Document of The World Bank

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Energy & Infrastructure Unit outh Asia Region Document of The World Bank IMPLEMENTATION COMPLETION REPORT (CL-44780) ON A LOAN IN THE AMOUNT OF U$ 105 MILLION TO INDIA FOR THE ECOND TAMIL NADU URBAN DEVELOPMENT PROJECT Report No: 32456

2 CURRENCY EQUIVALENT (Exchange Rate Effective November 30, 2004) Currency Unit = Rs. 1.0 = U$ 1 = FICAL YEAR April 1 March 31 Indian Rupee (Rs.) U$ Rs. ABBREVIATION AND ACRONYM BPL BT CA CA CBO CCP CMC CMDA DEA DMAW DTCP EC ERR ER FIs GF GI GO GOI GoTN HDFC HUDCO ICB ICICI ID IDC LACI MUDF NCB NGOs NLC PMR PMU PAD FC TA Below Poverty Line Board of Trustees Constitutional Amendment Country Assistance trategy Community Based Organizations City Corporate Plan Chennai Metropolitan Corporation Chennai Metropolitan Development Athority Department of Economic Affairs Department of Municipal Administration and Water upply Directorate of Town and Country Planning of GoTN Empowered Committee Economic Rate of Return Environmental and ocial Report Financial Institutions Grant Fund Geographical Information ystem Government Order Government of India Government of Tamil Nadu Housing Development Finance Corporation Housing and Urban Development Corporation International Competitive Bidding ICICI Limited Institutional Development Institutional Development Component Loan Administration Change Initiative Municipal Urban Development Fund National Competitive Bidding Non-Government Organizations Non-Lending Category Project Management Report Project Management Unit Project Appraisal Document tate Finance Commission Technical Assistance

3 TNUDF TNUDP TNUIFL TNULB TNIU TNHB TNCB TNUITC UIC UIR ULB W Tamil Nadu Urban Development Fund Tamil Nadu Urban Development Project Tamil Nadu Urban Infrastructure Finance ervices Ltd. Tamil Nadu Urban Local Bodies Tamil Nadu Institute of Urban tudies Tamil Nadu Housing Board Tamil Nadu lum Clearance Board Tamil Nadu Urban Infrastructure Trust Company Urban Investment Component Urban Infrastructure ervices Review Urban Local Bodies Water upply and anitation Vice President: Country Director ector Manager Task Team Leader: Praful C. Patel Michael F. Carter onia Hammam K. Mukundan

4 INDIA econd Tamil Nadu Urban Development Project CONTENT 1. Project Data 2. Principal Performance Ratings 3. Assessment of Development Objective and Design, and of Quality at Entry 4. Achievement of Objective and Outputs 5. Major Factors Affecting Implementation and Outcome 6. ustainability 7. Bank and Borrower Performance 8. Lessons Learned 9. Partner Comments 10. Additional Information Annex 1. Key Performance Indicators/Log Frame Matrix Annex 2. Project Costs and Financing Annex 3. Economic Costs and Benefits Annex 4. Bank Inputs Annex 5. Ratings for Achievement of Objectives/Outputs of Components Annex 6. Ratings of Bank and Borrower Performance Annex 7. List of upporting Documents Annex 8. Additional Information Page No.

5 Project ID: P Team Leader: Patricia Clarke Annez Project Name: TN URBAN DEV II TL Unit: TUDUR ICR Type: Core ICR Report Date: May 31, Project Data Name: TN URBAN DEV II L/C/TF Number: CL Country/Department: INDIA Region: outh Asia Regional Office ector/subsector: General water, sanitation and flood protection sector (65%); ub-national government administration (18%); Roads and highways (17%) Theme: Other urban development (P); Municipal governance and institution building (P); Municipal finance (P); Pollution management and environmental health (P) KEY DATE Original Revised/Actual PCD: 03/02/1998 Effective: 10/13/ /13/1999 Appraisal: 00/00/0000 MTR: 01/12/2001 Approval: 05/27/1999 Closing: 11/30/ /30/2004 Borrower/Implementing Agency: Other Partners: GOI/TNUDF/IDC TAFF Current At Appraisal Vice President: Praful C. Patel Mieko Nishimizu Country Director: Michael F. Carter Edwin R. Lim ector Director: Vincent Gouarne Frannie A. Leautier Team Leader at ICR: K. Mukundan Hiroaki uzuki ICR Primary Author: Patricia Annez 2. Principal Performance Ratings (H=Highly atisfactory, =atisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, U=ubstantial, M=Modest, N=Negligible) Outcome: ustainability: Institutional Development Impact: Bank Performance: Borrower Performance: L U QAG (if available) Quality at Entry: Project at Risk at Any Time: No ICR

