Cambodia: Road Asset Management Project

Transcription

1 Completion Report Project Number: Loan Number: 2406-CAM June 2015 Cambodia: Road Asset Management Project This document is being disclosed to the public in accordance with ADB's Public Communications Policy 2011.

2

3 CURRENCY EQUIVALENTS Currency Unit riel (KR) At Appraisal At Project Completion (5 December 2007) (13 February 2015) KR1.00 = $ $ $1.00 = KR3, KR4, ABBREVIATIONS ADB Asian Development Bank BME benefit monitoring and evaluation DMF design and monitoring framework IDA International Development Agency IPA independent procurement agent km kilometer MBPI merit-based pay incentive MPWT Ministry of Public Works and Transport OFID OPEC Fund for International Development POC priority operating cost PMU project management unit RAMO Road Asset Management Office RAMP Road Asset Management Project NOTE In this report, "$" refers to US dollars unless otherwise stated. Vice-President S. Groff, Operations 2 Director General J. Nugent, Southeast Asia Department (SERD) Director H. Iwasaki, Transport and Communications Division, SERD Team leader Team members S. Date, Senior Transport Specialist, SERD C. Clark, Safeguards Specialist (Resettlement), SERD M. Javier, Senior Project Assistant, SERD E. Lara, Safeguards Officer (Environment), SERD E. R. Nava, Operations Assistant, SERD A. Veron-Okamoto, Transport Specialist, SERD In preparing any country program or strategy, financing any project, or by making any designation of or reference to a particular territory or geographic area in this document, the Asian Development Bank does not intend to make any judgments as to the legal or other status of any territory or area.

4

5 CONTENTS BASIC DATA Page i I. PROJECT DESCRIPTION 1 II. EVALUATION OF DESIGN AND IMPLEMENTATION 1 A. Relevance of Design and Formulation 2 B. Project Outputs 3 C. Project Costs 5 D. Disbursements 6 E. Project Schedule 6 F. Implementation Arrangements 6 G. Conditions and Covenants 7 H. Consultant Recruitment and Procurement 7 I. Performance of Consultants, Contractors, and Suppliers 8 J. Performance of the Borrower and the Executing Agency 8 K. Performance of the Asian Development Bank 8 III. EVALUATION OF PERFORMANCE 9 A. Relevance 9 B. Effectiveness in Achieving Outcome 9 C. Efficiency in Achieving Outcome and Outputs 9 D. Preliminary Assessment of Sustainability 9 E. Impact 9 IV. OVERALL ASSESSMENT AND RECOMMENDATIONS 10 A. Overall Assessment 10 B. Lessons 11 C. Recommendations 12 APPENDIXES 1. Project Framework Status of Compliance with Loan Covenants Project Cost at Appraisal and Completion Outputs and Activities by Funding Source 26

6

7 BASIC DATA A. Loan Identification 1. Country 2. Loan Number 3. Project Title 4. Borrower 5. Executing Agency 6. Amount of Loan 7. Project Completion Report Number B. Loan Data 1. Appraisal Date Started Date Completed 2. Loan Negotiations Date Started Date Completed 3. Date of Board Approval 4. Date of Loan Agreement 5. Date of Loan Effectiveness In Loan Agreement Actual Number of Extensions 6. Closing Date In Loan Agreement Actual Number of Extensions 7. Terms of Loan Interest Rate Maturity (number of years) Grace Period (number of years) 8. Terms of Relending (if any) Interest Rate Maturity (number of years) Grace Period (number of years) Second-Step Borrower Kingdom of Cambodia 2406-CAM(SF) Road Asset Management Project Kingdom of Cambodia Ministry of Public Works and Transport Special Drawing Rights 3,761,000 PCR: CAM September October November November January February May December December February % per annum during the grace period and 1.5% per annum thereafter 32 years 8 years None

8 ii 9. Disbursements a. Dates Initial Disbursement 17 June 2009 Effective Date 18 December 2008 Final Disbursement 25 September 2014 Original Closing Date 31 December 2013 Time Interval months Time Interval months b. Amount ($5.56 million) Category or Subloan Original Allocation Last Revised Allocation Amount Canceled Net Amount Available Amount Disbursed Undisbursed Balance Civil Works-Contract AWP-CW-1.1 Civil Works-Contract AWP-CW-2.1 Civil Works-Contract AWP-CW-3.1 Civil Works-Contract AWP-CW-4.1 Training (Net Of Tax) Consulting Services - 1st Year Implementation Consultant (Net Of Tax) Consulting Services Implementation of M&E Activity New Subcategory. (Net of Tax) Interest Charge Unallocated Imprest Account Total AWP = annual work plan; CW = civil works; M&E = monitoring and evaluation Notes: 1. The difference between the original allocation and the revised total amount is due to the exchange rate variation between special drawing rights (SDR) and the US dollar. 2. The undisbursed loan amount of SDR151, ($214, equivalent) was canceled at loan account closing (13 February 2015). The difference in the amount canceled indicated in the table and actual canceled amount indicated in note 1 is due to rounding. C. Project Data 1. Project Cost ($ million) Cost Appraisal Estimate Actual Foreign Exchange Cost Total Financing Plan ($ million) Cost Appraisal Estimate Actual Implementation Costs Borrower Financed ADB Financed Government of Australia Grant IDA OFID Subtotal IDC Costs ADB Financed

