected and, consequently, benefits were deferred. 73. While considerable Improvement in the quality of housing in Manila was achieved, all sites and services units were used to meet relocation needs of squatters dislodged by public works, and thus represented no addition to the city's housing stock. This objective would be tho focus of the second project in which the bulk of the shelter investments were allocated to new housing. 74. Investments in urban transport, corresponding to approximately 37.9 percent of actual project cost ($34.8 million equivalent), increased by almost 20 km the carriage capacity of the Manila road system (through a section of 2.4 km in four lanes on highway R-10 and 5.1 km sections in two lames). Three bridges were built on these new roads and some 88 interconnected traffic signals were installed. In addition, the project equipped a central traffic control center aided by 400 mobile radio units (for traffic control by police) and created a Traffic Engineering and Management unit to administer the city's traffic system. Due to cost overruns of over 100 percent in the major roads and bridges component, the Borrower, vith the Bank's approval, deleted from the project the civil works elements originally planned under the traffic component. 75. The road access to Dagat-Dagatan area contributed to the increase by approximately 367 ha in the supply of urbanizable land in central Manila. This achievement in itself represents a major project benefit which unfortu- nately is not readily quantifiable.261 In addition, the project's urban transport investments facilitated traffic movement in and out of the Tondo Foreshore Area (including the International Port Area) and improved traffic management in the city as a whole. The PCR (paras. 7.03-7.05) estimates the economic rate of return on these Investments, based on savings in road trans- port cost, at 29 percent for traffic improvement and, tentatively, at 23 percent for major roads and bridges, since that component was not complete at the time the analysis was made. The Audit agrees with the PCR that both 26/ The total estimated urbanizable area at Dagat-Dagatan comprises 367 ha of which only a small portion (approximately 40 ha) was developed under the first urban project. 24 components have generally met appraisal objectives but lacks the appropriate data to coment on the specific ERR figures. 76. Project investments under the employment promotion components achieved mixed, if not wholly disappointing, results. Approximately 15.8 ha of saleable commercial and industrial land was developed at Tondo and Dagat- Dagatan at an est.mated cost of $8.7 million equivalent (or 9.4 parcent of total project cost). One year after the completion of this component, in December 1984, only 57 percent of the lots had been sold and demand since then has not much improved (PCR, para. 5.04). This outcome reflects the general retraction in business activity in the Philippines in the early 1980s but also suggests that investment risks for the component were unduly high due to lack of prior demand studies. Financially, the investment resulted in large losses for NRA, since expected sale revenues may not exceed one-third of the apprai- sal estimate. For the lots actually sold there is no information on whether or not enterprises are operating on them or on the number of jobs actually created. The PCR (par&. 7.02) estimates the economic return on this component at 25 percent (vs. 40 percent at appraisal) based on the market value of the land, but the Audit would expect the ERR to be somewhat lover due to pro- tracted sales and the eventual deterioration in the value of the Inventory.271 77. The second instrument used in the project to promote employment creation consisted in a line of credit extended to existing small businesses in Tondo. At project completion only 36 percent of the original allocation of $0.4 million equivalent had been disbursed under this program, loan collection rates were just over 55 percent, and evidence of employment promotion was lacking (PCR, par.. 3.15) or was ineffective under the circumstances. 78. Achievements under the technical assistance component, which accounts for approximately 6 percent of project cost ($5.4 million equiva- lent), consist primarily in strengthening NHA project implementation capacity. In addition, several special studies were completed, includings (i) municipal services for Metro Manila; (ii) Metro Manila Transport and Land Use; (Iii) fiscal base of the MMC; (iv) cadastral surnvy of HMA; (v) organization, man- agement, and financial planning for NHA; (vi) feasibility study for the second urban development project; and (vii) a National Transport Investment Plan prepared by DPWTC. In general, the studies were useful to the Borrower and led to the approval of two subsequent Bank-financed operations in the urban section (Loans 1647-PH and 1821-PH). The MMC staff training program did not materialize except for a number of ad hoc short courses (PCR, par&. 3.14). Physical Accomplishments of the Second Urban Development Prolect 79. Shelter investments under the second project emphasized new housing in Manila through sites and services and the expansion of the slum upgrading and resettlement program (SIR) to regional cities. Investments amounting to $39.5 million equivalent (or 58.3 percent of actual project cost) resulted in 271 While theoretically land values may be expected to keep pace with infla- tion, physically the inventory is likely to deteriorate due to lack of infrastructure maintenance and occupation by squatters. 25 the upgrading of approximately 7,500 housing units In Cebu, Davao und Cagayen de Oro end some 11,000 serviced lots, of which 76 percent were in Mnila and the remaining In Davao (18 percent) and Cagayan de Oro (7 percent) (PCR, Annex 4). Thee units were distributed in eight siteos a ninth site In Cebu (Carbon upgrading, 1 ha) was deleted from the project due to lend relamstion prob- ems. Substantial private investments in housing followed completion of civil works and land titling (PCR, para. 3.03). Average development cost per shel- ter unit, at approximately $2,100 equivalent, was slightly lower than under the first project. Overall, the shelter components met quantitative appraisal targets in upgrading and fell short by 11 percent In sites and services. 80. The provision of on- and off-site infrastructure at the housing sites in regional cities was completed to appraisal standards, although with large variations in work quantities due to underestimates at appraisal of land reclamation and off-site infrastructure works (PCR, para. 4.15). At Dagat- Dagatan land reclamation and on-site infrastructure works also vastly exceeded appraisal projections. 81. In general, the quality of the works completed was acceptable (PCR, pars. 6.28) even though all but one of the upgrading project areas were situa- ted on low lying, swampy coastal sites subject to tidal Inundation (PCR, pare. 3.01). The Audit notes that there was little flexibility in the choice of upgrading sites since squatters seldom have access to better terrain. Never- theless, the location of these sites remains a problem for the sustainability of the infrastructure investments since additional sea defense works will eventually be rquired to protect the sites from the gradual elevation in the sea level observed in the Philippines over the last few years. 