Actual Project Name : Second Community. Project Costs (US$M US$M): Sector Board : Water Cofinancing (US$M US$M): US$M): ICR Review

Transcription

1 Public Disclosure Authorized IEG ICR Review Independent Evaluation Group Report Number : ICRR Project Data: Date Posted : 11/08/2013 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Country: Sri Lanka Project ID : P Appraisal Actual Project Name : Second Community Project Costs (US$M US$M): Water Supply And Sanitation Project L/C Number: CH035 Loan/Credit (US$M US$M): Sector Board : Water Cofinancing (US$M US$M): Cofinanciers : Board Approval Date : 05/06/2003 Closing Date : 06/30/ /07/2011 Sector(s): Water supply (35%); Sanitation (35%); Other social services (10%); Sub-national government administration (10%); Central government administration (10%) Theme(s): Pollution management and environmental health (29% - P); Rural services and infrastructure (29% - P); Other human development (14% - S); Participation and civic engagement (14% - S); Decentralization (14% - S) Prepared by : Reviewed by : ICR Review Group: Coordinator : Kavita Mathur Robert Mark Lacey Soniya Carvalho IEGPS1 2. Project Objectives and Components: a. Objectives: According to the Development Grant Agreement (Schedule 2, p. 16) the objective of the Project was: to increase service coverage and achieve effective and sustained use of water and sanitation services in rural communities in Sri Lanka. The statement of objectives in the Project Appraisal Document (p. 2) is identical. b.were the project objectives/key associated outcome targets revised during implementation? No c. Components: There were five components. The fifth component was added under a "Part Z" adjustment to the Credit Agreement in February 2005 following the December 2004 Tsunami: A: Physical Investments (appraisal estimate US$38.4 million; actual cost US$34.10 million because US$6.70 million was reallocated to the Tsunami component (Z)). This component aimed to finance a share of the capital costs for service improvements in water supply, waste water, drainage and on-site sanitation. Beneficiaries, represented by community-based organizations (CBOs), were to be responsible for management and implementation of construction works and subsequent operation and maintenance of water supply and sanitation systems. The majority of the schemes were to be implemented in rural villages, communities in estate sector and small towns eligible initially on a pilot basis. In addition, the project would finance completion of selected sub-project initiated

2 by the government in the NW Region. B: Implementation Support (appraisal estimate US$4.4 million; actual cost US$3.39 million). This component aimed to support Partner Organizations and strengthen their capacity to assist communities in implementing water and sanitation works through: (i) publicity and awareness campaigns; (ii) Assistance to communities in preparing funding applications; (iii) formation of CBOs; (iv) work with CBOs to identify, plan, design, supervise, and implement water supply and sanitation schemes. In addition it was to assist local governments in (v) processing and supervising sub-projects; and (vi) provide in-situ training and capacity-building to CBOs to enable them to assume project management responsibilities and ensure effective use and maintenance of facilities created under the project. C: Program Management and Capacity Building (appraisal estimate US$12.60 million; actual cost US$11.64 million). This component aimed at building capacity of national unit, local governments and CBOs in implementing the project with the following sub-components: (i) Technical assistance and training; (ii) Consultancies, studies and policy formulation various technical studies; (iii) Incremental operating costs to support project administration and operating costs; (iv) A project-wide groundwater survey to identify potential water resource constraints; and (v) The detailed study to consider financing activities to address the water scarcity in Hambantota District of the Southern Province. D: North East Provincial Council (appraisal estimate US$7.00 million; actual cost US$6.54 million). This component was to finance various preparatory and implementation activities related to sub-projects in the NE Province including technical assistance, training and capacity building. Z: Tsunami Assistance for Recovery, Rehabilitation and Reconstruction Activities (appraisal US$7.00 million, actual US$6.70 million). This included two subcomponents. (a) Carrying out of emergency recovery and reconstruction activities in Tsunami-affected areas in various sectors including, but not limited to, health, education, infrastructure, roads, housing, capacity building, long-term hazard-and-risk management, training, and support for small-and-medium enterprises, by providing goods, services, training, operating costs, and sub-grants, and carrying out works required for this purpose; and (b) cash grants for livelihood support to individuals and families in the affected areas in accordance with acceptable assistance criteria including, but not limited to, various sectors, including health, education, infrastructure, roads, and housing for capacity building, support for small and medium enterprises, and establishing a civil society fund. d. Comments on Project Cost, Financing, Borrower Contribution, and Dates: Project Costs : The cost of the project was initially estimated at US$ 62.4 million, but at closing project costs were US$ million. This includes approximately US$7.0 million reallocated to the