9 iii Cost Appraisal Estimate Actual Subtotal Total ADB = Asian Development Bank; IDA = International Development Agency, IDC = interest during construction, OFID = OPEC Fund for International Development. 3. Cost Breakdown by Project Component ($ million) Component Appraisal Estimate Actual Civil Works Implementation Consultants First Year Implementation Consultant Supervision Consultant Project Management Advisor Institutional Strengthening Advisory Technical Audit Consultant Impact Monitoring Consultant Implementation of Monitoring and Evaluation Activity General HIV/AIDS Action Plan Resettlement Decree Road Safety Action Plan Merit-Based Pay Incentive Project Management Taxes (14% of civil works costs) Contingencies Financing Charges During Implementation Total Project Schedule Item Appraisal Estimate Actual Civil Works Contract Civil Works-Contract AWP-CW-1.1 Date of Award Completion Civil Works-Contract AWP-CW-2.1 Date of Award Completion Civil Works-Contract AWP-CW-3.1 Date of Award Completion Civil Works-Contract AWP-CW-4.1 Date of Award Completion Q Q Q Q June June March August February February December June 2014 Consultants 1st Year Implementation Consultant Q March st Year Implementation Consultant (replacement) Q July st Year Implementation Consultant (2nd replacement) Q March 2010 Implementation of Monitoring and Evaluation Activity Q April 2014 AWP = annual work plan; CW = civil works; Q = quarter. 5. Project Performance Report Ratings Ratings Implementation Period Development Objectives Implementation Progress From 16 May 2008 to 31 December 2008 Satisfactory Satisfactory From 1 January 2009 to 31 December 2009 Satisfactory Satisfactory From 1 January 2010 to 31 December 2010 Satisfactory Satisfactory From 1 January 2011 to 31 December 2011 Satisfactory Satisfactory From 1 January 2012 to 31 December 2012 Satisfactory Satisfactory From 1 January 2013 to 31 December 2013 Satisfactory Satisfactory From 1 January 2014 to 30 June 2014 Satisfactory Satisfactory

10 iv D. Data on Asian Development Bank Missions Name of Mission Date No. of Persons No. of Person-Days Specialization of Members a Fact finding Mission June a, d, f Appraisal Mission 26 Sep 9 Oct a, d, f, l, m, n Inception Mission Oct a, b, Review Mission 3 5 Feb a, b, e Resettlement Review Feb k Review Mission May a, b Review Mission Oct a, b, e Review Mission Feb a, b Review Mission May a, b, c, i, Review Mission Nov a, b, e Review Mission Mar a, b Review Mission Oct a, b Mid Term Review Mission Feb a, b Review Mission 4 8 Jun a Review Mission 3 7 Dec a Review Mission May f, g, h Review Mission Oct f Review Mission Mar f, g Project completion review 2 9 Feb f, g a a = transport economist and team leader, b = associate project analyst, c = sector director, d = senior project officer, e = resettlement specialist, f = senior transport specialist, g = senior project assistant, h = safeguards officer (environment), j = consultant (resettlement monitoring specialist), k = safeguards specialist (resettlement), l = counsel, m = financing partnership specialist, n = staff consultant.

11 I. PROJECT DESCRIPTION 1. On 21 January 2008, the Asian Development Bank (ADB) approved a loan of about $6 million equivalent from its Special Funds resources to the Government of Cambodia for the Road Asset Management Project. The objectives of the project were to (i) assist the government in periodic maintenance of about 950 kilometers (km) of the national and provincial roads administered by the Ministry of Public Works and Transport (MPWT); (ii) help build capacity at MPWT to improve the maintenance and management of its roads network; and (iii) provide institutional and management support and training to MPWT to strengthen its road asset management and quality assurance capacity at all levels, and to start a process of privatizing its force account units. 2. The institutional strengthening of MPWT included supporting a permanent and dedicated road asset management office (RAMO) within the general department of public works of MPWT. RAMO was to be responsible for systematic maintenance planning for the MPWT roads, and providing support and capacity development programs to MPWT to strengthen road asset management. 3. Apart from ADB financing, project co-financing was provided by (i) the Government of Australia, in the form of a grant of A$6.0 million ($4.80 million equivalent); and (ii) the OPEC Fund for International Development (OFID), which provided additional cofinancing of $7.0 million in 2009 to expand the scope of the project, based on the assumption that price escalation would not absorb the additional financing. 1 The International Development Agency (IDA) provided a credit of $30.0 million as parallel financing. The Government of Cambodia was to finance about $17.56 million, making the project cost $58.36 million. The project was completed on 30 June 2014, after a 6-month extension from its original completion date of 31 December II. EVALUATION OF DESIGN AND IMPLEMENTATION 4. The project had two components. The road maintenance component (A) encompassed two activities: periodic maintenance works implementation, and support to the implementing agency in managing and implementing these works. The capacity-building component (B) encompassed strengthening of MPWT s road management processes, systems, and staff capacity to undertake effective road asset management; privatization of five to six of MPWT s force account units; and provision of financial support to general programs aimed at mitigating negative impacts of road traffic. Component B was also intended to introduce a merit-based pay incentive (MBPI) in MPWT An indicative annual works program was prepared for the project s periodic maintenance works. These planned works were resealing, associated preparatory works such as crack and pothole repairs, and ancillary works such as shoulder reinstatement, line work, and installation of road signs on about 950 km (footnote 2) of the MPWT roads. Shoulder reconstruction and sealing was not normally considered part of periodic maintenance, but where necessary to secure the integrity of the network, up to 20% of the annual works cost could be used for this. The project did not intend to implement any works that would require resettlement. Force 1 The intention was to use the additional financing to increase the length of the road to undergo maintenance under the project by an additional 250 km, from 950 km (para. 5). However, the global financial crisis in 2009 caused an unusual surge in the price of construction materials, which absorbed all $7 million of additional financing. 2 Merit-based pay aimed to reward the work ethic and effectiveness of government employees. It was a joint initiative of the government and its development partners, including ADB and IDA.