82. The higher than average quality of civil works completed in this project, as compared to the first project, is the result of Improvements in construction supervision on the part of NHA and, in the case of regional cities, the assistance provided by resident planning and engineering advisors on site. Sanitation and basic infrastructure in upgrading areas substantially improved health conditions among resident populations, as documented in a study conducted in the main Davao upgrading ares.28/ 83. At Dagat-Dagatan, notwithstanding the physical site Improvements, the project failed to meet enviroumental standards advocated in Bank-financed operations. Land reclamation and subsequent filling resulted in the elevation of the site by several feet above the level of the surrounding built-up ares, exacerbating flooding problems in the nearby commnities (PCR, par&. 8.07 i). Project design should have anticipated this outcome and taken provisions to minimize its ill-effects. 281 Results from this study (cited in PCR, pare. 7.08) show that incidence of serious gastrointestinal diseases was reduced from 60 to 10 percent, infant mortality rates dropped from three times the national average to slightly above that average, and child malnutrition declined from 80 percent to only 9 percent, as a result of the project. 26 84. Community facilities at the housing sites were generally implemented according to the appraisal plan, except at Dagat-Dagatan where the number of primary schools was reduced by 60 percent (from 5 to 2 schools) to avoid cost overruns. Otherwise, the facilities actually built were those specified at appraisal, namelys 9 new or upgraded primary schools with approximately 160 classrooms; 5 now or upgraded secondary schools totalling 84 classrooms 10 new or upgraded health centers; 19 new multipurpose community centers; and 3 markets, all of which are presently oporational. 85. As in the first project, the second project succeedod in improving the quality of shelter for the urban poor but made only a modest quantitative contribution to the country's housing stock. Contrary to project expecta- tions, demolition of squatter housing continued unabated, adding to the demand generated by population growth and pent-up need for low-cost housing. Between 1975 and 1983 an estimated 63,000 squatter families in Manila and other Filipino cities were relocated by the Government, their number far exceeding that of families benefiting from in situ upgrading and sites and services programs carried out by NHA under Bank-financed projects. The 11,000 serviced sites built under the second project represent less than 5 percent of the country's demand from new growth alone (estimated at between 40,000 to 50,000 units per year) in the market segment served by NHA and were clearly insuffi- cient to relieve the squatting pressures in the capital region.291 86. The NHA estimated at 19 percent the economic rate of return on investments for t4e project as a whole (PCR, paras. 7.01-7.08) but unfortu- nately omitted from the report detailed analysis of returns per component. Considering the problems associated with the development of commercial and industrial properties, the Audit cannot confirm this figure and considers the ERR estimated by NHA exaggerated. For shelter components alone appraisal estimates of the ERR were 25 percent in Manila and 37 percent for upgrading and 13 percent for sites and services in regional cities. The Audit would expect actual returns to be probably lower, varying between 10 and 15 percent depending on the site, although it lacks the necessary data to replicate the appraisal analysis. 87. Project investments aimed at employment creation, in the equivalent of approximately $18.5 million equivalent (or 27.2 percent of total cost), performed poorly in this as in the first project. Commercial and industrial lots at appraisal comprised 56 ha in Manila and two regional cities. In fact, 82 ha were developed for this purpose, reflecting an increase of over 46 percent compared to the SAR figure.30/ Of the total saleable land developed (73.4 ha) 79 percent (58 ha) remained unsold at project completion. Sales 29/ See Malpezzi (1986) for several housing finance demand projections for the capital and regional cities. 301 Area increases occurred only in regional cities and may be attributed to the hope on the part of the local governments to maximize housing cross- subsidies with the profits expected from the sale of these properties and from the concomitant expansion of the municipal tax base. The changes made do not appear to have been an issue for Bank supervision missions. 27 have not Increased significontly since then and not much progress ts foreseen in the near future. As in the first project, demand analyses were never undertaken and insufficient attention has been given towards marketing (PCR, pars. 5.05). The PCR (para. 5.30) questions the wisdom of Incorporating large 8cale commercial/industrial development in future housing projects but still expects sizablo revenues to be generated from these properties on the projects being audited. This expectation appears optimistic to the Audit since long- term losses would likoly ensue from the declining value of the lnventory due to poor physical maintenance and/or occupation of the lots by squatters. 88. Similarly, the small business loans failed to meet appraisal objec- tives. Loan disbursements of approximately $0.46 million equivalent corres- pond to 21 percent of the original allocation, collections have been consis- tently low, and the employment Impact remains undocumented.311 In general, the business loans were administered by NHA and the local governments in the manner of a welfare relief fund, as suggested by the term 'livelihood' attached to the program. Government banks and small business advisory agen- cies of the Ministry of Trade and Industry never became Involved in the imple- mentation of the program, as expected at appraisal. There is no information regarding the outcome of NEDA's national small-scale business program anal- ysis, originally planned to improve performance of this program. 89. The flood control equipment, procured under the project at a cost of $2.2 million equivalent (or 3.3 percent of total project cost) and consisted of estero and sewer-drain cleaning and maiptenance equipment, assisted the Borrower in carrying out critical flood control activities, as planned. The city's flooding problem, however, has grown more severe over time. With limited resources, the MPWH has implemented several drainage improvement measures in addition to its normal maintenance activities, but a special project will be required to resolve the problem. Such a project is presently under discussion based on a comprehensive flood control program (including rehabilitation of existing works and construction of new works) prepared by MPWH with assistance from consultants financed by the project. 