system is being created, as it was under the DEMP, it is critical that a commitment for sustained investment is made. Neither the government nor ADB had any interest in a follow-on project. Even a 1-year extension to allow enrolments to continue their upward trend at the end of the project was viewed unfavorably. Innovation and reform require more long-term commitment to be successful.
The experience of the DEMP provides a good lesson on the importance of developing and operationalizing sound sustainability strategies. The project paved the way for the establishment of an innovative national online learning system, but the lack of a clear long-term plan was one reason why the concept eventually failed. If the best long-term strategy for sustainability is difficult to specify at the outset, specific support and additional time should be allowed in the project design to prepare and operationalize an agreed sustainability plan.
In most education projects, and the DEMP was no exception, the focus of project implementation is on the development of facilities, teacher training, and programs that are only completed near the end of the project. The final achievement measures, however, are generally framed in terms of the number of students intended to benefit from the program. Again, the DEMP was no exception. It takes time, however, for institutions to train staff and develop online courses—especially technical courses, to launch them, develop good online or face-to-face tutorial support, and then to attract reasonable enrolments. It generally takes at least one cycle of a program to convince students that new programs are worthwhile. In the case of the DEMP, additional time should have been built into the design, or it should have been a multiphase project. This would have allowed not only development of the network and associated courseware, but also time to allow the network to establish itself, experiment with different approaches, and develop into a mature system with good potential for sustainability.
Effective champions are critical for the success of any new initiative. This is even more true for innovative projects, where strong commitment to the project on the part of both the government and ADB are critical to success. The failure to locate the DEMP within a government agency with strong ownership of the project was one of the factors that led ultimately to its failure. The sustainability of initiatives beyond the lifetime of the project rests heavily on the commitment and support provided by the relevant ministry, in the form of formal institutionalization of systems and processes, further training, and regular follow up by relevant staff. Similarly, the failure of ADB to adequately invest time and resources in project implementation, and the failure to remain involved in ensuring network sustainability in the years after the project ended, when its fate remained in limbo, also contributed to the project’s limited success.
The DEMP failed to expand tertiary education opportunities and increase the employability of graduates to the extent intended. The question of how to achieve both of these objectives remains. A central issue in this regard is how to improve external degrees. One of the lessons of the DEMP is that Sri Lankan students may not yet be ready for fully online courses, at least for the longer degree courses. To raise the quality and range of offerings to external students, a blended learning approach would be most appropriate, combining online learning with a limited amount of face-to-face tutoring. This would allow the introduction of science and engineering as external degrees, currently not offered. Some universities have established their own open and distance learning centers, and these could be considered for future ADB assistance, along with support for further capacity expansion at the OUSL.
Apart from its innovative nature, the DEMP was also very broad, with the project inputs spread very thinly. The project included the construction of a national broadband network; mobilization of non-traditional partners including private institutions; capacity building and institution strengthening within the OUSL; development of online courses, which is notoriously difficult and slow; and implementation of these courses with the expectation of rapid, high enrolments. The DEMP may have had more chance of success if it had focused only on the OUSL, or only on NODES, with the strengthening of the OUSL as a separate project.
For innovative projects, significant resources are needed up front. This can include resources for assessing the feasibility of, and demand for, the project objectives and approaches for achieving them, as well as to lay the groundwork for ensuring their acceptability among different interest groups. Resistance to the innovations proposed cannot be overcome at the same time as project implementation is proceeding. The same applies to demand for the project outputs and outcomes. It is critical that a reasonable amount of consensus and commitment toward the project objectives is established before implementation begins.
Considerable time, effort, and emphasis must be placed on the development and monitoring of the risk matrix for innovative projects. This applies particularly to risk identification and potential risk mitigation strategies. A range of alternative strategies should be developed in detail for every identified risk. The risk matrix and mitigation strategies for innovative projects should be discussed extensively by ADB and the government to ensure a common understanding and commitment. The risk matrix should be a major focus of ADB-government discussions in project review missions during implementation, and should be updated at least annually through a formal process such as a memorandum of understanding. DEMP implementation would have benefited significantly from a sharper focus on emerging risks, development of appropriate strategies to address these risks as they emerged, and a strong joint commitment to implementation of these strategies.