12 2 account units were not intended to be employed for project works. All works performed under the project were planned to be implemented by private contractors, through competitive bidding. 6. The proposed first-year work program was firm, but the work programs for the subsequent 3 years were for planning purposes only. RAMO was to select road sections for periodic maintenance in these years, based on regular monitoring of road condition. The condition of the network is affected by the amount of traffic on the individual roads, the impact of maintenance activities undertaken outside the project, and future events, such as larger than normal seasonal floods. The ADB loan and the Government of Australia grant were to finance the first-year works program on National Road 5 from km 3.9 to the Kampong Chhnang western city limit, through two contract packages. In subsequent years it was anticipated that ADB and the Government of Australia would finance periodic maintenance of National Road 5 from Kampong Chhnang to Pursat. 7. Implementation support to RAMO was to encompass (i) a project implementation advisor to assist RAMO with project implementation; (ii) works supervision; (iii) a short-term consultant to assist RAMO prepare the first-year work program, encompassing contracting of civil works and contracting of supervision consulting services; (iv) an intermittent consultant to perform regular independent technical audits; and (v) independent financial audits of the project accounts. The ADB loan and the Government of Australia grant were to finance the short-term consultant preparing the first-year program and the technical audit consultant. Also, the Government of Australia grant was to finance a road safety intervention to mitigate negative impacts of road traffic, and resettlement training to the Interministerial Resettlement Committee established under the Ministry of Economy and Finance. IDA credit was to finance the remaining activities, including a priority operating cost (POC) reform (para. 8), for management and project implementation. 8. It was proposed that the MBPI would be replaced by a scheme to be introduced in This POC, planned in 2010, was intended to be an interim provision for replacing the MBPI, which at that time was proposed to support project staff incentives. POC was an initiative of the government s administrative reform program. A. Relevance of Design and Formulation 9. ADB s strategy for Cambodia s road subsector has emphasized restoration of the national road network to provide safe and economical travel between the main centers of population and economic activity, and to provide reliable land transport links to neighboring countries. The strategy has focused on road infrastructure rehabilitation and development of the institutional capacity to plan, manage, and maintain the transport system. The project was relevant to and consistent with ADB s country strategy and program The project was relevant to one of three strategic pillars in the Government s Rectangular Strategy. 4 ADB s overarching goal of sustainable poverty reduction was also consistent with the government s strategy. 3 ADB Country Strategy and Program: Cambodia, Manila. 4 Specifically, broad-based economic growth through investment in physical infrastructure, development of the financial sector, support for greater regional integration, sustainable development of small and medium-sized enterprises, and investments in agriculture and irrigation. Prime Minister Samdech Techo Hun Sen Rectangular Strategy for Growth, Employment, Equity and Efficiency, Phase II. Phnom Penh.