90. The technical assistance investments, estimated at $7.6 million equivalent or 11.2 percent of total project cost, were largely used to finance expatriate advisory services employed to strengthen the project implementation capacity of NHA and local governments. The availability of these services helped resolve several technical difficulties which arose during project execution and succeeded in building lasting technical expertise within NHA. In addition, NHA management was improved in several ways and a feasibility study for the Third Urban Development Project (Loan 1821-PH) Was completed. In regional cities the results of technical assistance were not as successful 31/ In the justification of the project the appraisal report (para. 7.08) does not quantify the expected employment impact of the project, although it states that most jobs created would benefit the urban poor. In the PCR (para. 3.08) the Borrower reports the creation of 15,000 new jobs in regional cities and incomq gains for some 9,000 households in Dagat- Dagatan as a result of the program, but these figures are not based on verifiable data. 28 because the local Institutional proble_ for exceeded the technical mandates of th expatriate advisors (PCRI, pare. 6.09-6.15). As alroedy mentioned (PPAH, pare. 89), technical assistance to MPWH resulted in the preparation of Phase I of the Metro Manila Drainage and Flood Control Master Plan. Planning assistance extended to NEDA was largely used in supporting regional cities in project planning and implementation and, with the exception of the study on business loans, achieved satisfactory results. No information Is avallable on the results of the technical assistance extended to MHS for policy planning and progrsm development. In general, the PCR (para. 8.07 j) notes, and the Audit concurs, that future technical assistance programs ought to place greater emphasis on technological transfer to local staff as well as on direct engagement of local consultants. 91. In both prolect, the credit line for buildins materials has been underutilized. Loan disbursements in relation to the original allocation were 36 percent in the first project and 21 percent in the second. Inefficient administration of this credit line by NRA and local governments accounts for the unsatisfactory performance. For the second project, the PCR (para. 3.09) notes that project beneficiaries appeared to have preferred acquiring their own materials from the private sector. Since the follow-on projects (Urban III and IV) contain similar provision, a thorough analysis of this component seems urgent to help Improve the utilization of this credit line. Financial Performance 92. The shelter components in both projects were intended at appraisal to provide both affordable housing for lower income families and a financially viable operational model that would permit NHA to expand the upgrading and sites and services program on a scale commensurable to tho PhilippineOs' hous- ing needs. Project results indicate that affordability targets have been met (PCR-I, para. 5.10 and PCR-II, para. 5.14) but financial viability goals had not been achieved (PCR-I, pares. 5.01-5.09 and PCR-II, pars. 5.02-5.11).321 93. Basically, the poor financial performance of the projects resulted from the combined impact on cost recovery of distorted pricing policies, poor sales performance, and inefficient collections. These problems were aggra- vated by the failure of the cross-subsidization scheme and by the general decline in the performance of the Philippine economy. 94. The pricina distortions are attributable to the foll --ing factors: (a) establishment of lot prices prior to completion of land acquisition and civil works, based on unreliable projections of what such costs would be; (b) modification in the form through which capital costs for utilities would be recovered, shifting these costs from the mortgage instrument to the respective 321 This section only summarizes the main financial problems encountered by both projects and the reader should refer to the cited PCR paragraphs for more detail. In particular, it should be noted that NHA is fully awaro of its cost recovery problems and has taken recently several measures that might indeed improve future financial performance. 29 utility tariffs without verifying If tariff lcvels Vero appropraste to actu- ally cover these costs, and without agreement from utility companies as to wvhther the charge. were accptable to them (a similar procedure was adopted for on-site road works); (c) application of large subsidies to the price of residential lots based on assumptions regarding profits from sales of comer- cillindustrial lots without testing the validity of the assumptions; and (d) owdssion from the pricing formula of provisions for bad debts, administrative costs and interest during construction. In addition, political conditions and Interference clearly aggravated HA's pricing proble_s. 95. Lot sales performance was poor primarily becauso project land was not acquired In advance and cumbersoms land expropriation procedures resulted In protractod land acquisition, delaying titling and mortgage origination. On land legally owned by NHA sales lagged due to mortgage documntation dolays and several other legal and practical difficulties, Including title award without concomitant mortgage signature. Accordiag to the lat data available roflecting sales as of December 1986 under the second project (PCR, Annex 4), 60 percent of the residential lots and 80 percent of tho CJI lots remain unsold.t3/ 96. Collections of mortgage lnstallaents from project beneficiaries averaged 56 percent of amounts due under the first project and are reported to be extremely poor under the second project, despite a concerted effort on ibe part of NHA to improve performanco as of 1984 (PCR-I, paras. 5.L, and 5.20 and PCR-II, para. 5.32). In regional cities where collections are the responsi- bility of local governments no roal attempt at initiating collections has yet boen made and it appears that the political will and the technical captcity to do so are lacking. 97. In addition, NMA has oncountered seemingly unsurmountable prbt'_ms in collecting payments from the government agencies on behalf of ihich tt incurrod significant expenditures by fiat without previous Interagency agree- ments. Similarly, the Authority's loans to local govornments represent large- ly uncollectable papers, although formal agreements were signed. The overall result of these various factors has been a sharp deterioration in NHA's debt service capacity which, thus far, has not been improved. 98. Tho cost recovery status under the second project gives a rough measure of the extent of the problem. On residential lots approximately 15 pereent of actual costs are expected to be recovered (vs. 75 percent estimated at appraisal) and on commercial and industrial lots the figure drops to 11 percent (compared to approximately 110 percent projected at appraisal). Similarly poor cost recovery performances are comon to *sveral public ector enterprises in the Philippines, and evidently the present difficult country conditions have aggravated the problem. However, it appears to the Audit that additional problems were created by the accumlation of housing finance and housing production functions within NBA. In the long-run, tho Audit would 33/ It must be clarified that unsoldO does not mean unoccupied lots, but rather lots which for various reasons are not yet under collections. 