6 3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The project development objective is to improve urban infrastructure services in Tamil Nadu in a sustainable manner. This objective is to be achieved through: (i)strengthening the managerial, financial and technical capacities of Urban Local Bodies (ULBs), through an institutional development program, in line with the urban sector reforms which the GoTN is implementing following the passage of the 74th Constitutional Amendment; (ii) mobilizing resources for basic urban infrastructure investments costing about U$ 183 million (water supply, sewerage and sanitation, solid waste management, roads, transports, storm water drains and street lighting, etc.); and (iii) ecuring sustainable funding sources for the urban infrastructure investment, through TNUDF and municipal bond issuance, beyond the Bank's line of credit operations. 3.2 Revised Objective: Objectives not revised. 3.3 Original Components: The project consists of the following parts : Part A : ub-loans (Line of Credit Component) The financing of ub-project through provisions of ub-loans to ub-borrowers Part B : Integrated anitation Program 1. The financing of chemes through provision of Capital Grants to Beneficiaries 2. Assisting PMU and ULBs in the preparation and implementation of chemes through the provision of consultants' services Part C : Institutional Development 1. The financing of Activities through the provision of Grants to Recipients. 2. trengthening the managerial, financial, technical and procurement capacity of ULBs, including the preparation and implementation of their City Corporate Plan, carrying out studies to promote the sustainability of institutional reforms, preparation of operational manuals, through the provision of equipment, materials, training and technical assistance. 3. trengthening the institutional capacity of DMAW and PMU through the provision of staff training, equipment and materials. 4. trengthening the capacity of TNUDF and selected ULBs to raise funds from the domestic capital market through, inter alia, provision of financial advisory services. 5.. Establishment and operation of the Project Management Unit. 3.4 Revised Components: Component; Cost; Rating UB-LOAN UNDER THE URBAN INVETMENT COMPONENT; $80,000,000.00; INTEGRATED ANITATION PROGRAM; $5,000,000.00; IDC - PROVIION OF GRANT; $5,000,000.00; IDC - TRENGTHENING ULB & CCP; $10,000,000.00; IDC - TRENGTHENING DMAW & PMU; $250,000.00; IDC - TRENTH. TNUDF & FIN ADV ULB; $500,000.00; NR PMU OPERATION COT; $1,000,000.00; FEE; $1,000,000.00; - 2 -

7 UNALLOCATED; $2,200,000.00; NR 3.5 Quality at Entry: 3.5.1).The project is rated satisfactory in terms of development objective, design, and quality at entry, primarily due to the strength of the design of the line of credit component. Quality at entry is more questionable for the other components, based on documentation available from the appraisal. The institutional development component seems to have been reworked into a more focussed program than that specified at appraisal ).The CA of 1997 identified both enhancing city management and municipal services as well as supporting state urban sector reforms that implement the 73rd and 74th Constitutional amendments as its objectives in the urban sector. This project DO's were broadly consistent with the first CA objective, but project design focused more on providing finance and TA to improve services than on promoting fundamental tate reform, as discussed below in paras )One key thrust of this project with respect to the development objectives was to provide continued support to the Tamil Nadu Development Fund (TNUDF). This fund had been created in November 1996 as part of the restructuring of the Tamil Nadu Urban project )The restructuring involved creating a new institution that was majority owned by Government, but also owned by three well-established Indian financial institutions all of which have private equity ownership, the ICICI, the HDFC, and the ILF. By the end of this project, their joint share ownership in the TNUDF amounted to 30% of total. Day to day management of the fund was delegated to an asset management company, the TUIFL, which was rewarded on the basis of disbursements, and which was majority owned by the three participating financial institutions (PFIs). Overall oversight of the fund was provided by a trust company, the TUIFC, which operated as a board to the TNUDF, with representation of all major shareholders. This somewhat complicated structure had been used before, particularly in cases when government entities are shifting operations from government to an autonomous agency. However, the TNUDF was the first public-private joint financial intermediary specializing in municipal financing without tate government guarantees. By supporting a public-private partnership of this kind for mobilizing resources to ULBs, the project sought to build financial discipline by imposing rigorous financial and technical assessment of projects ) From the time of restructuring of the TNUDP I project to project appraisal for TNUDP II, the TNUDF had operated effectively, hence a continued support to this institution was a reasonable approach. As an autonomous agency, the TNUDF had been able to process loans more rapidly than its government predecessor. By the time of writing the ICR for the first Tamil Nadu Urban Development Project in March 1998, a little over a year after its creation, the TNUDF had approved loans worth about Rs. 1,500 M. This compares to approvals of Rs M in the eight years of operation of its predecessor, the Municipal Urban Development Fund of Tamil Nadu (MUDF). At the same time, the TNUDF had maintained the strong performance on loan recovery already achieved by the MUDF, and thus continued to play an important role in establishing principles of financial discipline for investment projects undertaken by ULBs ) The project design included covenants to increase the resources mobilized from the private sector for infrastructure projects in ULBs. Expectations of private interest in lending to ULBs turned out to be optimistic in some respects. However, this optimism reflected a view broadly shared throughout India at the time, and as such, the project design was not ahead of general market perceptions. Thus, on the supply side of funds, the project design took a reasonable approach for mobilizing resources for funding projects in Tamil Nadu s ULBs ) Regarding the demand side for funding, i.e.ulb finances and autonomy, the project focused on institution building in municipalities, supported primarily through preparation of City Corporate Plans, GI studies, preparation of manuals for engineering and accounting. This component was not specified in much detail in the appraisal documents, but the component evolved to include support for introduction of accrual accounting and computerization, thus supporting greater transparency and rigor in municipal accounts, consistent with the 74th - 3 -