13 3 11. The project design was sound, and the project formulation process was adequate, with participation by all relevant stakeholders especially the Borrower, the executing agency, and the development partners in support of a stronger MWPT road maintenance regime. The project design generated a satisfactory level of ownership by MPWT. 12. Relevance of the project at appraisal became evident at completion through increasing reliance on the road network for the growing Cambodian economy. The project preparatory technical assistance (TA) designed two projects: 5 the project, approved in 2008, followed by the Greater Mekong Subregion Cambodia Northwest Provincial Road Improvement Project (approved in 2009; completed in 2014). 6 Although the latter project focused on rehabilitation of a provincial road, it supported sustainability of the MPWT road maintenance regime through one relevant output. Overall, the project preparatory TA was efficient, and prepared a relevant project design. 13. Resettlement training and POC financing had limited relevance, with the latter relevant only to the government s administrative reforms. B. Project Outputs 14. Component A: road maintenance. With financing from ADB, the Government of Australia, and OFID, MPWT successfully executed four contract packages. As planned at appraisal, two contracts were designed for the first year works, and two contracts were designed for works for subsequent years. Under these four contracts about 767 km of road maintenance work out of a target of 950 km were completed. The works were completed at various stages during project implementation, with the contracts awarded sequentially over a period of 4 years. 15. The ADB project completion mission found the works satisfactory overall, with the 4 respective road sections currently in good condition. Flooding in 2011 damaged some sections of National Road 5, and these were repaired in a timely manner through contract variations. However, technical audits have found that road markings are of poor quality. Thermoplastic paints and glass beads were used to aid night visibility, but despite the use of these materials, road markings were obliterated by bitumen stains from tire marks soon after painting. Therefore, the road markings do not seem to be effective. More attention needs to be applied to safety aspects of the roads. 16. Regarding IDA-financed road maintenance works, all seven contracts (for about 472 km of road maintenance) were completed over 4 years. During appraisal, the indicative length of road maintenance works planned for IDA financing was about 500 km. Although IDA-financed road maintenance appears not to have been achieved, the project completion review mission observed that maintenance of the 28 km that was apparently not financed by IDA may have been financed by the government (see para. 26). However, this distinction of roads lengths by financier is not clearly stated in the report and recommendation of the President. 7 The IDA 5 ADB Technical Assistance to the Kingdom of Cambodia for the Transport Infrastructure Development and Maintenance Project. Manila (TA 4691-CAM). 6 ADB Report and Recommendation of the President to the Board of Directors: Proposed Loan to the Kingdom of Cambodia for the Greater Mekong Subregion Cambodia Northwest Provincial Road Improvement Project. Manila (Loan 2539-CAM). 7 ADB Report and Recommendation of the President to the Board of Directors: Proposed Loan and Administration of Grant from the Government of Australia to the Kingdom of Cambodia: Road Asset Management Project. Manila (L2406-CAM).

14 4 mission for implementation completion, conducted 24 March to 3 April 2015, concluded the completed civil works quality is satisfactory. 17. Component B: capacity development. This component comprised six major activities: (i) organization of capacity development for road asset management, (ii) asset management technical capacity development, (iii) private sector development, (iv) provision of financial support to general programs aimed at mitigating negative impacts of road traffic, (v) resettlement training, and (vi) provision of financing for MBPI in MPWT. 18. The first two activities were fully implemented. The Road Data Collection and Management Unit, established under the newly established RAMO, received extensive training. Training was provided by international consultants on road condition data collection, data input, maintenance programs preparation, prioritization of programs for annual maintenance, preparation of budget proposals, construction supervision, quality assurance, and quality control. Nevertheless, the project completion review mission found that MPWT engineers who lack systematic training on conducting road maintenance to meet international standards failed to acquire sufficient knowledge under the project to ensure sustainable road asset management. Also, reporting capacity requires improvement, with regards to reporting recommendations and providing sound rationales for budget proposals and submissions. 19. Regarding activity (iii) (private sector development), the training provided for the private contractors under the project clearly improved the capacity of the contractors, which could be seen in bid preparations and submissions, compliance with financial requirements, and submission of claims during implementation. However, the MPWT force account units have not yet outsourced road maintenance activities as intended by the project. The first contracts of this nature, amounting to $5.2 million, are currently in the bidding process, with contract award expected during Activity (iv) (mitigating negative impacts of road traffic) was almost fully implemented, with road safety education for children of grades already completed. The guidelines for pedestrian safety have been corrected by including in the school curriculum guidance for students in the absence of a separate pedestrian sidewalk to walk on the road carriageway, against vehicular traffic. These materials are now being circulated. The practical training for students in the field will be implemented thereafter, in coordination with the Ministry of Education Youth and Sport. Road safety awareness campaigns focusing on reducing speeding and promoting the wearing of helmets were completed for medium-level management staff of relevant government agencies. 21. Activity (v) (resettlement training) was not implemented, because the officials from the line ministries and provincial authorities showed no interest in participating. Activity (vi) (funding for initiating the POC arrangements in MPWT) was proposed by the government in 2010 as an interim solution to abolishing the MBPI scheme for project management staff. The process of financing the POC was not clear at the outset, and was delayed by government formulation of clear guidelines on POC implementation, which was intended to be financed by the development partners financing the project. Although the government planned to implement this POC only until June 2012 as an interim measure, actual payments were only possible from mid as a result of the delay. Therefore, project management staff received no payments for about 18 months. This resulted in savings of special drawing rights (SDR) 450,000, from IDA resources, which was subsequently allocated for flood damage repairs and rehabilitation of National Road 7.