By definition, the knowledge, skills, and experience needed to guide and embed innovative change are likely to be scarce in both the borrowing countries and in ADB. It is evident that additional time and resources are required to support implementation, which did not happen in the case of the DEMP. This should include longer, more intensive, and more frequent review missions, and supplementation of missing technical skill areas by bringing in additional staff from across ADB or engaging external support for review missions in particular. Technical assistance to strengthen the expertise available under the loan and ensure ongoing support to implementation may also be necessary. If the sector department is not committed to providing additional resources and support to innovative projects, then these projects should not proceed.
Although it seems self-evident, this fundamental lesson is often overlooked and needs to be reiterated. Community involvement in planning and responsibility for O&M is a practical and cost-effective way of managing such small-scale systems. The government-enabled community ownership and capacity building approach employed by the project will be useful in other efforts to scale up agricultural development, for example in improving rural health, water supply and sanitation, and education. The project demonstrated that community mobilization combined with sustained agricultural extension services can deliver both higher rice yields and expanded irrigation areas through communities’ substantial and long-term commitment.
Sustainable O&M that guarantees water provision is an incentive for farmers to remain members of WUAs. Farmers tend to withdraw from WUAs if their fields do not receive adequate water flows. Without water delivery to their fields, there was no incentive for farmers to work on land conversion and therefore some refused to invest their resources in converting undeveloped land to paddy fields, which would have increased the irrigated area under the project. Farmers’ willingness to pay for irrigation services is affected if they cannot get adequate water flows to their fields, and if irrigation facilities damaged by extreme weather events are not expeditiously repaired by the government. The government’s response capacity for post-completion maintenance needs to improve by allocating adequate funds systematically.
Frequent damage to irrigation facilities caused by flooding and landslides is a big threat to the sustainability of infrastructure improvements. Considering the poor watershed management in the areas surrounding subprojects and incompatible land uses (e.g., hilly land corn farming above the irrigated rice fields), the construction of irrigation infrastructure needs to improve. Preventive and protective measures need to be taken to guard against recurring intense rainfalls and extreme weather events. Such measures could take account of environmental risks and ecological accidents and apply watershed management approaches. This would build long-term climate resilience, supporting the government’s Agriculture Development Strategy to 2025 and Vision 2030.
Accompanying TA would help clients meet key program targets, such as implementation of the IGF and medium-term budgeting. Policy-based loan design should ensure that TA findings are translated into actions. With the MFEM’s preference for policy-based lending, loan design should ensure strong links between TA and the program being supported. Given (i) the limited implementation capacity of the government, (ii) the approval of a policy-based loan supporting Cook Islands’ Disaster Resilience Program, and (iii) the MFEM’s preference for policy-based lending, further capacity support is needed.
As noted above, donors provide a significant amount of resources, including for capacity building. Various modalities of donor participation and coordination could be explored in the design of policy-based loans, such as inclusion of donor commitment and funding of specific programs and projects (especially in the social sectors, where donors are active) and TA (notably in PFM).
Many of the milestones for institutional reform are process related, e.g., cabinet endorsement of the IGF. The outcomes from the institutional reforms are not well-defined, e.g., higher economic and social rates of return. There is a need for closer monitoring and evaluation to determine whether the reforms achieved their objectives. The theory of change and intended results should be well articulated at the outset. Policy actions should be linked to development outcomes that are clearly defined, with measurable indicators and baseline data for future evaluation.
The 34 policy actions for a crisis response operation stretched out the government’s capacity in absorbing a substantial reform agenda. Constraints to capacity were reflected in the delays in complying with the policy actions, in particular with the policy action on financial ratios, which resulted in waiver of this action. Institutional capacity to manage the reform process and to implement agreed-upon reforms should be factored into the program design and continue to be reassessed during program implementation. The need to continuously strengthen the technical and managerial capacity of institutions, in terms of translating policy actions into workable, concrete measures and assessing reform options, should be carefully considered in designing reform programs.