Tsunami Emergency Recovery Project in 2004 and disbursed through a new category 'Z' added to the project because expenditures for tsunami -related activities under the Tsunami Emergency Recovery Project (P094205) were incurred from the parent project accounts. While the US$6.7 million spent on the Tsunami project was fungible, the Tsunami project spent US$34.6 million of infrastructure rehabilitation that was to include water supply and sanitation. The main effect of this diversion was that project expenditures were reduced, and the budget of component A was reduced to 89% of that planned. Service delivery to small towns (already delayed) was not undertaken. In addition, the detailed study under component C to finance activities to address the water scarcity in Hambantota District of the Southern Province, estimated to cost US$ 1.20 million, was dropped because the government financed it out of its own funds. Financing : At appraisal an IDA grant commitment of Special Drawing Rights (SDR) 28.9 million (US$39.8 million equivalent) was approved. Due to the Tsunami in December 2004, US$7.0 million equivalent was reallocated from the project in February 2005 to the IDA-financed Tsunami Emergency Recovery Project and US$ 6.7 million of this was spent. Actual disbursement was SDR 28.9 million equivalent to US$43.6 million equivalent because of appreciation of SDR against the US$. A small amount of SDR (14,856) was cancelled at closing. Borrower Contribution : At appraisal the Borrower planned to contribute US$ million equivalent; actual contribution was US$ 8.91 million equivalent or 78% of that planned. In contrast, communities were expected to contribute US$ million equivalent but actually contributed US$ million equivalent (145% of planned). Dates: The Grant Agreement was amended twice, once in November 2005 to change the definition of incremental

3 operating costs and also to enable disbursement for Tsunami related emergency equipment. A second amendment was made in mid-2007, to reflect the splitting of the original North East (NE) Provincial Council into two separate Councils. Due to the security situation; the project area was also moved to relatively safer locations within the NE province (Trincomalee and Ampara districts). The project, initially planned for 6 years, was to close on June 30, However, to enable the completion of works, the project closing date was extended by 18 months until December 31, Relevance of Objectives & Design: a. Relevance of Objectives: Substantial Objectives were substantially relevant to the needs of the rural population. Sri Lanka's rural areas house 19 million people that account for 80% of the population and 90% of the country s poor. At appraisal water access was low: 60% of villages and small towns and 70% of the estate population had no access to safe water sources. Access to sanitary latrines varied from 70% in the villages and small towns to about 40% among the estate population. On average, only 1 out of every 10 households had a house or yard water connection, half of the population had to collect water from outside for drinking purposes, and every fifth person was relying on unprotected water sources, rivers and streams. Water quality due to pollution of natural sources was a major problem in most areas. Additionally, the NE Province, apart from its many humanitarian problems as the result of civil conflict, was beset by chronic water supply problems due to its geographical location in the dry zone. Project objectives were relevant to the Government's Rural Water Supply and Sanitation Sector Policy and associated programs that focused on expanding access and decentralizing service delivery. The project s objectives are also consistent with the Government s vision for economic development laid out in the Mahinda Chintana 10-year Development Framework, which focuses on three main areas over the period : (i) achieving more equitable development through accelerated rural development; (ii) accelerating growth through increased investment in infrastructure; and (iii) strengthening public service delivery. The project s objectives are consistent with the Bank s Country Assistance Strategy and directly responds to the stated Strategic Objective 1: Expanding Economic Opportunities in Lagging Regions. This includes increasing access of rural households to improved water sources with special focus on the NE province. Additionally, the Bank's Strategy strongly emphasizes community -focused and community-driven projects facilitated through local and provincial governments. b. Relevance of Design: Substantial The project design built on the experience gained from the first community water supply and sanitation project and experience from India, and its results framework logically linked inputs to expected outputs and outcomes. Technical assistance and training was included to support government actors in the implementation of the project and provide training to increase the capacity of implementing entities, including the central Rural Water Supply and Sanitation Division and its Units in the Provincial Councils, the Pradeshiya Sabbas (local governments), partner organizations and community-based organizations. This was designed to bolster their skills and capacity for procurement, community mobilization, planning and financial management. A contribution of 5% of scheme cost from local government was relevant as it ensured their active participation in the project and their support for community contracting to minimize costs. The demand-responsive approach was highly relevant. Public