8 CAA s intent to devolve greater resources and responsibilities to ULBs ) Beside supporting the continued development of the TNUDF, the project also sought support to the reform process through two studies: (1) a management audit of selected ULBs to develop a phased program of institutional reform on staffing and personnel policies and streamlining of procedures; and (2) a second study to clarify the roles between ULBs and line agencies with direct responsibility for urban service delivery in municipal jurisdictions. In addition, the passage of a revised municipal act, the Tamil Nadu Urban Local Bodies Bill (TNULB) in November 1998, was considered a significant government commitment to reform that would assist in meeting project objectives. Project design did not take into account the possibility that the TNULB would never be implemented, as has turned out to be the case to date ) Based on the materials prepared before project approval, the project components focusing on the demand side received less attention, were less clearly specified, and involved less substantive dialogue with the GoTN than did the arrangements regarding the supply side of funds. The process of discussion and follow up on the two studies to be supported under the project to link them to ongoing project implementation, and to spur consideration of broader policy issues affecting the project was included only in the PIP without much prominence and specificity ) The risk that the tate would not sustain the reform process consistent with the intent of the 74th CAA was not identified at appraisal. The key risk was identified as ownership of reforms at the ULB level, while the role of the tate, which has proven significant during implementation of the project was not mentioned. A more thorough diagnosis of the constraints to ULB resource mobilization as well as a clearer understanding of the role of the tate in decision making pertaining to the approval and implementation of investment projects in ULBs would most probably have resulted in better project design and execution ) Experience in implementation also pointed out an important flaw in the design for pricing TNUDF s loans. The lagged pricing structure, high and inflexible mark ups, and the lack of any prepayment penalty were more consistent with a government-sponsored monopolist development bank model than the more market driven conception for the TNUDF envisaged under the project. As it turned out, Indian capital markets went through dramatic changes over the project period, and these changes penalized the TNUDF. (ee also section 4.5 on Institutional ustainability) ) The IP component also supported the objective of mobilizing resources for urban infrastructure investments. This component was targeted directly at poor communities, and was executed by the tate government through the PMU. This intervention supported and enhanced an ongoing tate program. The IP design focused on improving maintenance of municipally-sponsored public latrines. This objective was to be achieved through mobilization of community groups to maintain latrines, with penalties (loss of the grant financing) to those ULBs that did not provide support to community groups ) The project included a very high subsidy component for capital costs to direct beneficiaries (100%).The financing plan included a 50% contribution from ULBs, but there is little evidence of analysis of ULB capacity to mobilize the contribution, nor any specifics on timing and enforcement of collection of the ULB contribution. In practice, the non-bank portion of project costs was usually provided by a GOI sponsored program, VAMBAY. More thorough examination of the feasibility of the proposed financing might have revealed this extent to which this sub-component, as conceived by the borrower, was indeed a tate-sponsored top down initiative, fully financed directly or indirectly through tate, not ULB, resources ) The appraisal documents do not indicate that the design and costing of the units built under the IP reflect the results of a least cost analysis and attempts to use cost-effective designs. This may have been due to the emphasis on a participatory model for project design. However, a participatory approach is unlikely to do much to manage capital costs when beneficiaries are given a 100% capital subsidy ) Experience in implementation of the IP indicates a disconnect between the participatory, demand- driven concept outlined in the project documentation and the GoTN s plans for the project. GoTN planned to provide a - 4 -