30 expect NHA to stand a better chance of success by opersting solely as a hous- ing production agency supported by speciallzed housain finance institutions. Institutional Performance and Policy Impact 341 99. The projects helped establih NH as sn agency specializing in slum improvement and sites and services and, in particular, strengthened the agency's technical capacity. The NU remains the only public agency in the Philippines producing housing affordable to the lowoet two-thirds of the urban population. However, because attention (of both NSA and Bank staff) was focused primarily on physical output, administrative and financial aspects of NHA's development were comparatively neglected during sost of its first decade of operations. The Philippines' financial and economic crisis of 1983/84 and the subsequent restrictions imposed on the public budget have made it diffi- cult for the Government to continue to compensate for NRA' operational losses through budgetary transfers and the issue of the Authority's financial survi- val now appears to be a main focus of concern. 100. Besides directly executing several components, NHA was expected to coordinate the activities of all the other participating agencies, including national ministries and, in the second project, also local governments. The Authority was not capable of performing this coordinating role efficiently. Unsatisfactory interagency coordination Is apparent In the difficulties NHA faced in turning over infrastructure at the housing sites to the line agencies responsible for operations and maintenance. As noted in the PCR (paras. 6.19- 6.26), the NPWH has now accepted completed works for maintenance purposes but the financial asset transfer has not taken place. The MWSS remained unwilling to accept water supply and sewer works until a most recent Cabinet decision and Presidential Directive (dated December 18, 1986) clarified the maintenance responsibilities of each project agency. In regional cities, under the second project, the line agencies were directly responsible for building their own share of infrastructure and community facilities. This arrangement averted turnover difficulties but created serious scheduling problems since works by each line agency were not completed in time to fit NHA's needs. 101. The second project placed NHA in the position of a municipal lending bank responsible for financing local governments in the amounts required by the project. At the same time, the functions of selling lots, originating, holding, and collecting mortgages were left to each municipality. Local governments were expected to use the combined resources from tax revenues and direct cost recovery from project beneficlaries to repay the NRA loans. These expectations were frustrated and local governments appear unable to repay their debts to NHA. As noted in the PCR (pares. 6.09-6.15 and 5.22-5.24), the 341 For a more comprehensive account of NHA' institutional development see PCR-I, paras. 6.01-6.17. Assessments of the institutional performance of the other agencies involved in the projects are to be found for the first project in PCR, paras. 6.18-6.20 and for the second project in PCR-II, paras. 6.01-6.26. 31 experience clearly demonstrated that joint ventures of this kLnd aro $nppro- priate In the Philippines given present local govnroment conditions.31I The issue Is, of courso, broader *ince it undermines tho viibility of the national SIR, one of the central objectives of tho second urban project. 102. Neither project addresood sector olicv Issues dlrectly. Instoad, a demonstration effect wao planned whereby the *uccessful implementation by NRA of in situ alum upgradinp' and sites and services would persuade the Government to abandon Its slum oracication and relocation practices and instoad support shelter programs modeld aftor those in the projects. The deonstration effect failed to produce the desirable policy changes. As both projects wore difficult to complete, Bank assistance focused on strengthening NRA inter- nally, but neglected the external policy environsent surrounding that agency. Thus positioned, the Bank was unable or unwilling to establish effective working relationships with the other parts of Government working In the ahel- ter sector and meaningful policy dialogue with the Borrower seem not to have taken place until the mid-1980s. Ultimately, the projects had little positive impact on sector policies throughout the decade, the Government continued to eradicate slums and relocate squatters, and increased (rather than decreased) the shore of housing budgetary appropriations benefiting higher Income groups. In this context, NHA became relatively Isolated and primarily dependent upon continued Bank support for survival. Project Sustainability 103. The longevity of the housing infrastructure assets created by the projects will depend upon routine maintenance on-site. In addition, In low- lying coastal areas, periodical upgrading of flood control works would likely be required. Turnover from NHA to other agencies of completed infrastructure for maintenance purposes has improved as of 1985 but msintenance continues to be problematic. Given the current economic difficulties facing the country, these problem cannot be expected to be resolved on the short tem. As a result, at least a portion of the assets built (as for example the unsold commercial and industrial lots) are likely to deteriorate rapidly. In the first project inadequate maintenance has contributed to low mortgage payment levels at the Tondo (PCR, paras. 5.21 and 5.23). As an alternative to accel- erated decline in asset value resulting from lnadequate maintenance, the NEDA has suggested that means be devised for involving the comuzsity itself in maintenance activities. Bank Performance 104. The Audit agrees with the PCRs in noting deficiencies In Bank per- formance during the preparation and the supervision phases of both projects. The first project was appraised prior to complotion of detailed *ngineering designs and land acquisition leading to serious implementation difficulties 35/ The fourth urban project financed by the Bank In the Philippines con- tained the same NHA-Local Government joint venture arrangements, but in view of the unsatisfactory results of the second project it is being restructured so as to assign all financial functions exclusively to NRA. 31 (PCR, para. 6.28). Its supervision, although intensive (comprising 194 staff- weeks in the field), was marred by high staff turnover, general inexperience of Bank staff and consultants,, and excessive ephasis on civil works at the expense of other project lements. In the second project, Dank assaitance in project preparation Improved but *upervision problms recurred, vith the additional logistic difficulty of combining coverage of project implementation In regional cities with the preparation of follow-on operations in Manila (PCR, paras. 