awareness campaigns were organized to target the least served areas and build community interest in what the project could provide. Complementary hygiene education campaigns were conducted under parallel financing through the Bank -Netherlands Water Partnership. Given the scarcity of reliable water sources, water resource surveys were relevant as they ensured selection of sustainable water supplies from groundwater and surface water. Support and training for partner organizations was relevant to building local capacity within communities to begin the planning process to build water supply and sanitation schemes. These partner organizations assisted establishment of community-based organizations to undertake the planning, construction, commissioning and operation of schemes. To ensure community ownership, the demand -responsive approach required communities to make up-front contributions in cash and kind contributions of 20% of the cost of water supply facility chosen. This approach was relevant as it allowed poorer communities to choose technology they could afford and it provided an incentive for communities to look after their investment. Additionally, communities were trained in the operation of completed schemes and the setting and collecting of water tariffs to ensure financially

4 viable maintenance. Community contracting was relevant as it reduced costs and increase ownership. The phased approach through five successive batches was relevant as it allowed the project to build on lessons learned. Importantly, the project supported community -based and partner organization advisors over the 23 months it took to move from scheme pre-planning to post-construction support. The Sanitation Revolving Fund that contributed towards the cost of sanitation on a cost -sharing basis was relevant as it reinforced community stewardship of grant funding. Pilot projects designed to learn more about the special circumstances in Sri Lanka were relevant. While the majority of schemes were expected to be implemented in rural villages, communities in the estate sector were also eligible for project funding for pilot water supply schemes. It was expected that estate schemes would require modifications to the subproject cycle in response to lessons learned. This was important as estate workers had no experience in organizing themselves for development of infrastructure, and had never experienced handling large sums of money. In addition, the project financed a strategy for delivery of water supply and sanitation services to small towns and up to 2 pilots were to be implemented before scaling-up. The challenges posed in the conflict-affected North-East were to be met through a phased approach to reduce risks; preliminary work was to set up of Rural Water Supply and Sanitation Units in the Provincial Council followed by a replication of the demand-responsive approach used elsewhere in Sri Lanka. The proposed Kirindi-Oya river basin study - a minor project component - was ill-suited to project development objectives and was of low relevance given the institutional and scaling -up challenges to be faced by the rest of the project. The additional component added to the project as a result of the Tsunami Emergency effectively earmarked part of the project credit for recovery and reconstruction activities. IEG's Review of the Implementation Completion Report of that project (P094205) rated relevance of design as substantial. 4. Achievement of Objectives (Efficacy): To increase service coverage and achieve effective and sustained use of water and sanitation services in rural communities in Sri Lanka. Each of the two parts of this objective statement is assessed separately below. 1. To increase service coverage : Modest The Tsunami of December 2004 was an exogenous event that adversely affected the project. About 18% of the project grant and 22% of the government's potential funding was reallocated to the Tsunami recovery efforts. The tsunami further delayed commencement of the component in the NE Province that was severely hit by the tragic disaster. Subsequently, the same region was adversely affected by the resumption of civil conflict and the subproject areas had to be relocated to safer areas within the same province. Outputs: Increase in access to improved water service coverage : This was achieved in two steps. First the construction of the water supply schemes; second the connection of the water supply pipework to communities and households. Around 709 water supply schemes were completed compared with the target of an achievement of 75%. Of the 805 schemes approved, 59 were abandoned before commencement for various reasons (including unwillingness of some community members to contribute to costs ). 43 schemes in the fourth and last batch could not be completed by project closing (December 2010). On the basis of the project's commissioning experience with earlier batches, it is expected that all schemes will be completed by the end of Indeed, by the end of May 2011, 20 of the incomplete schemes had been commissioned (ICR, page 46). Almost 28, or 68% of 41 projects initiated in the North East Province (Trincomalee and Ampara District) were completed and operational at project closing (ICR, page 26). There were no target values. Drinking water was supplied by various technologies including : 1,054 piped systems, 1,965 private wells, 36 tubewells, 116 hand pumped wells and rainwater harvesting tanks. Some 922 community water points were constructed or rehabilitated under the project, meeting almost 82% of the target of 1,121. Water points catered to poor people from the community who could not afford household connections. It was expected that the project would increase access of 311,800 households to improved water supplies - this included 220,000 rural households, 90,000 households in government schemes completed by the project and 1,800 households in the small town pilot (which was cancelled in 2005). By the end of the project, however,