9 specific allotment of IP units was to each municipality and town panchayat as part of IP execution. The GoTN s concept was to target supply of units to communities and use CBOs to implement projects thus chosen. In some cases when a CBO could not be found, municipal councilors took on this role. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: DO 1. trengthening managerial, financial, and technical capacities of ULBs, in line with the implementation of the 74th Constitutional Amendment Act (74th CAA). Achievement is rated satisfactory ) The focus of the institutional development component shifted in the course of implementation from the original emphasis on CCP s to accounting reform, following UAID s lead in the FIRE program. The accrual accounting system has been rolled out to all corporations, municipalities, and town panchayats (now reclassified as special village panchayats. 49 of the special village panchayats are due to be re-classified as municipalities to reflect their urban status.). The computerization of accounts is largely complete. This achievement is quite substantial and is unparalleled in India. While the starting position for Tamil Nadu was better than in other states, due to the tradition of strong accounting, the change is nonetheless significant, and required strong commitment from GoTN. For example, chartered accountants were hired to move the transition forward. everal ULBs have been visited in the course of supervision missions and the ICR mission, and it is generally agreed that the change is a substantial one. Computerization of accounts has reduced the time to close the accounts, and morale of municipal accountants has improved because they have access to computers for their work ) Not only have municipal accounts been computerized, but information kiosks for use by the public have also been introduced in all municipalities and corporations, providing residents with information on the amounts due for property tax and water charges, for example, at computer terminals provided at tax payment centers. Government Orders have been issued requiring municipalities to offer payment only through these automated centers. This provides transparency to taxpayers and reduces the elements of discretion exercised by tax collectors in their interface with the public ) To support the transition to the new accounting system and computerization, the PMU completed a major training program. Under the project, 11,400 training man days were provided per year. This compares to an average of 400 per year in the period An external evaluation of the training program (prepared by the Madurai Institute of ocial ciences) indicates that 71% percent of 160 revenue officials in ULBs interviewed were either fully or partially familiar with the new accounting system. For a sample of 25 ULBs, the number of audits taking thirty days or less increased from 8% to 32%, and the number of audits taking between 91 and 200 days declined from 28% to 12%. It is disappointing to note, however, that the audit lag for ULB accounts remains significant. Only 4 of a total of 102 municipalities have completed their audits for as of November 2004, and 34 have been completed for Reducing the audit lag was not explicitly targeted under the institutional development component, but it is important to the ultimate goal of achieving market access ) The implementation of the training component was delayed for some time, and much of the training took place in the last year of the project implementation. The actual number of man days of training provided undershot the targets in the Project Implementation Plan (PIP) by about 36%. Costs per participant were roughly as planned. Delays in mounting the training component and difficulties in freeing up staff from their immediate work responsibilities were the key constraints to reaching the training objectives fully. For the design of future projects, it should be noted that the external evaluation consistently noted that the trainees would have preferred more use of local language rather than English in the training courses ) The financial capacities of municipalities have been increased over the project period, largely due to the transfers provided to ULBs in the tate Finance Commission (FC) allocations. Between FY2001 and FY2003, total revenues in municipalities have increased by about 42% and in corporations by 49% (13% and 14% annually).this performance indicates a very modest growth in ULB revenues in relation to tate GDP.(Average annual growth of nominal GDP over the seven year period between 1995 and 2002 (latest available data) was about 12%.) While the tate cut back and delayed its transfers to ULBs in during a financial crisis, - 5 -

10 since then these transfers have resumed and were increased during the period of the econd FC award ) ULB performance on own revenues shows a more mixed picture. Only water charges (accounting for less than 10% of total revenues in ULBs) have increased at rates that exceed inflation. The property tax, the main own revenue source for ULBs, has only increased at about 5%, and in municipalities, the annual growth of property tax revenues has declined in the period as compared to GoTN s decision to delay the property tax revaluation planned for 2001 has played an important role in this regard. Because the new municipal law has not been implemented, the shift to an area-based scheme for property taxes, envisaged at appraisal and expected to increase property tax revenues, has not been implemented ) In spite of the constraints experienced due to the slow pace of devolution, the TNUDF has worked with its borrowers to encourage them to improve their finances in the course of project appraisal. They pioneered with ULBs the practice of obtaining advance deposits from users in some water and sewerage projects to defray part of the capital costs. everal ULBs raised their water tariffs in response to TNUDF evaluations of the financial viability of their proposed projects. In some cases, this has resulted in increased revenues. In others, however, revenues have declined due to reduced collection efforts. Nonetheless, TNUDF s efforts to improve resource mobilization have contributed to sending a message to ULBs and their constituents about the links between cost recovery and services, and have helped the TNUDF to understand how to address these issues more effectively in their future lending. DO 2: Mobilizing resources for basic urban infrastructure investments costing about $183 million (water supply, sanitation, solid waste management, roads, transport, storm water drains, and street lighting, etc.) atisfactory Line of Credit Component ($173 million) 4.1.9) The project succeeded in mobilizing considerable resources for investments in urban infrastructure. Under the line of credit component, nearly $160 million of investments were supported, against an estimate of $173 million at appraisal. A total of $62 million were financed directly by the Bank line of credit (estimated at $80 million at appraisal) while over $97 million in other funds was mobilized for urban investments, overshooting the appraisal target by about 5%. Additional Annex 8 provides details on the benefits of these investments as projected at the time of appraisal by the TNUDF ) The line of credit component performed exceptionally well for the first three years of the project. Loan approvals and disbursements from the fund picked up quickly after project approval, and were ahead of schedule after a year. Project preparation costs were reduced from over 13% of project costs under the MUDF to about 7.5% in this project. This roughly translates into savings of Rs. 230 million (over U$ 5 million) over the project period. Cost overruns on the larger ULB sub-projects were about 3% under the project, as opposed to 10% otherwise. Loan recoveries were excellent at close to 98-99%, better than projected at appraisal. (With regard to measuring overdues, the TNUDF follows Reserve Bank of India (RBI) guidelines for recognizing non-performing assets. Thus arrears start to be recognized after 180 days.) GoTN provided the TNUDF with access to intercepts of tate transfers to ULBs for a major tate sponsored program for improving roads in ULBs, and this greatly facilitated high recovery rates. Both the rapid disbursements and recovery performance of the TNUDF are outstanding in the Indian context. Projects were executed more rapidly than when executed by government departments. For example for many roads projects, they were completed in two to three months, whereas these often took over a year when executed by the state road department ) However, after the mid-term review, the TNUDF faced several difficulties that slowed the progress of disbursements under the line of credit. These difficulties all took time to resolve, and as a result, disbursements of the Bank line of credit at completion were only about 77% of the original projections. In FY , the GoTN experienced a fiscal crisis, and had to reduce its transfers to ULBs, and paid them late. The uncertainty this created, as well as the immediate financial impact, significantly slowed down the - 6 -