15 5 22. In general, the implementation of all project components had no serious efficiency issues, except for the delays caused by heavy flooding in areas where the project s civil works were ongoing. To compensate for these delays, the loan closing date was extended for 6 months, until 30 June The project benefits expected from the two components may be considered substantially achieved, with the exception of the target road length of 950 km to be maintained under the project. The project instead achieved about 767 km. This underperformance resulted from the non-award of contracts via government financing, through outsourcing force accounts (paras. 21 and 29). 23. Environment safeguards. Although previous loan review missions observed minor issues of non-compliance with environment safeguards for civil works implementation, 8 during field visits the project completion mission noted that all work-closing requirements related to environment safeguards compliance were properly completed. 24. HIV Prevention Program. During project design, a $200,000 allocation for an HIV/AIDS prevention program, financed under IDA, was proposed. During implementation MPWT discovered the allocation was insufficient to complete the program to cover civil works for year 4, but no additional financing allocation was made. Therefore, this program was partly successful. C. Project Costs 25. The project cost at appraisal was $58.36 million (including taxes and duties, and charges during implementation) while the actual project cost was $47.62 million. The original financing plan included an ADB loan of SDR3,761,000 (about $6 million equivalent) from ADB s Special Funds resources (about 10.3% of the total project costs), and the Australian government provided a grant of A$6 million ($4.8 million equivalent) (about 8.2% of the total project costs). IDA provided a credit of $30.0 million (about 51.4% of the total project costs) as parallel cofinancing. In 2009, OFID provided additional cofinancing of $7.0 million (about 12.0% of the total project costs). The Government of Cambodia was to finance about $17.56 million of the project cost (about 18.1% of total project costs). Financing details (at appraisal and completion) are shown by outputs and activities in Appendix 3, and by the financiers who financed the outputs and activities in Appendix The difference of projected project costs at appraisal and actual costs at completion is due mainly to non-implementation of outsourcing of the force accounts under government financing. The reason was an agreement reached between IDA and the Ministry of Economy and Finance (MEF) that outsourcing of the force accounts might not be necessary. This was based on a misunderstanding that MPWT is already outsourcing the force account for road maintenance contracts. However, MPWT's outsourcing was actually in compliance with requirements that pre-dated project appraisal (outsourcing was not on a competitive basis; and lacked proper documentation, transparency, accountability, and supervision or quality control by MPWT). Also, the aforementioned agreement between IDA and MEF was not known to MPWT's project director, or other financiers and/or co-financiers. As a result, when it was revealed during a loan review mission, it created an atmosphere of mistrust among parties involved in project implementation that was not remedied until project completion. In addition, other minor activities were not fully implemented, 9 which resulted in cost savings, but did not impact the project 8 Issues of non-compliance comprised inadequate road safety during construction, workers not wearing safety gear, and minor housekeeping shortcomings (e.g., spills of fuel and bitumen in camp sites). 9 Costs savings included allocations for resettlement training, un-contracted civil works costs, and unused project management costs. These funds were not used for a number of reasons. For resettlement training the allocation

16 6 outcome. D. Disbursements 27. The loans and grant proceeds were disbursed in accordance with the ADB Loan Disbursement Handbook (2012, as amended from time to time) through reimbursement, the statement of expenditure method, and direct payment procedures. 10 MPWT established an imprest account for consulting services, with the statement of expenditures procedure used for liquidation and replenishment. For civil works contracts, the direct payment procedure was applied. 28. At project completion, utilization of financing was as follows: ADB loan: $5.564 million (92.7%), including $0.11 million in interest during construction, out of the original allocation of $6 million; Australian government grant: $4.598 million (95.8%), out of the original allocation of $4.8 million; and OFID loan: $6.984 million (99.8%), out of the original allocation of $7 million. ADB finalized the financial data with MPWT and closed the loan account on 13 February 2015, and the grant account on 27 March E. Project Schedule 29. ADB approved the loan on 21 January 2008, and the loan agreement was signed on 18 February Loan effectiveness was extended twice, because the IDA credit was not approved until 21 May 2008, and the agreement signed on 13 June The loan became effective on 18 December The original loan closing date was 31 December However, the project suffered delays, mainly caused by delays in consultant recruitment during initial stages of project implementation, and heavy flooding in 2011 in areas where project civil works were ongoing. 11 To compensate for this delay, the loan closing date was extended for 6 months, until 30 June The delay was not significant. F. Implementation Arrangements 30. As at appraisal, MPWT was the executing agency and responsible for the overall technical supervision and execution of the project. The Directorate General of Public Works was the implementing agency and responsible for direct project supervision and day-to-day project implementation through the newly established RAMO. This implementation arrangement did not change until project completion. 31. RAMO was staffed from various divisions, and the employees day to-day tasks included other work required by their original divisions. This was a serious issue for staff, in that they had to follow orders from more than one chain of command, and resulted in delays in performing project tasks. This was clearly the case in financial and accounting management, administration and procurement. Project Management Unit 3 (PMU3) which has administered all ADB-financed road projects, helped overcome this problem. 32. There were initial problems recruiting the targeted 75 staff for implementation startup was not used because coordination between MPWT and IRC was not effective. The other major reason was the difficulty that MPWT faced in accurately projecting the project cost at completion, and using the loan and/or grant proceeds efficiently. 10 ADB disbursed the loans and grants of ADB, the Government of Australia, and OFID, while IDA disbursed their own loan proceeds. 11 Although there were initial start-up delays, implementation was on schedule in 2011, when the floods occurred.