5 constructed water supply schemes had only the capacity to supply 176,326 household yard-taps, or 52% of the target. Even then, only 92,000 of these 176,326 households could be connected to a private connection before the end of the project. The ICR states that all project households could be connected within a year (ICR, page 6) but notes that about 18% of targeted households were unwilling to connect, many preferring their own water sources. Fewer households than planned could be served because of diversion of the 18% of the project grant to Tsunami recovery effort and because the unit cost of the demand -responsive schemes was much higher than anticipated - first because beneficiaries preferred household connections and second because of cost inflation partly induced by the post-tsunami reconstruction activities. At appraisal it was expected that 65% of the beneficiary population would opt for piped water distribution systems connecting water supply facilities to individual households - normally to a private tap in the yard/garden of the house. Actual household yard-tap connection under the project was 95%. At appraisal the cost of the yard-tap household connection was 30% more than a piped communal point source. It was expected also that cheaper common dug wells would be demanded by 20% to 35% of beneficiaries; in fact fewer than 5% of beneficiaries chose this option. As a consequence of these preferences, the average cost of water supplies was considerably higher than planned. In addition, the cost inflation index for civil engineering works, reported by the Central Bank of Sri Lanka in 2012, rose from in September 2003 to in November 2010, a cost inflation of 212%. The average rate of inflation over the life of the project weighted by the phasing of the scheme construction, 78% of which occurred in the last three years of the project, gives an average cost inflation of about 180%. Thus fewer schemes and household connections could be constructed within the project budget. Discounting the actual budget expenditure of US$34.1 million by the amount of cost inflation produces an adjusted project budget that was 55% of that planned. Actual achievement is potentially 52% of the planned target - and this is in line with the available budget. However, as noted above, actual household connections at project completion were 25% of the original target and half the revised target. In mitigation, the Borrower s comments in the ICR state that: Obtaining of house connections takes place, as previously experienced, within 4 6 months from the commission of the water scheme. This is proven by the following figures. During the last five months (since the project closure), 22,979 new house connections have been completed by the beneficiary HHs. When that amount adds to the coverage figures at the closure, 52% rises to 65%, while pipe borne coverage of 50% rises to 64%. This shows that the coverage figures are increasing rapidly. Notwithstanding the Borrower s observations, the Region did not provide updated household connection data in its January 2013 comments on this Review, two years after project closure. Tsunami Emergency Recovery Project. Water supply and sanitation rehabilitation works were not undertaken because they were dropped from this operation's infrastructure component. Increase in access to sanitation : The project built 45,660 latrines ( 91% of the appraisal target of 50,000) through the Sanitation Revolving Fund. The ICR states a higher target of 54,500 but there is no discussion of why the target increased or details of its formal approval. Outcomes: Access to water supplies : As a result of the project, about million people, or 32% of the targeted 1.23 million people, achieved access to safe water sources. The ICR argues that the target was reduced to 0.72 million beneficiaries because of the US$ 7.0 million allocated to the Tsunami recovery and therefore the achievement was 53%. However, the target was never formally revised. However, taking into account exogenous factors (as discussed above) a revised pro rata target would be about 714,000 people. On this basis, achievement by the end of the project was 54%. Based on a small sample size of only 100 sub-projects (a 14%sample), 88% of beneficiaries with completed systems indicated they were satisfied with their access, and 44% of beneficiaries were very satisfied (ICR, page 12). Chlorination is not practiced generally and only 33% of water samples collected passed bacteriological tests,

