Selectivity

Global public-goods programs meet most criteria. While largely supply-driven, most Banksupported global public-goods programs (MLF, GEF, ProCarbFund and CEPF, CGIAR, TDR, UNAIDS, Stop TB, Roll Back Malaria, Global Forum for Health Research, and GAVI's Global Research Funding) largely meet the four Development Committee criteria for selectivity. Most global programs also largely meet the approval and eligibility criteria for Bank involvement. CGIAR does not meet the arm'slength criterion; the Bank did not involve developing-country stakeholders in CEPF's establishment or its global-level governance; the Bank did not do a thorough analysis of the expected level of Bank resources required for the health programs, or of how to implement and manage this new commitment. These are exceptions to the general rule, however.

The corporate advocacy programs meet the Development Committee selectivity criteria. This is largely because the criteria are broad and difficult to apply precisely. For example, the first criterion - "an international consensus that global action is required," which all programs claim as their raison d'être - provides no basis for selectivity because the concept of international consensus is amorphous and loosely applied. The case studies illustrate that the consensus is often driven by constituencies in donor countries and the staff of international agencies. At the same time, few of the networks demand links to country operations, one of the most important criteria, before approval, nor do they track them during implementation.

The Bank deploys its comparative advantages more at the global level than at the country level. Financial and reputational risks and budgetary and staffing implications are rarely sufficiently assessed. The international consensus on the existence of a problem is usually strong; consensus on what collective action is required is often weak. Many global programs are implicitly (sometimes explicitly) established to promote consensus, to "harmonize" donor approaches to specific problems, to delineate donor comparative advantages in addressing those problems, and to give the donors specialized knowledge to use on the problems. Capacity building in the recipient countries is secondary in such projects.

Evidence is lacking that the programs are exploiting economies of scale and scope in such activities as knowledge creation and dissemination, capacity building, technical assistance, and donor coordination. It is also not clear whether the knowledge they disseminate is sufficiently evidence-based, quality-tested, and contextual to add value to what the Bank's client countries themselves do, need, or want or what the Bank can achieve working through country-level partnerships. Performance indicators to assess changed donor or international agency behavior do not exist. Performance indicators, when they exist at all, are focused on the behavior of developing countries. IEG was able to identify only a few program-specific indicators of changed Bank and donor practices, procedures, and actions in response to the advocacy of global programs. In the case of corporate advocacy programs, the needs of the Bank's client countries should be the prime consideration for Bank involvement.

The voices of developing countries, or even those of the Bank's operational Regions, are inadequately represented in the international consensus. The case studies of corporate advocacy programs show that including developing-country voices at the concept stage enhances program ownership, makes the organizational design more effective, and increases program impacts. Based on the evidence IEG has provided so far, management has acknowledged the need to strengthen the role of developing countries and the Bank's operational Regions in global programs.

Value Added to the Bank's Development Objectives

Evidence on value added to the Bank's development objectives varies, but is increasing. Some programs lack clearly defined objectives, and others have many unstated objectives; this makes it hard to judge what value they have added. It is hard to assess many young programs that have not had time to demonstrate impacts. However, evaluations are increasing, in part prompted by the DGF, and are beginning to affect program design and implementation. When programs do not meet all three requirements for effective evaluation - clear, shared, and measurable objectives; appropriate methodology; and measurable evidence - their global impacts remain unclear.

Programs delivering global public goods often add value. Global public-goods programs (CGIAR, TDR, MLF, parts of GEF, and even some new global-health programs) rate well in their impacts on reducing poverty or on focusing on the policy, institutional, infrastructural, or technological constraints developing countries face in achieving sustainable economic growth. Adding value on the ground in client countries is typically a joint product of global and country-level activities. For example, CGIAR, like TDR, has demonstrated impressive poverty-reducing impacts in part because the Bank, donors and some governments made complementary investments at the country level. However, as country-level investments have shrunk, donors have tried to compensate by encouraging CGIAR to move downstream. They have offered restricted funding tied to research programs that demonstrate immediate impacts, to push CGIAR toward more national- and local-level applied and adaptive work. Management agrees that the activities of several CGIAR research centers now resemble those that regular Bank instruments would support through country-level investments.

Programs close to the Bank currently add more value. Not surprisingly, the programs for which the Bank is an implementing agency are more closely linked with Bank operations. This is in part because the Bank is better at absorbing and using information and findings produced internally or nearby. The Bank needs to devise ways to increase its links to programs more distant from it. Keeping the governance of global programs at arm's length from the Bank and maintaining clear accountability for program performance offer the greatest potential for bringing new information and fresh perspectives to Bank operations.