11 development of a project pipeline. The tatewide revaluation exercise used for property tax assessments, to be undertaken in 2001 was delayed, and given the upcoming elections in 2006, no immediate plans are underway to complete it. This decision has reduced the buoyancy of property tax revenues. The annual rate of growth of property tax revenues was only 6% in FY01-FY04, while it had been about 18% from FY98-FY01. Meeting the covenant requiring that the TNUDF mobilize $25M in sub-project co-financing from the three participating financial institutions(fis) caused difficulties. (ee also paras ) Eventually, the issue was satisfactorily resolved through amendments to the covenant, but the uncertainty over several months retarded project development. The process of TNUDF sub-project clearance, a responsibility of the Commissioner of Municipal Administration(CMA) office of the GoTN, slowed down considerably, and this problem was only recently resolved. Conditions in credit markets in India had shifted considerably, and with them, the TNUDF s market position. At appraisal, credit markets were tight, and the TNUDF s cost of funds was somewhat advantageous in comparison to market rates. By , interest rates were declining rapidly, financial markets were quite liquid, and interest rates have only started to increase recently. The pricing formula established for the TNUDF responded to market shifts with a one year lag, which put the TNUDF at a distinct disadvantage. It should be recognized that the TNUDF s vulnerability to competition in the market was at least partially a product of its own success in introducing the ULBs to financial markets. However, the inflexibility in their pricing formula made it difficult for them to respond to this competition. Two years ago, the 611 Town Panchayats were re-classified as pecial Village Panchayats. As such, they are not allowed to borrow from the TNUDF. Now, some 49 of those are in the process of being reclassified as municipalities to recognize their urban character. It is not clear how much of the potential borrowing base of TNUDF was affected by this decision, but this grouping includes many of the suburban areas surrounding major metropolitan cities, who often have borrowing capacity. It took a year for a new CEO to be appointed to run the TNUDF, after the first CEO left for a different position. This delay was particularly costly for the TNUDF because it coincided with the other adverse developments noted above, and doubtless impeded the formulation of a vigorous response to them ) Given the number of adverse factors affecting the project pipeline, it is not surprising that they were not all resolved in time to meet the original project targets. Nonetheless, the overall result indicates a substantial mobilization of funds for urban investments. The roughly $32 million per annum of investments supported in the context of the credit line component of the project compares favorably to total investments in ULBs in the pre-project period, estimated at $55 million in FY Integrated anitation Component ($10 million) ) This component successfully constructed all the units envisaged, with details shown in table 3 of Annex 8. Although considerable delays were experienced in the beginning of the project, these were made up in the later stages of project implementation. As noted earlier, the direct contribution to capital costs from ULBs was not the 50% envisaged at appraisal, and the Bank funds financed more than the 50% originally envisaged. This result is due, in part, to the fact that no ULB contribution was required of the town panchayats, and the Bank financed 70% of the capital costs for them. However, ULBs were required to provide municipal land for the construction of the facilities. In many of these ULBs, the land has substantial value, in some cases amounting to the equivalent of direct construction and equipment costs. As such a substantial hidden subsidy is provided by the ULB to the program. In some cases, the land provided by the ULB was far enough from the target users residences that the utilization of the unit was compromised. The user groups were expected to recover only out of pocket maintenance costs, including electricity, but the waste removal costs are borne by the municipality. These costs can be substantial, and depending on usage, can be close to the direct costs borne by the user group ) While the subsidies on the IP are large, GoTN considers these projects a substantial improvement over the other sanitation projects typically sponsored in ULBs. Those projects do not mobilize user groups, and recover no user charges for operation and maintenance. Often these facilities are no longer used within months of - 7 -