6 and 52% passed physical/chemical tests (ICR, page 27). Access to improved sanitation : An independent audit found that 97% households had latrines (ICR, page 27). About 10% of this total may have come from other sources because the same survey found that only 87% of community-based organizations had utilized the Sanitation Revolving Fund. According to the government's ICR (ICR page 51), beneficiaries who could not afford to pay all their contribution in one lump sum were assisted by the community -based organizations with loans without interest or part loans with grant components or complete grants. Overall, 471 villages achieved 100% sanitation coverage. There is no information about the utilization rate of latrines. 2. To achieve effective and sustained use of water and sanitation services in rural communities in Sri Lanka : Substantial Outputs: After the initial public awareness campaigns, 755 community-based organizations were established and they were supported by 57 partner organizations that provided technical advice and assisted preparation of funding applications, planning, design and supervision of the construction of the water supply and sanitation infrastructure. Communities were trained in the operation and maintenance of the water supply system, tariff setting, procurement, scheme management and record keeping. In addition, the partner organization promoted health and hygiene education and raised awareness on environmental conservation practices. At project completion 709 community-managed projects were in operation (this is 75% of the target of the 940 schemes started). People's audits of community-based organizations took place, supplementing those by the Auditor General. Outcomes: Water metering is universal. Water use increased. Average daily per capita water consumption (based on household surveys) increased by almost 59% from 28 liters to 68 liters by the end of the project. The majority of beneficiaries have sustainable daily water supplies. 46% of the schemes have a continuous water supply (ICR, page 15), and 78% of households receive water every day (ICR page 34). Household connection resulted in time savings of around 50 minutes per household per day and the main beneficiaries were women and girls (ICR, page 13). These time savings provide a powerful incentive for sound maintenance of water supply facilities. The institutional and project processes have improved the ownership of the interventions by the community (ICR, page 13). Overall, communities contributed 46% more to capital costs than was required: US$11.90 million equivalent was expected but actual contribution was US$ million equivalent. This was important as local governments were only able to provide 2% of project funding instead of the 5% agreed at appraisal (ICR page 18). In an independent Bank survey found that at least 60% of community-based organizations meet regularly with a 70% attendance by members of whom 40% are women. Women also hold over half the officer positions. It was found also that these organization have become de facto development centers in villages. People's audits increased financial transparency and ownership of project facilities. An independent audit found 83% of the schemes were financially sustainable, as they generate sufficient revenue to recover the costs of operation and management, ICR (page 7). This is based on a sample of 14% of sub-projects, and not covering the NE Province. Households were willing to contribute US$18 or Rs.2,000 per latrine under the Sanitation Revolving Fund. At project completion, cost recovery to the Revolving Fund was 76%. The estate sector pilot found that in the conflict -affected NE province the focus on water and sanitation has brought communities together in pursuit of a common objective.

7 5. Efficiency: Overall, the investments generated benefits of time savings for the community in fetching water and increased water consumption. While there were associated health benefits, these cannot be quantified. Economic rates of return were estimated ex-ante and ex-post: The ex-ante (PAD, page 17) cost benefit analysis was conducted to estimate the time savings generated by the water supply investments based on 7 different technology scenarios which was projected to constitute 80% of the total investment cost. The ERR for all technology options ranged from 14.9% to 51.1%. The estimated time savings based on different technology schemes ranged from minutes/day. Sensitivity analysis also showed that a significant reduction in time saving would make investments economically non -viable. The ex-post economic analysis (ICR, p.28) was computed for the two most popular piped water supply technologies (piped gravity and pumping schemes) that provided private yard-tap connection for households and accounted for 91 percent of the water capital costs. On average the achieved time savings were estimated to be 70 to 77 minutes per household per day much less than the appraisal estimate for the yard -tap connections of 120 minutes. While benefits at appraisal only included time savings, the value of incremental per capita daily water consumption, which increased from 29 liters to 68 liters, was factored into the ex post calculation. The resulting ERR was 30% and 18% for gravity and pumping schemes at completion, slightly lower than the appraisal estimates of 43% and 25%. Capital costs of construction were far higher than estimated at appraisal and this gives the appearance of inefficiencies. Gravity piped yard-tap connections cost Rs 30,000 per household, or 76% more expensive than the appraisal estimate of Rs 17,000. Similarly pumped piped yard-tap connections cost Rs 44,000, some 83% more expensive than the appraisal estimate of Rs 24,000. However, as noted in Section 4, cost inflation for civil works was about 180%, mainly because of the Tsunami. Overall, the project efficiency is rated as substantial. a. If available, enter the Economic Rate of Return (ERR ERR)/Financial Rate of Return (FRR FRR) at appraisal and the re-estimated estimated value at evaluation : Rate Available? Point Value Coverage/Scope* Appraisal Yes 25.2% 80% ICR estimate Yes 18% 91% * Refers to percent of total project cost for which ERR/FRR was calculated. 