17 7 from the MPWT structure because of: (i) lack of interest on the part of staff to work on the project, as they would not receive incentive payments, compared to what staff working for the PMUs would receive; (ii) the delay in POC implementation; and (iii) difficulty in finding qualified staff from within MPWT for road asset management. Therefore, only five key staff members were recruited during the first year of implementation. This increased to 75 during the project mid-term, when POC was available, and staffing remained at that level for the rest of the implementation period. 33. Overall, the MPWT structure proposed for project implementation was not fully effective or efficient. However, the project director, who happened to be the project manager for PMU3 (and subsequently became its director), exhibited strong ownership and dedication to overcome RAMO lapses with PMU3 support. He made efforts to coordinate among development partners, MPWT, and the Borrower to resolve confusion that arose as a result of the independent procurement agent (IPA), the POC, and HIV/AIDS program financing. This resulted in satisfactory project implementation. G. Conditions and Covenants 34. There were no significant delays in meeting the conditions of loan effectiveness through attention to procedural requirements in Cambodia. All loan covenants were assessed to be relevant. All general and specific sector, financial, environmental, social, and other covenants were complied with. No covenant compliance was delayed or breached, and no covenant has been modified, suspended, or waived. However, covenants do not appear to be detailed or specific enough to ascertain the sustainability of MPWT s road maintenance regime, and the future of the project initiatives. 35. The Borrower and executing agency generally complied with reporting requirements. There were no changes in amortization between the Borrower and executing agency, or conversion of the loan to equity. H. Consultant Recruitment and Procurement 36. There were no issues pertaining to procurement activities under contracts financed by ADB, the Australian government, and OFID. However, the IPA method adopted by IDA experienced delays as a result of the procurement decision-making process and lack of capacity in IPA. MPWT found this approach inefficient. These delays affected activities funded by ADB, the Australian government, and OFID, as well as IDA-financed activities, because of interdependency in implementation. 37. A critical delay in the recruitment of the implementation adviser adversely impacted the early stages of implementation, because the services included engineering design and preparation of bidding documents. Following the failure of the consultant that was initially appointed, and termination of their contract, an implementation adviser was eventually fielded in March 2012, almost 3 years after project commencement. Recruitment of the supervision consultant was delayed until February 2011, about 15 months after the award of the year 1 civil works contract. 38. These delays were mitigated by the flexibility of the ADB-administered activities, which enabled the first-year implementation consultant to be engaged for the intended services. The consultant was originally appointed for a 3-month period to facilitate quick startup for the design of the year 1 civil works, but extended to cover supervision of the works, and preparation of the

18 8 year 2 and year 3 engineering designs. Eventually, the consultant s services were extended for over 2.5 years, and this mitigated the delay in recruitment of the implementation adviser and the supervision consultant teams. This arrangement could not provide full technical support for those consultant teams, but enabled progress in the year 1 civil works, and facilitated the bidding of year 2 and year 3 civil works contracts. In addition, an intermittent consultant to perform regular independent technical audits was also recruited. I. Performance of Consultants, Contractors, and Suppliers 39. Although there were some lapses in the performance of consultants, there were no issues with respect to the performance of contractors and suppliers. Overall, performance was satisfactory. 40. Consultants. The initial individual consultant recruited for benefit monitoring and evaluation (BME) did not perform well, and left the assignment unfinished, without even submitting an inception report. A second consultant was recruited, delaying the BME process. This consultant completed the baseline report in 2012, which was the midpoint of project implementation. The BME consultant completed the final report in September 2014, and updated it in February 2015, but it still contained errors because of a lack of attention to detail. This provided valuable information for the project completion report. MPWT mobilized the technical audit consultant to complete the audit of completed works. The consultant only audited 90% of completed works. Apart from these two BME consultants, others performed satisfactorily to contribute to implementation of the project. 41. Contractors. The contractors executed the works satisfactorily, on time and of satisfactory quality. The four road sections maintained by the project are currently in good condition. Due to flooding in 2011, some sections of National Road 5 were damaged, but these were repaired in a timely manner through contract variations. However, the technical audits (para. 15) found that road markings are of poor quality (road markings were obliterated by bitumen stains from tire marks soon after painting). J. Performance of the Borrower and the Executing Agency 42. The Borrower and executing agency performed satisfactorily in implementing the project. The assessment of the executing agency's capabilities at appraisal was reasonably accurate. The executing agency was highly proactive, technically competent, and closely monitored procurement processes and civil works to avoid delays. 12 The executing agency cooperated well with ADB and other parties in resolving issues that did arise. There were no issues regarding counterpart financing or co-financing. Therefore, overall performance of the Borrower and executing agency is rated satisfactory. K. Performance of the Asian Development Bank 43. ADB performed well in implementing the project. ADB acted promptly during approvals, disbursements, and monitoring, and there were no disagreements with the Borrower or executing agency on terms of reference, bid documents, awards, or other matters affecting the implementation procedures, project costs, or implementation schedule. Further, ADB assisted the Borrower and executing agency to resolve project implementation issues constructively. Therefore, overall performance of ADB is rated satisfactory. 12 The delays caused by the use of IPA were beyond the control of the executing agency.