12 construction ) The IP component has succeeded in mobilizing some communities, and achieved health benefits in the neighborhoods served, reducing open defecation by 80%. The results for the IP component nonetheless show that the efficiency achieved is less than projected at appraisal, and somewhat disappointing overall. The units are only used by 63% of allotted users in the municipalities and 50% of allotted users in the town panchayats. In only 25% of the units in municipalities were all facilities (bathing, washing and latrines) being used fully; this was the case for only 10% of the units in the town panchayats. 90% of the units were managed by CBOs and functioning as of the time of the independent evaluation in November, However, only two thirds of the units were assessed to be in good working order in municipalities and 45% in the town panchayats. This performance is disappointing, and compares unfavorably with a similar component in the Bombay ewage Disposal Project in which all units were operating at project completion. Regarding collection of user charges, 60% of the units in municipalities were collecting from 100% of users, while only 20% were doing so in the town panchayats. Problems with usage, maintenance, and cost recovery appear to have been particularly acute in the town panchayats. ocial marketing to promote the concept of using latrines rather than open defecation is particularly important in these less densely populated areas. The CBOs appear to have had limited impact on the design and selection of locations, and the uneven usage of facilities reflects this. Although a rigorous comparison is beyond the scope of this ICR, nonetheless evidence from two other slum sanitation programs, one supported by the Bank in Bombay and another undertaken by NGOs in Pune, suggests that it may be possible to provide similar services at considerably lower costs. ( Bombay ewage Disposal Project ICR June 12, 2004 p.6,7 indicates a construction cost of roughly Rs per beneficiary as against an estimated Rs per beneficiary in this project. "Pune lum anitation", Manuscript April 2004, Water and anitation Program indicates that per seat toilet costs were closer to Rs. 50,000 vs. an average of Rs. 100,000 per seat in this project.) In spite of these shortcomings, the project has channeled considerable additional resources for these critical services for the poor. DO 3: ecuring sustainable sources of funding for the urban infrastructure investment, through TNUDF and municipal bond issuance, beyond the Bank s line of credit operations. atisfactory ) The spirit of this objective has been met, although not as envisaged at appraisal. Unquestionably, the TNUDF has sought to link ULBs with broader capital markets in India. Through the use of the grant fund, the TNUDF was able to assist some ULBs in exploring innovative project structures and new sources of financing. The TNUDF, in response to a covenant, mobilized funding through a bond issue, although this funding came at relatively high cost in comparison to other sources of funds. Direct municipal bond issuance has not taken off as was expected at the time of appraisal,. Although TNUDF did issue one bond, they also facilitated several other innovative transactions: Madurai Ring Road, a toll road financed by TNUDF and taken out by the market (U $ 6.4 Million), resulting in a savings of Rs 6.5 Million (U$ U$ approximately) a year to the municipality. Karur toll bridge (Rs Million) on BOT, with a concession period of 14 years. Alandur ewerage with an investment of Rs.80 million on the treatment plant by the operator of a total project cost of Rs. 420 million Tirunelveli and Cumbum Bus tands financed by the project, with loan repaid on receipt of cash payments from potential tenants. This complex in Tirunelveli also includes outsourcing of Operation & Maintenance (O&M) resulting in an annual saving of Rs. 2.5 Million. Facilitated the first tax free Municipal Bond for Chennai Metro Water upply & ewerage Board (CMWB) to finance the Chennai Water upply Augmentation Project I (Rs. 420 Million) A pooled financing bond issue to provide water facilities to several towns ) In the course of project implementation, the expanded involvement of the participating financial institutions (PFIs), who were shareholders in the restructured fund did not materialize. The policy purpose of the covenants requiring additional funding from the PFIs was to draw them into a financing relationship with urban local bodies. It now appears that the PFIs (and the Government of India) have defined their mission so as to exclude this kind of local government development financing, at least as regards smaller scale projects of the type - 8 -

13 supported by TNUDF, and have thus not developed the capacity to process them. There are good reasons for huge IFIs like ILF and ICICI to concentrate on the financing of major infrastructure projects and wholesale support for financing systems instead, and the tax treatment of the dividend was a serious disincentive to further equity contributions. It is to the credit of the TNUDF that they sought out satisfactory alternative approaches to meeting the intent of the covenant requiring additional municipal funding from Indian capital markets ) It is important to assess the achievement of this objective in the context of the evolution of the decentralization process in India and the optimistic assessment (since revised) of the prospects for private urban infrastructure finance in India that was prevalent in the latter half of the 90's. In practice, the tates in India, and Tamil Nadu has been no exception in this regard, have proceeded slowly with the process of ceding both revenue sources and autonomy to ULBs. As a result, ULBs in Tamil Nadu are still very highly dependent on GoTN support for any important decision. The high degree of ULB dependency on the GoTN, combined with the small revenue bases of ULBs, greatly reduce financial institutions appetite for investing in a direct relationship with most ULBs at this stage. In such a situation, an intermediary such as the TNUDF plays a role in linking the ULBs to financial markets. The discipline that the TNUDF exercises in using commercial criteria to appraise their projects is valuable in inculcating fiscal discipline and preparing the ULBs for a more direct relationship with the capital markets as the gradual process of decentralization unfolds. It now appears that this period of transition for both the ULBs and the TNUDF will be longer than expected at the beginning of the project. The slow pace of decentralization also underscores a delicate balance for the TNUDF. It must manage a relationship not only with multiple ULBs but must also establish a strong working relationship with the GoTN, because they are the TNUDF s key client until substantially more autonomy, both financial and administrative, devolves to ULBs ) Thus, the TNUDF, with support through the project and aided by favorable capital market conditions, has forged important new links between the ULBs and capital markets, and has helped to promote innovative transactions. Nonetheless, it is important to note that the overall policy environment in the tate has restrained the autonomy, both financial and administrative in ULBs, and thus limits how much investment can be funded from capital markets. 4.2 Outputs by components: Credit Line: 4.2.1) The table 1 in Additional Annex 8 provides details on the physical outputs of all sub-projects financed under the credit line, as projected at the time of sub-project appraisal. Under the line of credit component, nearly $160 million of investments were supported, against an estimate of $173 million at appraisal. A total of $61.5 million were financed directly by the Bank line of credit (estimated at $80 million at appraisal) while over $97 million in other funds was mobilized for urban investments, overshooting the appraisal target by about 5%. Table 4 in Annex 8 shows the sector-wise distribution of approvals of sub-projects supported under the credit line. Integrated anitation Component 4.2.2) Please see para Also please note Table 3 of Annex 8 for specifics on the number of units built and costs. Institutional Development Component: 4.2.3) Please see paras Net Present Value/Economic rate of return: 4.3.1) No economic rate of return was computed for sub-projects before project approval, although this was provided for in the lending guidelines. ince no base line before project and under project improvement scenarios were developed at appraisal, it is not possible to assess benefits ex-post. TNUDF does not systematically follow up on the operations of their projects, since their main concern is that the loans be repaid, and this has not been a problem thus far. As mentioned above, costs and delays have been reduced in the projects supported by the TNUDF, which indicates that there are clear economic benefits to using the TNUDF as an intermediary. Better - 9 -