6. Outcome: The relevance of project objectives and relevance of design was substantial. While infrastructure achievements were in line with the reduced project budget, only about half of households were connected to improved water sources, and efficacy of this objective is rated modest. Conversely, for schemes that were operational, the evidence indicates effective and sustained use of water and sanitation services in rural communities, and the efficacy of this objective is rated substantial. The project efficiency is rated as substantial. The outcome is rated moderately satisfactory. a. Outcome Rating : Moderately Satisfactory 7. Rationale for Risk to Development Outcome Rating: Project beneficiaries were very satisfied with the increased access and 83% of sampled schemes are financially viable. While the infrastructure created was of good quality, there are unresolved concerns in some schemes concerning source protection, water yield, consumption and water quality management (ICR, page 16). At project closing, nascent Community Based organizations continued to require support in revenue and expenditure management, and in mobilizing the large number of unconnected households. The Region in its comments on this Review notes that after project closing, the Government established the National Community Water Trust as an entity to ensure sustainability of Community Based Organizations. In the Region s view this

8 signifies the government s commitment to sustain the model pioneered under the project even without Bank-financing. Although Provincial Councils of the Central and North -Western province have made budget allocations for retaining their Rural Water Supply and Sanitation Divisions beyond the project, there is no indication of such provision being made for the North East province. There is also divided opinion on the future role of Rural water Supply and Sanitation Departments at the provincial and local levels. a. Risk to Development Outcome Rating : Moderate 8. Assessment of Bank Performance: a. Quality at entry: The project was formally reviewed by the Bank s Quality Assurance Group (QAG) for Quality Assessment at Entry in June All aspects of the QAG assessment were rated as satisfactory with the exception of the Bank s inputs and processes. This was rated moderately satisfactory because of poor staff continuity during preparation, and inadequate attention at appraisal to affordability issues, particularly by provincial and local government. Even so, project design was sound and factored -in the experience from the preceding project and of similar projects in India and Nepal. The design of the M&E framework was uneven and some of the indicators included were too ambitious (e.g. for health outcomes). Similarly, the inclusion of the Kirindi Oya river basin water resources study was not particularly relevant to the project's objectives that were already institutionally ambitious. Quality-at at-entry Rating : Moderately Satisfactory b. Quality of supervision: There were two Task Team Leaders over the 91-month project. Despite the institutional nature of the project, a social specialist only started providing inputs two years into implementation. Supervision budget and staff weeks on the project almost doubled in the last 2 years before the project closing (ICR, page 40). According to the project team, the reason for high supervision costs was the rapid growth of the project in its last two years, during which 300 of the 700 water supply schemes were implemented. Thus supervision efforts increased in response to increased needs. Fiduciary and safeguards generally received adequate attention after 2005, although reporting on safeguard issues was weak. Although the development of alternative institutional models for water quality monitoring was not satisfactorily concluded and water quality issues remained a concern, the project team informed IEG that the Government has now put in place a much improved water quality monitoring system. The project team further stated that the metering policy, which was not envisaged for the project, was developed with the Bank's support, and is being implemented. Also, though the formulation of the National Sanitation Policy was not part of the original project, the policy supported by the Bank has been approved by the Cabinet. The Bank Team found the "small town water strategy which was prepared by the consultant as unsatisfactorily. The team later helped the Government complete this strategy by using a WPP Trust Fund. The main shortcoming of supervision was that key indicators and targets were not formally revised after the impact of the Tsunami Fund diversion and national price inflation became clear. Instead, revisions were made unofficially a few months before project closure. Quality of Supervision Rating : Moderately Satisfactory Overall Bank Performance Rating : Moderately Satisfactory 9. Assessment of Borrower Performance: a. Government Performance: Government's commitment was reflected in the formulation of the 2001 National Policy for Rural Water Supply and Sanitation. Significant government reorganization caused the ministry responsible for the project coordination to change four times during implementation (Ministry of Housing and Plantations, Ministry of Urban Development and Water Supply, Ministry of Urban Development and Sacred Area Development, and Ministry of