6.31-6.32). In contrast, the remedial financial measurec afforded by the inclusion of both projects in the Bank's 1983-84 Special Assistance Program greatly facilitated the completion of both projects. Lessons Learned 105. The experience of Urban I end II has been valuable to the Borrower and the Bank. The useful operational lessons learned are clearly stated in both PCRs and may be sunmmarised here as followss36I (i) as a demonstration project the Tondo upgrading was overambitious and overwhelmed the implementation capacity of the Borrower and of local contractors* However, an alternative smaller target area wuld probably not have been politically acceptable in view of the high priority assigned by the Borrower to improvements in the Foreshore area of the capital; (ii) a two year interval between the inception of the first and the second projects proved insufficient to consolidate NHAs operational capacity and at the same time expand the program on a national scale. The repetition of errors might have been avoided by postpon- ing the follow-on operations until substantial results were achieved under the first project;271 (iii) belated land acquisition and cumbersome expropriation procedures led to serious implementation difficulties in both projects, reemphsisz- ing the importance of land ownership as a necessary condition to project success; (iv) the availability of detailed engineering designs for civil works under the second project, combined with on-site advisory services and more efficient construction supervision by NHA staff, contri- buted to reduce the technical implementation problems of that proj- ect as compared to the first; nevertheless, insufficient attention 361 For the first project see PCR, para. 8.10 and for the second project see PCR, paras. 5.28-5.32. 37/ After the Second Urban Development Project in 1978 Bank loans to the Philippines supporting urban sector programs were approved for the 'Regional Cities Development Project,' in 1980 (for $72.0 million equiva- lent) and for the 'Municipal Development Project,t in 1984 (for $40.0 million equivalent), raising total Bank financing for the urban sector in the Philippines to $250.6 million equivalent in a ten year period. 33 continued to be paid to non-civil worki elemnts In project prepara- tion and exzcution, a deficiency that ought to be correc4ed in on- going and futur* projects; (v) le than successful institutional arrangemnt for project Implemen- tation revealed unreallstic apprdisal expectations regarding the interagency coordination capacity of NU and the level of financial managgemnt and policy coherence required by the projects at both the national and the local levels. Financial and Institutional Issues are presently being addressed on a sectorwide context, but satisfac- tory solutions have yet to bo devised. 106. Other points of special interest highlighted in the PCRs Include the need for explicit policies towards renters (who comprise a large part of the population In upgrading sites) and various pricing measures designed to Improve cost recovery. 107. At the soctor level the Audit reiterates the adverse effect on project performance caused by lack of an effective policy dialogue betwen the Borrower and the Bank. In addition, by not slowing down the lending pace in face of unsatisfactory performance under the first project, the Bank might have in fact contributed to delaying the necessary sector reforms. 34 ATTACHMENT A SORROWVR'S COMMENTS The draft PPAR was forwarded to the Borrower for comments. The response is reproduced in Attachment A to the report. The summary of the comments is as follows: (1) It Is observed that while now new lessons were presentod, the short- comings of the project emphasise the need for better project preparation and, in particular, more careful consideration of local situation *nd culture. (ii) The report provides inadequate focus of the social benefits of the project which parallelled the physical achievements. Note: While these achievements are noted in general terms throughout the report, available data did not per-it measurements in the degree of Improve- ment in the standard of living of the affected population. (iii) The unsatisfactory cost recovery is noted and the insufficient analysis of the risks associated with the cross-subsidization assumptions is correctly noted. Adequate considerations of affordability is also questioned. (iv) It is suggested that community participation in the maintenance of facilities be considered. Note: The report notes (PPAM, para. 103) that such action was under con- sideration by NEDA. 35 ATTACHMET B REPUBUC OF THE PHILIFINES Page 1 of 2 NATIONAL ECONOMIC AND DEVELOPMENT AUTHORITY NEDA sa Pasi&, Amber Avenue Pasig, Metro Manila CA Addrbs: KEDAPHIL P.O. Box 419, Oreenb T.k. 6.0.31 to 50 1 Decesber 1987 Mr. Alexander Nowicki Chief, Policy-based Lending Industry, Public Utilities and Urban Sectors Operation Evaluation Department World Bank 1818 H Street, N.W. Washington, D.C., 20433 U.S.A. Subject: Project Performance Audit Report: Manila Urban Development Project (Loans 1272-PH/1282-PH) and Second Urban Development Project (Loan 1647-PH) Dear Mr. Nowicki: Thank you for sending us a copy of the abovecaptioned report and for allowing us to give our camments. While we concur in general with the findings of the PPAR, we have a few comments on the report as follows: On Implementation Performance Overall, the report offers no new lessons but reiterates the same lessons learned from the implementation experience of past projects, i.e., the need for thorough and adequate project prepara- tion especially in the case of the subject project, which was experi- mental in nature and the need for the executing agency to have adequate institutional capability to carry out the project. It appears high time for both the government and the Bank to place emphasis on these aspects in future projects. In addition, to the 'technical'factors that accounted for the poor implementatior. performance of the project, we note that the experimental nature of the project, as well as the sensitive issues involved, e.g., relocation of squatters, necessitate a more careful consideration of local situations and culture. On Project Results We note with disappointment the report's lack of focus on the project's socio-economic results. It would have been more worth- while if the report considered not only the achivement of physical 36 ATTACHMENT A Page 2 of 2 target but those of socio-economic outcomes of the project, particularly on addressing inequalities in the distribution of urban services in support of the urban poor, improvement in the quality of life and effects on beneficiaries' attitudes and behaviors. Relatedly, the issue of cost recovery should be seriously looked into. The report mentions the lack of political will and technical capacity of local governments for collection of mortgage installments. While the latter may be true, afford- ability by target beneficiaries (e.g., urban poor) should also be considered. It is noted that the targetted cross-subsidy assump- tions made at appraisal did not materialize on account of low sales of commercial and industrial lots. The risks involved with cross-subsidies should have been foreseen at appraisal, and the level of anticipated risk should have been incorporated in rate of return analysis. It would also be helpful if estimates of financial internal rates of return could be presented in the final report. On Maintenance of Project Assets We agree with the Bank's observation on the need to expedite operation and maintenance arrangements for the project to sustain the positive results that have been generated. In this respect, maximum participation by the beneficiary commmities in operation and maintenance should be explored. We hope that our comments will be useful in the