14 monitoring of actual vs. projected project outcomes and assessment of economic returns as a funding criterion for sub-projects would have been desirable, and this point will be followed more closely in the follow-on project. 4.4 Financial rate of return: 4.4.1) The TNUDF did not compute a financial rate of return for many of its projects, since debt service was often funded out of tax revenues rather than specific user charges, and debt service capacity was their primary concern. However, for two projects, the Karur Bridge and the Madurai Ring Road, debt service was to be funded out of tolls, and thus performance was tracked by the TNUDF. In both cases, the FRR projected at appraisal was 18%. In the case of the Madurai Ring Road, the actual revenues have fluctuated between 44% and 50% of projections, indicating an unsatisfactory rate of return. On the other hand, the Karur bridge has achieved collections that are nearly 70% of projections. In both cases, the hurdle rate of 14% FRR is not likely to be met unless there is considerable improvement later. Both of these projects are indicative of the risks of PPP s in the Indian environment. In both cases, projected tolls did not materialize due to resistance to paying tolls amongst the local population. 4.5 Institutional development impact: 4.5.1) The institutional development impact of this project is rated substantial. As noted in the discussion of the DOs, the TNUDF has achieved a qualitative improvement in the intermediation services available to ULBs in Tamil Nadu. In the early years of the project, the TNUDF achieved good financial results, broadly in line with projections at appraisal, and exceeding them in one year. They have maintained strong repayment discipline throughout. However, as noted earlier (para ), while disbursements quickly exceeded projections in the early years of the project, this momentum was not maintained past the mid-term review. Not only did disbursements decline very substantially, many loans were prepaid, as another GoTN sponsored intermediary (TUFIDCO) provided financing at more attractive terms to ULBs. As the balance sheet shrank, profitability suffered. If it had not been possible to prepay the loans from GoTN without penalty, the impact on profitability could have been much more serious. ee Annex 8 Table 2 for a summary of TNUDF financial performance ) The massive refinancing of TNUDF s existing loans is largely attributable to the inflexibility of the pricing formula (interest rates were reset only once every six months) and loan tenor used by the TNUDF (all sub-loans had a 20 year tenor although the project documents only specified this as the maximum maturity) in a highly liquid financial market. TUFIDCO was able to mobilize funding from commercial banks and then offer refinancing on favorable terms to the ULBs. A Government Order in 2003 authorizing ULBs to avail themselves of this facility gave momentum to the refinancings. This process has benefited ULB finances, but it is unfortunate that TNUDF was neither able to seize this opportunity itself nor utilize the reflows that arose from the refinancing for new lending to ULBs. However, GoTN has shown its commitment to revitalizing the TNUDF in recent months by appointing a new CEO, addressing the conflicts between TNUDF and the CMA s office, and agreeing to a more flexible pricing formula for TNUDF operations in the context of appraisal of the upcoming TNUDP III project. The three PFIs have maintained their, admittedly modest, shareholdings, and the lead PFI, ICICI demonstrates interest in seeing the TNUDF succeed ) With regard to details on the impact of the Institutional Development Component, please see the discussion of DO1, paras Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: 5.1.1) As the capital market became highly liquid and interest rates came to decline rapidly over the course of project implementation, this affected the project, both positively and negatively. Financial institutions across the board were looking for new investment opportunities, so it was easier for the TNUDF to facilitate transactions that brought finance to ULB projects from new sources. The presence of the grant fund for project preparation made it easier to seize these opportunities by subsidizing the structuring of transactions. At the same time, the onlending supported by the World Bank became less competitive in the declining interest rate environment