9 Water Supply and Drainage). Contribution from regional and local government was less than half that expected (2% of investment costs instead of 5%), and there was one instance of delay which caused problems for the District Councils and community-based organizations that had to borrow from moneylenders to continue construction on the basis of expected contributions of the Provincial Councils. Government Performance Rating Moderately Satisfactory b. Implementing Agency Performance: The counterpart agency for the project was Rural Water Supply and Sanitation Division and its offices in the provinces. Staffing continuity was good with the exception of the accounts section which experienced a high turn-over. Quick follow-up was provided in response to Bank mission recommendations, such as drafting of sanitation, tariff and metering policies, revision of cost sharing arrangement, responding to staffing needs in the field and providing implementation support, especially for the NE Region. The Division actively promoted stakeholder consultation, and facilitated linkages between suppliers and community -based organizations. Even though the contribution from regional and local government was less than half that expected (2% of investment costs instead of 5%), beneficiaries willingly contributed more than their allocated share. The implementing agency had problems with ensuring sound M&E, resulting in data issues and limited use of M&E during implementation. Financial compliance lagged with regard to establishment and use of financial management software. Non-compliance with the Bank's safeguard policies necessitated cancellation of a small component. Implementing Agency Performance Rating : Moderately Satisfactory Overall Borrower Performance Rating : Moderately Satisfactory 10. M&E Design, Implementation, & Utilization: a. M&E Design: The M&E design (PAD, page 12) was based on a management information system developed by the Rural Water Supply and Sanitation Division. The capacity to undertake project M&E was not assessed at the provincial and local government levels. It included input, process, output and outcome /impact monitoring, to be coupled with independent audits, periodic health surveys (later omitted on the suggestion of QAG) and independent sociological impact assessments during project implementation and at completion. Water quality, though identified repeatedly as a problem, had no indicators and was not monitored. M&E design did not make a clear distinction of the indicators by sector, project, provincial, and sub -project level and the project impacts could not be disaggregated. Project benefit and beneficiary assessment was missing. Similarly, there were no measures to assess capacity development improvements resulting from the project. b. M&E Implementation: Implementation was weak. The NE province was not monitored as it was not part of the M&E framework. Many planned M&E activities were not undertaken (socio-economic impact assessments at mid-term and project completion, independent audits for technical quality of sub -projects, safeguard compliance monitoring, implementation of grievance redress mechanism etc.). There was a three-year lapse when the old computer-based M&E system was replaced by a new paper -based system; and the data responsibility of M&E shifted to the central government agency rather than local government as envisioned during preparation. Data quality control remained a constant problem due to shift between paper and computer systems, and lack of dedicated staff (ICR, page 9). Even so, there were some valuable outcome assessment surveys conducted towards the end of the project. Indicators for the project were changed to align with the Core International Development Association water sector indicators and only three of the original indicators were retained. Finally, key target values for the indicators were only revised in May 2010, seven months before project closing. The geographical aspect of the project was not integrated into the M&E, and as a result evaluation cannot conclude where the project made the most impact. c. M&E Utilization:

10 There was limited use of M&E for decision-making. M&E Quality Rating : Modest 11. Other Issues a. Safeguards: Environment : The project was classified as an Environmental Category A under OP 4.01 Environmental Assessment at appraisal because the project included a water resource study for Kirindi -Oya river basin in the Hambantota district. However, the required environmental assessment of the Kirindi -Oya river basin was not carried out before starting work, and the Bank cancelled this component. The ICR reports (p. 9) "The project required B categorization, but for a water resource study for Hambantota district, which promoted A categorization. The Hambantota study was dropped and hence Bank policies related to Category A were not applicable". A Sectoral Environmental Assessment was prepared and highlighted the problem of drinking water quality and scarcity. Even so, only limited water source protection was taken up in the project due to lack of funding and low capacity at the local government level. The ICR neither describes the overall monitoring of Environmental safeguards during project implementation, nor indicates if there were any compliance or capacity problems. Social: The project faced no significant social safeguard issues. There is no discussion on the beneficiary assessment proposed to be undertaken during mid-term and at project closing, or of the grievance or redress mechanism to deal with complaints under the project. b. Fiduciary Compliance: Procurement : The project mainly involved small works, undertaken by communities and coordinated by provincial an d local government. Delay in