finalization of the PPAR. Very truly yours, SOLITA COLLAS-MONSOD Secretary of Socio-Economic Planning & Director-General THE WORLD BANKIINTERNATIONAL FNANCE CORPORATION 37 OFFICE MEMORANDUM EAST ASIA & PACIFIC REGIONAL OFFICE PCR DATE December 31, 1986 TO Mr. D. Ahmad, Acting Chief, AEPUW FROM D. Haldane, AE f EXTENSION 61931 SUBJFCT PHILIPPINES: Second Urban Development Project Project Completion Report (PCR) 1. Attached is the PCR on the above project prepared by the National Housing Authority (NHA). This memo supplements the PCR in order to: (a) highlightsome of the more important issues and lessons of Urban II and add the Bank's perspective; (b) cover points omitted from NHA's report; and (c) update the reader on the status of recommendations made in the PCR. 2. The attached report attempts to minimize duplication of material covered under the PCR on Urban I issued last April 11, 1996, on which little, if any, change has occurred. It -focuses therefore mainly on features specific to Urban II or requiring follow up in the Urban I PCR. The environmental context and institutional assessments of the NHA's first ten years will be found in the Urban I PCR. Thus for a full picture of the NHA's Bank-assisted program, both reports should be taken together. 3. The central new feature of Urban II was the introduction of the agency's first regional operations, and one of the project's most important lessons related to NHA's unsatisfactory experience with joint ventures with the local governments. Recommendations for revised institutional arrangements for future national shelter development have played an important part in the shelter sector rationalization, the first stage of which was recently signed into law by President Aquino. Thus, despite serious problems encountered under Urban II with regionally dispersed operations, the project experience has usefully contributed to what will hopefully be a more rational and efficient shelter delivery network in future. Key Issues Issue I: Institutional Arrangements 4. The project's financial arrangements placed NHA in the de-facto position of a municipal lending bank, with NHA making loans to local governments which originate, hold and collecting on the mortgage assets built. Local governments, however, proved to be unstable partners, with Cagayan withdrawing from the project shortly after local elections installed a new mayor. Much time was lost to rebuilding the project's momentum. In addition, sites were turned over to local governments following physical completion, making cities responsible for sales and collections. A combination of poor administration 38 and lack of political will has resulted in negligible cost reco,very on the regional component as outlined in para 6 below. 5. This experience helped crystalize the decision that local governments represented an inappropriate institutional base for the management of financial instruments such as mortgages. Particularly so for serviced sites where participation is voluntary and eviction for non payment both possible and necessary. The question is less clear for upgrading where participation is largely but not entirely voluntary, and eviction extremely difficult. While many mortgages in upgraded areas are viable, a proportion are not. The appropriate institutional home for such relatively uncollectible paper is still under discussion. However, recommendations that NHA disengage from joint financial (mortgage) ventures with local governments have been implemented, and the Bank's Urban IV project, which contained such provisions, is being restructured to place 100% of the financial responsibility with NHA. Local governments continue to play important roles in nonfinancial matters. Issue II: NHA's Debt Crisis/(Cost Recovery/Sales) 6. Collections were just underway at the time of the Urban I PCR, and only a year has passed since then. However, further work has been done on the combined (compounded) impact of deficiencies in the Authority's pricing structure, sales performance and collection efficiency. Tables I and II present the compounded effect of these three problems on overall cost recovery. Actual recovery on residential units at the time of writing is about 15% vs. 75% estimated at appraisal. For commercial/industrial units the comparable figure is 11% vs roughly 110% estimated at appraisal. A significant and unforeseen proportion of this gap is due to slow sales, meaning document signature which enables startup of repayment. At appraisal, immediate sale of all lots was anticipated. However, in reality titles appear to have been awarded in large politically motivated events without simultaneous signature of the mortgage document. Thus many more lots are reported as awarded than are "sold" in the sense of signed and under collection. Sales in upgraded areas are particularly troubled with only 17% of completed units sold as of June, 1986. Obtaining signature is difficult on completed and occupied upgrading sites and evicting occupants who have not yet signed is difficult. For serviced sites, only 56% of completed units are sold (under collection) but 68% are reported as allocated. It is not clear if the remaining one third are illegally occupied. It appears both NHA and the cities lack the political will to enforce eviction in serviced site areas on unsigned units. (Table III summarizes sales as of June, 1986.) A diagnostic study is needed to check this data, pinpoint the bottlenecks and lay out a program for resolving NHA's sales problem. 7. This lethal combination of institutional weaknesses triggered NHA's debt crisis which threatens the institutions very survival. The Authority's debt service ratio was 0.25 in 1985 and has further deteriorated in 1986. Sales of existing assets and annual budget allocations are helping bridge the gap until new policies and procedures can be put in place to rectify the situation. 8. The Authority is aware of these problems, and some progress is being made in resolving them. Specific recommendations and their status are listed below: 39 With regard to pricing: (a) costs of both on-site and off-site water, sewerage and roads infrastructure previously charged off to other agencies (subsequently not accepted under both Urban I and II) will in future be charged to beneficiaries and included in the mortgage unless a signed contractual agreement exists stipulating the amount and terms of payment to NHA by another agency. In the latter case, NHA will function as a contractor to the owner agency. As a result, unit costs could double in future projects. The MWSS indicates that it has refused to accept water and sewerage works because it's pricing structure is inadequate to cover the associated costs. Thus charging beneficiaries would not result in double charging. (b) construction in future will not begin until NHA holds title to a property and therefore knows the acquisition price; (c) administrative costs and interest during construction will be added to the sale price, both of which were omitted from Urban I and II; and (d) a 20% "default surcharge" has been added to the base cost. The new pricing policy is already being implemented in the Pasig site of Urban III. With regard to sales: (a) in future it is proposed that NHA function as a developer, selling its mortgages to a financial institution. Thus sales must