The project was prepared through a sequence of four processing missions, without the benefit of dedicated project preparatory TA. In retrospect, given the weaknesses in the DMF, the four departures from the original scope of work, and the methodological shortcomings in the EIRR calculation, the project could have benefited from more careful preparation. Due to the unexpected long delay between the end of the last processing mission and the project’s approval following the political instability during 2000– 2002, there was ample opportunity for a more complete preparation. Small-scale project preparatory TA should have been considered during this period to better prepare the project.
The project processing missions could have consulted the key project stakeholders more effectively. For example, a decrease in the port service charge by 50% in 2009 and its elimination by 2010 was cited as a project benefit, but this would have required agreement of the shipping agents and shipping companies to implement. The same applies to the Lautoka reclamation component, which should have been undertaken only after commitments by shipping companies and agents had been obtained. More thorough consultation would have ensured that all parties were aware of their commitments and resulted in better project design.
Related to the above lesson on better consultation, the project should have done a better risk assessment. A key impact assumption overlooked was elimination of the port service charge by shipping agents and companies. Another assumption that shipping agents would not construct their own container yards near Lautoka Port should have been highlighted as a key output assumption for sustainable utilization of the reclamation area financed by the project.
Higher priority should have been given to ensuring that baseline data was collected and reported during project implementation. The data that should have been collected includes (i) benchmarking data, to provide MPAF management with objective reference points for impact evaluation corresponding with the needs of the port users and consumers; (ii) benefit monitoring data, to ensure that the project benefits actually accrue to the port users; and (iii) post-project performance evaluation data, to assess overall effectiveness of the project. The PCR identified this as a lesson, and this is reconfirmed at the evaluation stage. Subsequent to the PCR, Fiji Ports Corporation Limited and Ports Terminal Limited have introduced a more systematic performance management system, but port statistics provided to the IEM were still insufficient. The absence of these baseline data targets made independent evaluation of the project considerably more difficult.
(i) If the road asset management tools developed under the project were used as envisaged, this would improve the performance of the road sector in the long term. (ii) It is essential that investment in road safety is based on knowledge of road safety conditions, which need to be measured so remedial actions can be proposed and assessed. (iii) Traffic data and data on the performance of road assets is critical for performance measurement, future planning of the network, as well as for road asset management.
A more rigorous stakeholder analysis would have identified potential weaknesses and more emphasis could have been given to capacity development from the outset.
An ambitious target for the rehabilitation of 60% of significantly damaged schools by July 2018 was set under the program. The Ministry of Education did not have the capacity to implement construction projects on such a large scale. This necessitated the establishment of a construction implementation unit within the executing agency, the Ministry of Economy, that, along with Fiji’s transparent albeit complex construction procurement process delayed the start of the school rehabilitation component, with most of the works completed only towards the end of the project implementation period. The delays prompted the construction implementation unit to develop a data management system, which when completed and operationalized, could potentially be a best practice, particularly in smaller Pacific economies.
The government did not draw down the EAL proceeds until after the award of a substantial number of contracts for the rehabilitation of schools, although it had earlier withdrawn the DPL to meet the Help for Homes financing requirement. The lack of a definite schedule for EAL disbursements required ADB to maintain close coordination with the government to ensure that disbursements, when needed, were made in a timely manner.
EAL implementation is just like that of most other loans, except that processing is expedited to support the financing of early recovery efforts. Early processing of the EAL by the Asian Development Bank (ADB) and close alignment with the development policy loan (DPL) of the World Bank gave the government of Fiji the necessary confidence to proceed with supporting the poor and vulnerable families and homeowners and commence a relatively large-scale rehabilitation program for public buildings, including schools. The positive impact of the EAL is evident in the country’s sustained economic growth. Despite damages and losses equivalent to 27.6% of its gross domestic product, Fiji’s economy grew by 2.6% in 2016 and 5.2% in 2017, and is estimated to grow by 4.2% in 2018. Drawing on this experience, ADB should retain the EAL as a distinct product that may be availed by developing member countries prone to disasters.