A steady, step-by-step pace of reforms may be more suitable in the Tongan context, given prevailing capacity constraints. The process of implementing reforms takes time, especially for reforms that require specialized skills, such as those in PFM and those that are structural in nature; these types of reforms take time to implement and cannot be completed by a single quick-disbursing stand-alone program. The program’s sequence of policy actions in the reform areas of PFM, PSEs, and business environment could have prioritized the binding constraints, such as the legal and regulatory issues, before getting into the operational aspects of the reform issues. However, this was not the case for the policy actions on social protection, which targeted vulnerable groups without an established system for identification, delivery, and monitoring. Also, ADB’s coverage of areas for policy reforms under the program could have focused on key needs and constraints. A gradual approach that considers more realistic timelines and proper sequencing could allow more flexibility in prioritizing and refining key policy actions in successive stages of the reform process. This could help strengthen the enabling environment for reforms, foster learning, and build capacity for policymaking and implementation.
The operation and maintenance of agricultural infrastructure built through projects needs to be sustainable. Routine O&M costs have been reduced by the project’s concrete lining of the canals. In addition, the choice of this technology reduces water losses from canals and thus enhances irrigation efficiency and the reliability of water delivery. Where flooding is a predictable annual event, as in the Quang Binh subproject, concreting the top of the dykes and covering it with gravel may prevent damage. Concrete lining should be promoted, given that older parts of the existing irrigation infrastructure would have deteriorated without the lining done under the project. Piloting other technologies, such as buried pipe irrigation and closed concrete canals, should be explored as such systems are cost-effective in terms of O&M and more resilient to extreme weather events, leading to lower maintenance cost in the long term. The extraordinary damage caused by typhoons and consequent flooding requires further research with different technologies, especially given the emerging risks of climate change.
The government should take effective financing measures for O&M and repairs to support IMCs and the cooperatives, especially through the provincial people’s committees. Mechanisms to ensure that IMCs and WUGs receive funds at the right time for repairs are needed. At a minimum, WUGs and cooperatives must work closely with the IMCs to ensure timely and effective O&M. The newly established Disaster Management Authority within MARD should play a role in this.
Close monitoring of the condition of the infrastructure is needed and cost recovery measures should be instituted for O&M of: (i) tertiary systems managed by cooperatives, and (ii) primary and secondary systems managed by IMCs and IMEs. The findings of the monitoring exercises should be regularly shared with the central and provincial governments to help MARD and the DARDs prepare budgets and mobilize resources.
Although their autonomy was short-lived, WUGs functioned well using hydraulic boundaries. Project WUGs had their own bank accounts and stamp and the right to sign contracts. They received training in irrigation planning and management and financial management. Empowered water users became fully committed to the infrastructure that was crucial to their agriculture livelihood. It was acknowledged by the IMCs, the cooperatives, and even by DARDs that some WUGs constructed tertiary irrigation canals more quickly and/or at lower cost than PPMUs. Since cooperatives have absorbed the role of WUGs, they should capitalize on the WUGs’ enhanced capacity to boost PIM. Maximizing the use of the irrigation management knowledge embedded in the WUGs and cooperatives would improve tertiary system management. If cooperatives work closely with the IMCs and IMEs, this would also help define O&M needs and systems.
The project introduced a project performance monitoring system that made use of GIS and was supposed to be applied by the PPMUs. However, since benefits started to materialize only upon completion of the project, the government was more interested in construction monitoring. After the project, no funds were available for the continued use of the project performance monitoring system and it has been abandoned. During the project, the ADB resident mission and MARD’s central project office introduced a detailed project implementation planning system, which made a substantial contribution to limiting the delay in project completion to 7 months, avoiding the need for a formal extension of the project. This practical tool should be used in other projects whenever feasible.