19 9 III. EVALUATION OF PERFORMANCE A. Relevance 44. The approach of preserving MPWT s existing road stock in an optimal manner through a road maintenance intervention, such as the project, was very timely in terms of Cambodia s economic development. Although the project did not focus on the sustainability of financing MPWT s road maintenance regime, the project laid the foundation of a solid initiative that needs continuous further support. The project initiatives attracted a greater allocation of financing for road maintenance from the government. 13 However, the project design should have identified the partnership issues of the IDA for a better preparedness in project implementation. This might have reduced the delays in implementation. Given this background, the project is rated relevant. B. Effectiveness in Achieving Outcome 45. As the initial step to resolve the issue of road maintenance sustainability, the project focused on improved road maintenance planning, and road maintenance, with less attention to the sustainability of financing for road maintenance. The overall approach was considered optimal at project design, and all monitoring framework indicators were achieved (para. 54). The project is therefore rated effective in achieving its outcome. C. Efficiency in Achieving Outcome and Outputs 46. The project achieved its outcome with a delay of 6 months, as a result of delays in consultant recruitment, and natural events beyond the control of MPWT, ADB, or other project stakeholders. The economic internal rate of return of the project investments is 87%, compared with the 97% projected at appraisal. Therefore, the project is rated efficient. D. Preliminary Assessment of Sustainability 47. Project sustainability is less than likely. The project not have much room to focus on sustainability of MPWT s maintenance regime in terms of sustainable financing, given the project being first such similar intervention. This cannot be considered a serious design shortcoming it was essential to first focus on the necessary technical aspects of a sustainable road asset management regime for MPWT and achieving sustainability requires further interventions, with a stronger focus on sustainable financing gradually. E. Impact 48. Overall, the project had some impact on poverty reduction, and economy, because timely road maintenance has resulted in a more efficient road network for year-round access. There were limited environmental impacts during construction (e.g., noise and dust). However, there were almost no social impacts, as there was no resettlement, and the HIV prevention program focused on minimizing related risks during project implementation. The project included environmental and safeguard measures that were effective in preventing accidents, although there were minor non-compliance issues during implementation. 13 In 2007, when the project was conceived, the maintenance budget of MPWT was about $7.0 million. This increased over 7-fold by project completion in 2014.

20 The project was classified as safeguard category C for involuntary resettlement and indigenous peoples throughout implementation, confirming that no impacts eventuated, as anticipated during appraisal. The grievance redress mechanism was established (Appendix 2, item 17), but there were no complaints received. IV. OVERALL ASSESSMENT AND RECOMMENDATIONS A. Overall Assessment 50. The analysis of the design and monitoring framework (DMF) and the BME is as follows. 51. Impact. The project contributed to sustainable economic growth; although Cambodia s gross domestic product growth declined by 0.1% due to the global financial crisis during 2008 and 2009, gross domestic product growth from 2010 to 2014 was 6% 7%. The percentage of national roads that is paved increased from 26.3% in 2009 to 71% in This resulted from the impact of the investment in infrastructure development; it was partly contributed by the project, but the precise impact is difficult to quantify. The traffic volume on project roads was about 3,000 15,000 vehicles per day in 2014, an average increase of 19% from 2008, when the range was about 1,800 15,000 vehicles per day. The asset value of MPWT s rehabilitated roads has more than doubled, from $1.1 billion in 2008 to $2.5 billion in Outcome. Compared with the project outcome at appraisal (MPWT effectively and systematically plans and implements road asset management on its rehabilitated road network; the Ministry of Economy and Finance budgets accordingly and disburses without delay; and road safety is improved), the project s achievements have been: (i) $2.9 billion savings in vehicle operating costs over the project period; (ii) the percentage of roads in good to average condition has increased as follows: primary roads, from 57% in 2008 to 89% in 2014; secondary roads, from 35% to 52%; and provincial roads, from 6% to 53%; (iii) private sector participation in road maintenance increased only marginally, from 10.9% in 2009 to 11.1% in 2014; this is much less than the target of 50%, which was unrealistic given the slow improvements in capacity; and (iv) the casualty rate target of the project for 2014 was 12 casualties per 10,000 vehicles (for all of Cambodia), which was achieved, as the casualty rate in 2014 was just 7.7 per 10,000 vehicles. However, this resulted from high rates of vehicle registration, rendering the indicator inappropriate. The casualty rate per 100,000 inhabitants has increased, and the project can be considered to have had no impact on road safety, particularly in light of the small road safety program implemented under the project, against what would be required to have a nationwide impact. 53. Regarding regular routine, recurrent, and periodic maintenance instituted on MPWT s rehabilitated road network, the project s contribution has been as follows: (i) under-achievement of the targeted periodic maintenance of 950 km of roads, as only 767 km of roads were maintained by the project. This resulted from the failure to privatize the force accounts; (ii) regarding the indicator of rehabilitated roads receive adequate routine, recurrent, and periodic maintenance by 2010, maintenance budget provisions increased steadily during , but the budget is not satisfactory. A revised indicator is needed (e.g., funding per km); and (iii) regarding the indicator of outlays for repair of failed pavements and emergency repair of flood damage on the rehabilitated road network are reduced by 50% by 2013, the budget provision was $0 for 2008, and increased steadily to $5 million in The incidence of flood damage has increased yearly due to climate change, making this indicator meaningless to assess the quality of road maintenance.

