Box 11: Characteristics of Successful Education Projects
• The projects were consistent with the education strategies of DMCs.
• Borrowers made counterpart funds available as required and complied with loan covenants.
• Committed executing agencies were able to plan, manage, implement, and monitor the projects and benefited from capacity-building support. Institutional readiness is an important driver of project success.
• A series of projects and consistent ADB involvement over a long period contributed to successful outcomes, particularly when sector/subsector reforms were pursued.
• Participatory approaches were used for project design and implementation and to build alliances and shared ownership by engaging with a broad range of stakeholders.
• Basic and secondary education projects are generally not able to recover costs, so adequate budgetary support is essential for project sustainability; technical, vocational, and higher education projects recovered some costs and generated revenues.
• There was an emphasis on cross-cutting themes, especially poverty reduction and gender concerns.
• Project facilities were well utilized and maintained.
• Consultants and contractors performed well.
• ADB supervision missions resolved implementation issues.
In countries blending ADF and OCR funds, the success rate of OCR-funded projects (66% overall) has been higher than that of ADF-funded projects (52%). That held for projects approved in the 1970s, 1980s, and 1990s, with no evidence of a convergence in success rates. A significant difference in sector mix provides part of the explanation. Agriculture and the social sectors, which have had modest success rates, account for 79% of ADF projects in blend countries compared with 42% of OCR projects. In contrast, transport and energy, sectors with high success rates, account for 45% of OCR projects but only 18% of ADF projects.
The 2006 Annual Evaluation Review examined the characteristics of successful projects in five sectors: roads, power, water supply and sanitation, education, and irrigation and drainage. Naturally some of the characteristics noted are sector specific, but a number cut across several sectors. Some factors are dependent on the local situation, and some on the situation at ADB. Strong government ownership of a project is important. This translates into commitment and good performance of the executing agency, and provision of adequate counterpart funds for implementation and for operation and maintenance. The executing agency must be able to work together with the consultants, contractors, and ADB staff; must be familiar with ADB requirements; and must be able to learn from previous projects. For projects involving public works, the performance of supervision consultants and contractors is crucial. Continuity of ADB engagement in the country and sector, and of ADB staff involvement—ADB’s long-term commitment—is salient. Also on the ADB side, quality at entry is a recurrent theme, covering quality of project preparatory TA, participatory approaches to project design, a flexible design that allows a project to be adjusted during implementation, and the incorporation of lessons learned from previous projects. Regular supervision and project review missions from ADB were found to contribute to project success.
Program Lending. While project lending has constituted the majority of ADB’s assistance, program loans are also an important instrument and were used extensively in response to the 1997 financial and economic crisis. Program loans are disbursed relatively quickly to cover the adjustment costs of policy reforms. Program loans have been made to 31 countries, although most program lending is concentrated in a few. From 1978 to 2002, 133 program loans were approved for a combined ADF and OCR total of $18.1 billion. Of these, 92 have been evaluated: 48% were rated successful, 49% partly successful, and 3% unsuccessful. Most of the program loans approved after 2002 are still being implemented and have not been evaluated.
The trend in program ratings, by year of approval, is depicted in Figure 2. The success rates were unacceptably low for program loans approved through 1991, particularly regarding those approved from 1989 to 1991, when none of the 15 programs was rated successful.19 However, the success rate improved for programs approved after 1991, reflecting the positive effect of initiatives to improve quality, greater experience with the use of a policy-based lending modality, and revision of ADB’s program lending policies in 1996.
Figure 2: Trend in Program Ratings by Year of Approval (% Success Rate Based on 3-Year Moving Average)
BOP = balance of payments.
Source: OED.
Program loans are a more common lending modality in ADF countries than in OCR countries. Nearly 75% of program loans rated (68 of 92) were ADF funded. This is not surprising, as the need for policy and institutional reform is greatest in countries that are eligible to borrow from ADF. The analysis confirms that country factors, such as economic performance, development priorities, quality of governance, and strength of institutions, contribute to the success rates of programs. In all, 67% of the OCR-funded programs have been rated successful compared with 41% of ADF-funded programs. These results highlight the difficulty of formulating successful program loans in ADF countries.
Program lending has been most commonly used in the agriculture/natural resources and finance sectors and has rarely been used in the transport/communications and energy sectors. However, the sector distribution has changed significantly. More than 75% of the 78 program loans approved since 1996 have been concentrated in three sectors: law/economic management/public policy, finance, and multisector. The success ratings by sector and source of funding presented in Table 3 are difficult to compare because of the small numbers of OCR-funded program loans in most sectors. Most agriculture, industry, and social infrastructure program loans were in ADF-eligible countries, and there was no OCR-funded transport program loan. On average, ADF-funded agriculture programs performed poorly (worse than agriculture projects), with a low success rate of 30%; however, from 1995 onward, significant improvements were achieved. The finance sector is the one sector that had a comparable number of program loans funded by OCR and ADF. In this sector, the success rate for ADF-funded program loans (53%) was significantly lower than for those funded through OCR (90%). The disappointing performance of some program lending has been attributed to (i) the absence of government commitment to the principles and concepts underlying the reforms, (ii) the lack of a comprehensive sector analysis prior to the formulation of the policy reform package, (iii) a lack of clarity in program objectives and policy measures and their interrelationships, (iv) inadequate capacity of the executing agencies to carry out the required studies and monitoring, and (v) unsustainable policy reforms. Sometimes, the reform agenda was overoptimistic about what could be achieved during the time frame of the program. Delays in tranche releases were common.
Table 3: Program Loan Performance by Sector and Source of Financing (% of Program Rated Successful, Loans Approved from 1978 to 2002)
– = not applicable.
Source: OED.
The 2006 Annual Evaluation Review also reported on two themes that are consistent with the commitments made by ADB to ADF donors: (i) relating staff incentives to achieving development results; and (ii) factors