Global programs have revealed major gaps in investment. Evidence indicates that investments in health research have substantial poverty reducing impacts. The current global policy and aid environment has huge investment gaps at the global level in the provision of global health research, as well as gaps in complementary investments at the country level. Health research, like agricultural research, is a long-term activity unlikely to be addressed by the private sector on the scale needed.

Global programs have also revealed gaps in global public policy. Several global programs highlight the existence of global public-policy gaps - often involving industrial-country policies in trade, aid, finance and intellectual-property rights - that affect developing countries. Few programs regard it as within their mandate to address these policy gaps. If changing the international ground rules is the objective of the programs and if advocacy is the means to achieve it, then the programs should be assessed on their ability to deliver changed policies or a changed global environment from the perspective of the poor.

Governance, Management, and Financing

Governance is weak in several programs. While pure shareholder models of program governance are being replaced by stakeholder models, programs are still struggling to balance legitimacy and accountability for results with efficiency in achieving them. The permanent members of the programs' governing bodies, who tend typically to be the major international organizations and donors, have greater de facto responsibility, relative to the rotating members, to ensure that programs are successful. But such responsibility and accountability are rarely clearly articulated. Lack of effective governance and management must be addressed if the Bank's financial support is to continue.

Management arrangements can alter perceived and actual responsibilities. When the Bank or another international organization chairs programs that they house, this reduces the responsibility for shared governance. When programs are housed in the Bank or in another international organization, the program manager often reports both to the programs' governing body and to a line manager in the housing organization. This situation often places responsibility for both management and oversight in the same management chain, which in turn creates real or perceived conflicts of interest in monitoring performance.

Global programs have increased overall aid very little. At the aggregate level, the global programs reviewed have added little new money to official development assistance. Exceptions include funds from private sources for the Prototype Carbon Fund, from the Gates Foundation for health, and small amounts from pharmaceutical companies through new public-private partnerships for drug and vaccine development. Given the opportunity cost of ODA funds, the Bank's involvement in programs with important goals but little demonstrated value needs reconsideration. In some cases, too close an association with the Bank has hampered mobilization of other funds for these programs. It is time to move from "letting a thousand flowers bloom" to assessing which programs deserve continuing Bank support and which do not.

World Bank Performance

Bank performance in global programs is better at the global than at the country level. As a partner in global programs, the Bank has managed programs and mobilized resources better at the global level than at the country level. Other partners view the Bank's leadership role, its financial clout, its access to policymakers, its operational support, and its fiduciary oversight as a seal of approval, giving them confidence to invest in global programs, both in-house and externally managed. Even at the global level, though, the Bank's performance can be improved, particularly with respect to strategy, independent oversight, and global-country linkages.

The recent reforms are promising. The establishment of the Global Programs and Partnership Council, together with the GPP Group, is a positive development. In line with the Phase 1 report's recommendation, the GPP Council could help oversee the development of the Bank's global strategy, anticipate changes in the global environment, and help set priorities and funding strategies. It can move global programs from the current network perspective to a Bankwide perspective and establish Bankwide standards for global programming and performance. The Bank still needs to strengthen its appraisal of new programs and to make its selectivity, oversight, evaluation, and exit strategies more transparent and results-based. Finally, assessment and oversight of complex global partnerships requires expert knowledge and input, not only from the program managers who promote them but also from other partners, developing countries, and experts in the field.

Independent oversight is needed. The Bank needs to institute independent oversight of all its programs - in the case of in-house programs, by senior managers outside the line management of the vice presidency handling the program. Oversight of both externally managed and in-house programs needs to be guided by clear terms of reference, have the necessary budget, and have accountability for performance. Independent oversight is particularly important early on to ensure that programs get off to a good start. Bank management also needs to institute routine procedures of quality assurance, internal audits, risk assessment, and risk management.

Exit strategies of programs are not working well. The Bank's record in managing the separation of in-house programs from the Bank needs improvement. For example, the mechanical, hands-off, three-year rule for DGF Window 2 programs has not facilitated orderly financial exits. More attention needs to be paid to strengthening governance and sustainable financing of programs being spun off.

The Bank's strategy for global programs is poorly defined. The Bank has lacked, but clearly needs, a global strategy that is developed in conjunction with its key partners and draws on the capacity that exists in its central vice presidencies, network anchors, and Regions to do so. The strategy needs to address the coherence, or lack thereof, between global expectations (particularly in the donor community) and the needs of developing countries. At its center, the global strategy needs a clear focus on sustainable, poverty-reducing growth in the Bank's client countries; on global policy issues that prevent such growth; and on mobilizing incremental, unrestricted funding to address global issues that are high-priority for developing countries. Such a strategy will not simply emerge from improved selectivity or oversight of individual global programs; it must be worked out. Furthermore, strengthening oversight in the absence of an overall strategy risks micro-managing the global program portfolio.