21 RAMO is established and operating effectively, demonstrating completion of civil works, contracts on schedule, in compliance with conditions of contract and all defects rectified by end of defects liability period. RAMO is headed by a project director, a project manager, an office manager, and has a dedicated Finance Unit with supporting staff. RAMO used the existing MPWT set up with clear delineation of responsibilities and authorities. 55. Adequate pavement design and quality of works were achieved during maintenance of the rehabilitated network, through systems established to effectively implement: (i) identification of work needs, (ii) preparation of works, (iii) procurement, (iv) contract administration, (v) civil works supervision and management, (vi) monitoring and reporting, (vii) financial accounting, and (viii) technical auditing. 56. Outputs related to completion of road safety campaigns to raise public awareness and improve enforcement of traffic and road right-of-way have been achieved through implementation of the National Road Safety Action Plan. 57. Finally, HIV/AIDS awareness and prevention programs have been completed on the project roads. Vulnerable groups and contractor personnel participated in HIV/AIDS awareness and prevention programs before the start of civil works. The HIV/AID awareness program was implemented in parallel with the civil works. The program targeted not only the civil works workers, but also the indirect employees, along with the community in a corridor of 200 meters wide on either side of project roads. As a result of the program, awareness regarding HIV/AIDS improved from 50% to 100% among roadside residents. Awareness among females increased from 24% to 100% in some sections of National Road Although there were difficulties during project implementation, overall the project was implemented as conceived, and difficulties were generally resolved. Therefore, the project is rated successful, in accordance with the definitions and guidelines provided by the Independent Evaluation Department. B. Lessons 59. The following lessons emerged during project implementation. 60. Capacity building (planning). Road Data Collection and Management Unit staff are still unable to complete a road maintenance planning report with solid budget recommendations, which needs to be improved (para. 18). 61. Capacity building (implementation). The overall institutional setup for asset management, construction supervision and quality assurance and/or quality control, still needs improvement to be practically usable (paras. 9 and 18). In addition, the project completion review mission agreed with MPWT s recommendation for further training of RAMO staff as necessary in International Federation of Consulting Engineers-based contracts; the English language; technical reporting; procurement activities; road safety; and social and safeguards aspects, including gender. 62. Road safety audits. Similarly, the project completion review mission found the need for road safety audits to be undertaken as part of the pavement design process, with additional work to solve problems with ineffective lane markings (para. 15). 63. Road maintenance funding. Sustainable funding was not an objective of the project.

22 12 Also, sustainable road asset management was not given a sufficient focus by the government, with adequate annual budget allocation for road maintenance. These issues need to be explored in the proposed future project to achieve sustainability of the MPWT road maintenance regime (para. 74). 64. Project coordination and project readiness. Where multiple cofinanciers are involved in interventions, careful selection (and attention to transparency in collaboration) of co-financiers during project design is required, to enable close and effective coordination during implementation; this is especially the case for parallel cofinanciers (paras. 28 and 36). The project design should focus on project readiness, including climate resilient design of pavements (flooding has become more frequent than it was during project design). 65. Privatizing force accounts. The failure to privatize force accounts remains an issue to be resolved during the proposed future project. MPWT reported that the privatizing activities are now ongoing, although there has been no successful bidding process to date (para. 19). 66. Priority operating cost. POC was a confusing process for MPWT staff, which was implemented with delays, and it is recommended that a similar approach not be used for future intervention (para. 21). 67. HIV prevention program. Although HIV prevention program was implemented as designed, the budget allocated by IDA at project design was insufficient. Better project design with more detailed focus on an HIV prevention program is needed for future interventions (para. 24). 68. Independent procurement agent. This modality was not successful, and is not a good example to follow in future interventions, as shown by the issues that MPWT faced with the IDAfunded IPA (para. 36). 69. Project management. RAMO staff were not located in a single area, but scattered across their respective offices. This meant the project director could not supervise them in the course of day-to-day activities (paras ). 70. Robustness of cost estimates. Some cost categories were fully used and others insufficient (such as for the HIV prevention program), and more careful cost estimates are needed. 71. Link between cost estimate and financing arrangements. There should be a stronger linkage between cost estimates and financing arrangements to achieve the project scope as designed (para 25 and 26). However, this may have been possible only given sufficient ADB financing, ideally close to that provided by IDA. C. Recommendations 72. Project-specific and general recommendations that may affect projects or may be generally applicable to ADB practices are discussed below. 1. Project Related 73. Future monitoring. It has been a practice in road sector assistance in Cambodia that ADB focuses on sector issues common to all ongoing projects to establish a common