be complete before NHA can obtain payment for its mortgage. This should create a strong incentive to execute sales immediately upon completion, if not before. Since land title will be in hand before construction, lack of title would no longer delay sales; (b) a system of pre-sales combined with downpayments is to be tried out under which lottery winners will begin savings accounts to build a downpayment (10% agreed; 20% recommended by the Bank). Deposits of those unable to meet the downpayment requirement after the allowed period will be refunded together with earned interest, or the application may be extended until the next lottery; (b) the practice of title award without simultaneous document signature must be immediately terminated. With regard to collection efficiency: (a) as noted above, NHA will no longer collect under future arrangements. Collections will be handled by a financial institution with a private Bank interface between the public agency holding the mortgage and the individual mortgagee. However, it is possible that as long as public agencies hold loan assets, payment may be poor in view of the relaxed payment culture existing in the Philippines which is well described in the attached PCR. Much depends on the financial agency's 40 administrative efficiency and willingness to enforce sanctions; and (b) loan loss reserves must be established by all institutions holding uninsured mortgages. COA has concurred in this recommendation which is to be implemented in 1987 for NHMFC and 1986 for NHA. 9. If all the above recommendations are implemented, cost recovery on residential units could theoretically rise from the present 15% to 103%, as shown in Table I assuming 95% sales and 90% collection efficiency. The pricing changes are already in effect, and the Authority is in the process of shifting to a policy of not initiating construction until land has been acquired and the price is therefore known. These two policy shifts represent fundamental and quite difficult changes. However, tightening sales procedures, enforcing sanctions and cleaning up the unsold (undocumented) inventory backlog represent several of NHA's most compelling and difficult tasks, particularly under the present political climate. The task is complicated in Urban II by the local governments' ownership of the sites in Davao and Cebu. These may simply have to be written off to experience. The institution's very survival will depend upon its success in addressing these procedural and financial issues. Issue III: Turnover 10. The Urban I PCR noted NHA's difficulties in turning over infrastructure completed under both projects to MWSS, MPWH and others. As noted in this PCR, the MPWH has now accepted all completed works for maintenance purposes, though the financial asset transfer has not taken place. The MWSS, however, remains unwilling to accept works and the legal documents under both Urban I and II are vague on this important point. As a result, NHA remains at least formally responsible for servicing the associated debt. In future, legal documents should clearly stipulate which agency ultimately will own the asset built and bear the associated debt burden. Due to this gap in project documentation, it was necessary to obtain a Cabinet approval and Presidential directive, signed December 18th, clarifying maintenance responsibilities and directing MWSS and others to accept completed infrastructure. Though the matter is therefore in principle resolved, the directive has yet to be implemented and will doubtless take time to execute. 11. Under the regional cities component of Urban II, an experimental approach was tried out with MPWH, MWSS and others directly building their own off-site infrastructure. While this successfully averted the turnover problem, it predictably generated difficulties of coordination and timing of these works to dovetail with NHA's needs. Until sites are functional, they cannot be occupied and collections started. Having now tried both approaches, the NHA has concluded that self-contained projects are preferable, and that the extra costs must, as noted above, be added to the mortgage. Standards will be trimmed back to meet affordability constraints. Issue IV: Cross-Subsidies 12. Both Urban I and II entailed substantial cross-subsidies, utilizing projected profits from commercial/industrial (CI) lots to subsidize prices on the residential units. These proved unworkable in both projects, and have contributed significantly to NHA's current debt service crisis. The essential 41 weakness in the cross-subsidy theory is that it ignores business cycles, and differential demand for residential and commercial/industrial properties. A steady demand curve was assumed at appraisal lasting through completion, a period anticipated at four years and spanning in reality 7 years (Table IV). The worldwide economic recession of the 1980-81 period, followed by the 1982- 83 crisis flattened demand for CI property, and eroded affordability on residential units. Demand moved downward toward cheaper units such as NHA was building. Accompanying high inflation in 1984 and 1985 as sites were completed depressed economic expansion, but stimulated demand for inexpensive residential real estate as a classic store of value in inflationary periods. Some sales also occurred in CI property, but this represents a far less liquid, longer-term investment prospect. Thus NHA encountered heavy demand for the subsidized residential units (many of which are occupied though not documented) and quite depressed demand for the Cl units which were to provide the requisite profit. The Authority's inventory value will doubtless keep pace with inflation if it can afford to hold the investment until market conditions improve, but inventory contributes nothing to NHA's acute cash needs for debt servicing. 13. Commercial/industrial properties in the regional cities are a particular problem, with only about 5% of the existing inventory sold. As a result, proposed CI investments under Urban IV have been deferred until the existing inventory is reduced, and demand picks up. 14. Experience under both Urban I and II has therefore forced NHA to the conclusion that internal cross-subsidies are excessively risky, and cannot be relied upon to provide cash when needed. Borrowings secured against these and other properties have reached a maximum prudent limit. The business environment on which these arrangements rely is too cyclic, volatile, and unpredictable for NHA to risk its institutional viability. The Bank fully concurs in NHA's judgement. It is in fact, our view that the private sector should undertake commercial/industrial property development, and that NHA should disengage from this type of activity. The new Executive Order on rationalization of the housing sector appears to approve this change, but the language is not explicit and a clarification is needed. Issue V: Business Loans 15. Urban II included a small business loan (SBL) component which was to be implemented by a Bank, though firm arrangements had not been agreed at the time of appraisal and were left to be worked out during implementation. In actuality, the two banks identified at appraisal declined the role and NHA ultimately implemented the component. Experience with SBL's has been exceptionally poor. Only a third of the amount earmarked was taken up, and collections total only 20% of the amount due. Though NHA is committed to the program which is politically popular, business lending has not been professionally managed and has become blurred with the Authority's social objectives. Ultimately the success of any business lending program must be judged by whether the funds are recycled into a replicable program or become a grant to the initial loan recipient. By this criteria the component must be judged a failure. The Bank has recommended to Government that a specialized institution be named into which such programs could be transferred, and that NHA terminate further business lending. The new Executive Order concurs in this mandate directive. 