The project steering committee, responsible for overseeing and monitoring all aspects of project implementation, was established at project commencement. It reviewed and approved subprojects but did not fulfill other requirements and had little involvement in implementation monitoring and supervision. Not all quarterly reports had been submitted, and the quality often inadequate for the purpose of full monitoring. Reporting on the indicators in the design and monitoring framework was extremely lacking. Serious deficiencies in the quality and regularity of project reports prevented timely and appropriate interventions to address potential and actual problems. Future projects should ensure that mechanisms are in place for necessary data gathering and monitoring and that loan covenants on project performance monitoring and evaluation are complied with.
ADB and the executing and implementing agencies must ensure that adequately skilled and experienced consultants are fielded to provide the inputs required for effective project design and implementation. The Report and Recommendation of the President, the project proposal document, explicitly noted that, based on experience in emergency assistance projects in the Pacific, adequate consultant support would be particularly important for disbursements. Nonetheless, the consultants engaged for the project design proved to be inexperienced and not suitable for the task. Given the weak capacity environment of Fiji and the high cost of providing consultancy services in the Pacific, the consultant selection process must be robust. Executing and implementing agencies need to check candidate references. The use of quality-and cost-based selection procedure, which may have discouraged some of the more experienced and better suited candidates to bid for consultancy assignments in this project, should be reviewed and analyzed across more projects.
Asian Development Bank (ADB) disaster response operations often involve executing and implementing agencies with limited capacity even before or due to the impacts of an emergency. The project inception report identified this as an issue facing Fiji’s Ministry of Works, Transport, and Public Utilities, the main implementing agency for the project. Inadequate follow−up was however made on the use of consultants, which was embedded in the project design to address capacity constraints. While shortened, the design and approval process for an emergency response project must rigorously assess the adequacy of executing and implementing agency capacity and install the necessary measures to address the constraints. Raising institutional awareness of ADB procedures should also be made a part of the project preparation process.
ADB did not take firm measures to ensure compliance to non-financial reporting covenants. Upon takeover from the WSD, the Water Authority of Fiji (WAF) was given time to familiarize with ADB policies and procedures while overseeing the remaining project implementation. Stringent reporting requirements should be enforced on WAF to ensure effective M&E and improve the implementation of the new investment facility.
Lessons identified in ADB’s 2013 special evaluation study, highlighting the importance of: (i) using performance-based information for decision making; (ii) promoting local ownership of operational reforms; and (iii) building strong stakeholder support to identify acceptable solutions, increase transparency and sustain implementation of reforms ─ remained valid under the project.
The civil works cost estimates for this project were prepared using preliminary designs. Costs went up during the detailed design stage as the project scope increased and prices escalated substantially after 2006. ADB’s introduction in 2015 of the project design advance financing facility is expected to facilitate more realistic cost estimates during the appraisal stage. Effective use of the facility, which provides resources for detailed engineering design and preparatory work, including feasibility studies, due diligence, safeguards and other pre-project implementation activities, supports the successful implementation of projects.
A reimbursement-only disbursement procedure, followed by the government during the first 3 years of the project, contributed significantly to implementation delays. ADB indicated this risk early enough. The risk surfaced in 2006, with an inadequate allocation for the Water and Sewerage Department (WSD) hindering the award of new project contracts. ADB strongly advised the government to switch to a direct payment mode in all review missions up to late 2007. When the interim government agreed, implementation accelerated.
The staffing and capacity of the project management unit for works, fully funded and composed of individuals selected by the executing agency (EA), was limited in the early stages of the project. The Asian Development Bank (ADB) considered this the main cause of the implementation delays. The loan covenants were insufficient to address the constraints and the human resource-related conditions for loan effectiveness did not help much. The EA’s capacity and resources should have been assessed at project appraisal to identify risks and allow for adequate mitigation measures to be taken promptly.