training, compounded by the absence of a dedicated implementing agency procurement officer, adversely affected overall procurement management. According to the ICR (page 10), "at the Credit closing, two cases of misprocurement were declared." The two cases involved 0.7% of total procurement, according to the Borrower's comments on the ICR (ICR, page 47). However, the project team subsequently informed IEG that "the implementing agency made a genuine mistake in procuring two items. The Bank did not declare misprocurement; instead, it made these expenditures ineligible." Financial Management : At appraisal, the Financial Management risk was rated as substantial due to capacity constraints of community-based organizations and local government. High implementing agency staff turnover, and delayed and partial commissioning of the computer based financial management system were the main issues. At project closing there were no formal audit arrangements to assist community -based organization to manage the large overhang of project funds (Sanitation Revolving Fund accumulations, ongoing contributions from households ). According to the ICR (page 10), financial management reports were improved in 2010, and no major issues were reported in the official Audits. Due to the Tsunami in December 2004, US$7.0 million equivalent was reallocated from the project in February 2005 to the IDA-financed Tsunami Emergency Recovery Project. The compliance of these funds with financial auditing requirement is not discussed in the ICR as these funds were disbursed under a separate project (Tsunami Emergency Recovery Project). c. Unintended Impacts (positive or negative): d. Other:

11 12. Ratings: Outcome: Moderately Satisfactory Risk to Development Moderate Outcome: ICR IEG Review Moderately Satisfactory Moderate Reason for Disagreement /Comments The relevance of project objectives and relevance of design was substantial. While infrastructure achievements were in line with the reduced project budget, only about half of households were connected to improved water sources, and efficacy of this objective is rated modest. Conversely, for schemes that were operational, the evidence indicates effective and sustained use of water and sanitation services in rural communities, and the efficacy of this objective is rated substantial. The project efficiency is rated as substantial. The outcome is rated moderately satisfactory. Bank Performance : Satisfactory Borrower Performance : Satisfactory Quality of ICR : Moderately Satisfactory Moderately Satisfactory Unsatisfactory Project output targets should have been made more realistic following the impact of the Tsunami. There were only partial arrangements to ensure completion of household connections after the project closed. The implementing agency had problems with ensuring sound M&E, resulting in data issues and limited use of M&E during implementation. Financial compliance lagged with regard to establishment and use of financial management software. Non-compliance with the Bank's safeguard policies necessitated cancellation of a small component. NOTES: - When insufficient information is provided by the Bank for IEG to arrive at a clear rating, IEG will downgrade the relevant ratings as warranted beginning July 1, The "Reason for Disagreement/Comments" column could cross-reference other sections of the ICR Review, as appropriate. 13. Lessons: The following four lessons are derived from the ICR : Adequate financial analysis of provincial and local governments ability to contribute should be assessed prior to implementation. In this project, these levels of government provided less than half the amount stated at appraisal and the shortfall had to be made up from additional community contributions.

12 The Sanitation Revolving Fund model can be effective. Because of its innovative design the Fund was able to mobilize community action to improve sanitation access and leverage funds for other economic activities. Including conflict -affected areas within a national project requires flexibility in project design and implementation arrangements. Special arrangements are required to take into account the differing socio-economic and institutional challenges present in conflict -affected areas - one set of arrangements for a project will not fit all circumstances. And IEG adds this lesson: When projects are significantly affected by natural disasters attention should be given to adjusting expected outputs and outcomes as soon as possible. In this project this was not done until the very end and a consequence was that expectations became unrealistic given the reduced budget. 14. Assessment Recommended? Yes No Why? The project transition arrangements were not in place at project closing, it would be interesting to learn what course the project sustainability took. Also, to verify increased household connection rate. 15. Comments on Quality of ICR: The overall quality of the ICR was uneven. The description of components is inadequate and the subsequent discussion is overly focused on implementation and outputs with not enough attention to outcomes. The rationale for revising project targets is neither clear nor objective. Compliance with financial auditing requirement is not discussed adequately, While the report on the findings of the independent audit on outcomes was useful, it is not clear when the audit took place, the methodology used, and why the NE Province was not included. Safeguard and financial management issues should have been more thoroughly discussed. a.quality of ICR Rating : Unsatisfactory