42 Issue VI: Resettlement 16. Urban I and II focused largely on establishing a national upgrading program which, as noted in the NHA's PCR, brought about an important modification in a 30 year old Government policy of squatter relocation. It did not, however, terminate relocation which has continued to dominate an excessive amount of NHA's institutional capacity. Of the approximately 133,000 units built by NHA in its first ten years, only 6,000 to 10,000 (or a maximum of about 7%) have contributed to expanding the housing stock. All of these were under Urban II and III. The 10,000 new low cost units served less than 2% of the demand from new growth alone estimated at between 60,000 to 80,000 units per year in the market segment serviced exclusively by NHA and the informal sector. Thus squatting pressures in fact intensified over the decade. Production of new housing stock must be significantly expanded. While the NHA's new six year program should reflect this shift, private sector participation is ultimately the only means capable of delivering the volume needed, and the Housing Sector Loan is intended to help develop the necessary mechanisms to achieve this goal. Issue VII: Flood Control 17. The PCR correctly points out that the project's modest contribution of additional equipment and TA did not resolve the flooding problems of Manila which have, in fact, grown more severe. A comprehensive review of the causes of flooding in Manila is needed along with a project focused on resolving these problems. Such a project is under discussion between the Government and the Bank. A separate PCR report prepared by MPWH is available on file detailing results of the flood control component. Issue VIII: Land Acquisition 18. Land costs were significantly underestimated at appraisal, and land acquisition problems plagued the project throughout implementation. Appraisal estimates were based on the assumption that the prices frozen under the Areas for Priority Development (APD) Order would prevail. In actuality, expropriation took years and in many cases is still unresolved. Despite protracted litigation in the courts, prices were finally set far closer to market than expected. This enabled NHA to get on with s 4%division and titling. For years lack of land title prevented finalization of t sales as a Deed of Sale with Mortgage was being utilized requiring title. Thus lack of title resulted in NHA's inability to document the sale and initiate collections. In 1984 the NHA shifted to a Conditional Contract of Sale in which title was not required, thus removing legal obstacles to completing sales and starting collections. A further and more important policy shift is also taking place under which the Authority will now undertake construction of a site only after title is acquired and the final price known. Adjusting to this new policy will be quite difficult, with several years of intensive land acquisition and little construction, but over the longer term is expected to yield substantial dividends. Issue IX: Disbursements 19. A learning curve is clearly evident in the contrast between the Urban I 43 disbursement profile and the Urban II profile (Chart V). The first project closely paralleled the overall urban profile, but the second significantly improved on it, cutting over two years from the profile. The fact the project did not achieve the highly optimistic appraisal schedule (of substantively 3 years) should not be allowed to detract from this Improvement. Issue X: Economic Evaluation 20. The 19% return noted in NHA's revaluation is probably quite optimistic and suffers from some methodological constraints inherent in the nature of the instrument rather than the method or data utilized by NHA. The evaluation utilizes "market" values of commercial/industrial property based on a few sporadic sales in an otherwise highly depressed market in which little movement was occurring. As noted above, even the Authority's own CI property remains largely unsold to date. The market's general suppression may have been partly mitigated by demand generated by speculative buyers seeking inflation hedges during this period, especially for residential properties. However, the reliability of the data as a true judge of the market may be arguable, and the economic rate of return similarly affected. Issue XI: NHA's Financial Management 21. The Authority's financial management and performance remains weak -- a protracted problem which virtually endangers the agency's survival. Loans under Urban III will severely strain NHA's debt service capacity and dramatic steps will be needed to rectify the pricing, sales and collections problems now depressing NHA's cash generation if the Authority is to regain its capacity to again borrow and repay external loans. 22. In conclusion, the project achieved its key objective of nationalizing the upgrading and serviced site program. However the institutional model utilized (NHA loans to local governments) proved unreplicable. The projectbs failures (notably the joint financial ventures with local governments, and the performance of small business loans, home material loans and cross subsidies have contributed to both the NHA's and the Bank's understanding of the sectoral institutions strengths and weaknesses, and key issues leading to the development of the Housing Sector Rationalization Plan, the first stage of which was approved in December, 1986. It is hoped that with the adoption of a significant proportion of the policy agenda under this Plan, many of the problems encountered under both Urban I and II will be rectified, and the sector placed on a sounder policy and financial footing to achieve its longer term objectives. 44 TabLes I/EI PHIUPPINES SECOND URBAN DEVELOPMENT PROJECT Cost Recvey Appaial Estiabe vs. Actual RESIDETI UNTS 200 180- 140- 120- 100 - Jso - mmmin Raskientk Unft SAR 60 - 40 _i ti,3 20 r ____-1-% CR Cost Pcin Scoes Co CR Ersiou ftom Prcing Sales & Coil COMMERCLAL/INDUSTRIAL UNITS 200 1160- 140- 120-- Cmm m mvnm m m u n j. w 40- 20 codt Pricino SlsCoilecorm cR Erosion from Pric sing es & conl ~bIld Bonk-31018&2 TABLE III Tl[E 1: S
Groupe de la Banque mondiale · Project Performance Assessment Report
Philippines - First and Second Manila Urban Development Projects
Voir le document original
Le texte intégral est hébergé par l’organisation qui le publie. lawenc.com indexe les métadonnées et renvoie vers la source officielle.
Texte intégral
Informations clés
Organisation
Groupe de la Banque mondiale
Type de document
Project Performance Assessment Report
Pays
Philippines
Source
Banque mondiale