15 5.1.2) The ruling of the Inland Revenue of Government of India against the tax exempt status proposed for the fund had an important impact on the appetite of the PFIs for further equity contributions. Moreover, the shift in mandate for the ICICI away from project finance, affected their appetite for co-financing ) The broader fiscal crisis that GoTN faced in led to a reduction and interruption of devolutions to ULBs. This clearly affected sub-project pipeline development. 5.2 Factors generally subject to government control: 5.2.1) GoTN allowed the TNUDF to avail of an intercept facility for recovering sub-loans in the line of credit component made in support of a large GoTN sponsored ULB road resurfacing project that accounts for about $43 million of TNUDF disbursements, making it easier to manage this large set of subl-loans. This facility is also available to other intermediaries upon request from GoTN. It is important to note that the TNUDF has performed very well in recovering loans overall, not just those that used the intercept facility ) The decision to delay indefinitely implementation of the TNULB passed in 1998, due to stiff union resistance to the provisions affecting public sector employment, sent a strong signal that the feasible pace of decentralization would be slower than originally anticipated ) The decision to delay the revaluation exercise for property tax assessment purposes for the entire project period, and now planned for after the elections in mid-2006, signaled strong hesitation to tax urban areas and to bolster the ULBs most significant own revenue source. This action demonstrated, to financial markets among others, how much ULBs depend administratively and financially on GoTN decisions ) The GoTN failure to appoint a permanent CEO during the better part of 2004 affected TNUDF s ability to respond to important market developments during that time. This problem was recently corrected ) ee para regarding delays in technical sub-project approvals in the CMAs office. After about two years of virtually no sub-project approvals, the GoTN stepped in to address the issue, which it has done recently ) ee para on the GO authorizing refinancing of TNUDF loans. 5.3 Factors generally subject to implementing agency control: No significant factors. 5.4 Costs and financing: No significant factors. 6. ustainability 6.1 Rationale for sustainability rating: ustainability is rated likely based on the strength of the IDC and the commitment to the TNUDF model )Please see paras and on the sustainability of the TNUDF. The institution has maintained reasonable profitability, and in the course of appraisal of TNUDP III has agreed to measures that will permit the TNUDF to respond better to changes in the market )Regarding the IDC, measures have been taken to institutionalize the progress made. The training program has built up a skill base amongst the staff of a wide range of ULBs, and the CMA s office has issued orders that there are to be no transfers of municipal accountants trained under the project before 2006 to ensure continuity in the implementation of the new accounting system. 6.l.3) The sustainability under the Integrated anitation Component is less clear. The independent evaluation has raised questions regarding long-term sustainability. As noted in para , a significant share of the facilities are already not in working order or utilized fully. A need for foolproof systems for accounting and revenue

16 management to avoid leakage of revenues has been identified. Only 52% of the CBOs in the municipalities and 20% in the TPs felt they had received all the support required of the ULBs. The performance of the CBOs in collecting user charges is quite uneven. ee para Transition arrangement to regular operations: The PMU staff under the project has been moved into the CMA s office and its functions will be handled in the CMA's office in the upcoming project. 7. Bank and Borrower Performance Bank 7.1 Lending: This is rated satisfactory based on the strength of the line of credit component. As noted in the section on quality at entry, the Bank developed a reasonable project design with the borrower, building on earlier efforts to revamp the MUDF into the TNUDF. This concept proved its worth in the results that the TNUDF was able to achieve. The design could have been more robust had the role of GoTN in both approving ULB decisions and affecting the ULB revenue base been factored more clearly into project design. Moreover, given the private management of the fund, more flexibility in pricing and terms might have been offered the TNUDF. The IDC and IP components would have benefited from more detailed technical preparation and clearer agreement with the GoTN on the objectives and means of achieving them. 7.2 upervision: atisfactory The Bank showed useful flexibility in addressing the issues related to the second covenant on equity contributions from the PFIs, and in adopting the accrual accounting work launched by UAID in the IDC. It is unfortunate that this covenant was such a preoccupation just at the time that so many other factors were affecting the project pipeline and TNUDF s business prospects. More attention by the Bank to the many other issues that were masked by the impact of the uncertainty regarding the second covenant might have prompted a quicker response to other problems that were surfacing. Fortunately, these issues have come to the fore and are being addressed as preparation of the TNUDP III project proceeds. 7.3 Overall Bank performance: atisfactory Borrower 7.4 Preparation: atisfactory GoTN was proactive and creative in the approach they took to the restructuring of the TNUDF, which was the linchpin of the preparation of the TNUDP II. 7.5 Government implementation performance: atisfactory The GoTN s support through implementation was uneven. In the early phase of the project, GoTN ensured that implementation moved forward satisfactorily. However, later on a number of actions were taken that were detrimental to the TNUDF, and which slowed implementation, as mentioned in paras , As noted above for the Bank, in the period after the mid-term review, while the GoTN worked proactively to resolve the issues raised by the second convenant, the other issues affecting the project could have received earlier attention. However, overall performance is rated satisfactory since the GoTN has taken steps to address these issues now. 7.6 Implementing Agency: