Lies And Exaggerations.

The Meltzer Commission report runs roughshod over the facts.

In recent articles in this magazine,(1) Adam Lerrick, senior advisor to Chairman Allan Meltzer of the International Financial Institution Advisory Commission, has been sharply critical of the World Bank, arguing that our lending is misdirected, our performance deteriorating, and our financial costs exorbitant. These are very serious charges indeed.

The good news is that they are not true. The reality of what we do, how effectively, and at what cost is quite different. Lerrick's findings derive from selective use of facts -- many from an earlier era -- and faulty analysis. Nor do they support his core conclusions that we should terminate lending, switch to grants, and bypass local authorities in providing social and infrastructure services.

But far more serious than the breaches of logic would be the consequences of implementing Lerrick's recommendations. They would undermine our ability to help the nearly 3 billion people in developing countries who subsist on less than $2 a day. This at a time when our performance is improving, reflecting the major changes that the Bank has implemented over the past five years.

Does this mean that all is perfect within the World Bank? That we can't do better? Absolutely not. We are midway through a dynamic reform program designed to improve the Bank's internal performance and external impact -- and though there has been much progress, there is still an unfinished agenda.

In looking to the next phase of reform, it is timely to step back and consider important strategic issues such as our role in emerging market economies, our instruments, and the division of labor with the International Monetary Fund, the multilateral development banks (MDBs), and other partners. But it is best to do that reflection, and have that debate, with the facts in hand. Hence, this note to set the record straight.

DO WE LEND TO THE WRONG COUNTRIES?

Lerrick's arguments about Bank lending are misleading. Take the charge that the Bank inappropriately channels the majority of its flows to several large developing countries. Yes, the eleven countries Lerrick cites received about 70 percent of new International Bank for Reconstruction and Development (IBRD) commitments during the 1993-99 period. But these eleven countries are home to over half the world's people, and to almost 80 percent of the people -- and of the poor people -- who live in IBRD countries (Figure 1). It should hardly be surprising that much of IBRD lending during the 1990s was in such countries: that is where the bulk of the global development challenge lies.

[Figure 1 ILLUSTRATION OMITTED]

Lerrick also makes much of the declining share of IBRD lending to non-rated countries -- from 41 percent in 1993 to 1 percent in 1999 -- suggesting that at best Bank lending is not needed, and at worst it crowds out private providers. But what drove this shift was the fact that many previously non-rated borrowers began to enter the capital markets for the first time during this period, and therefore obtained "high-yield" ratings (Figure 2, pictured on the next page). Thus a better measure for assessing the changing balance of IBRD lending is the share going to countries with investment-grade ratings; this share remained broadly constant at about 30 percent throughout the 1990s.

[Figure 2 ILLUSTRATION OMITTED]

To be sure, countries' migration from "non-rated" to "high-yield" ratings points in a healthy direction that will ultimately have implications for the Bank's relationship with these borrowers. But it does not mean they no longer need support from the Bank during a transition period. Having a rating -- especially a high-yield rating -- does not ensure full access to the needed volume and terms of market financing. Rated countries do not have sustained access to credit markets, and the maturities they can obtain tend to be inappropriate for long-term development projects. Moreover, many countries gain initial market access in recognition of policy improvements, which can erode quickly unless they are bolstered by continuing development efforts. Immediately cutting back on the Bank's development assistance to such countries is not the way to ensure the long-term sustainability of reforms -- or to promote the adoption of reforms in the first place.

SHOULD WE NOT LEND AT ALL TO THE EMERGING MARKET ECONOMIES?

The World Bank Group aims to catalyze the private sector, not to compete with it. The strategic objective of much of our lending is to help developing countries and countries in transition to put in place the enabling environment in which their economies, and the private sector, can flourish. So while Lerrick sees displacement of private finance in the Bank's lending -- for example, to Brazil for public sector administrative reform and to the Philippines for financial sector reform -- we see enhancement of the scope for future private inflows. The irony is that in the end, once reform efforts are successful, it is Bank lending that is displaced by private capital, rather than vice versa as Lerrick argues. And this is exactly as it should be.

The typical pattern is for a country to climb the ladder of policy and institutional reform, which in turn makes it attractive to foreign creditors and investors, preparing the way for graduation from the Bank once it gains sustained access on reasonable terms. This pattern is apparent both at the country level -- for example, in Chile, Korea, Malaysia, and Poland -- and at the aggregate level. Indeed, as more countries did in fact climb the ladder through the 1990s, private financing to the developing world increased and IBRD lending became less important. But the journey is not always along a straight line, as Korea illustrates. During the 1960s, Korea's low per capita income and lack of creditworthiness made it eligible for concessional assistance from the International Development Association (IDA). It went on to borrow from IBRD in 1968 and to become an IDA donor during the 1970s. And it graduated from the Bank in 1994, but returned to the IBRD window in 1997-98. Indeed, the East Asia crisis demonstrated quite vividly the vulnerability of many middle-income countries to volatile private flows; it also heightened the need for vital structural and social reforms, for which a number of countries sought IBRD financing.

Lerrick concludes, however, that the Bank should stop lending to middle-income countries, and should focus its work with these clients exclusively on advisory services. This would be a mistake. No one believes that lending is a silver bullet, but it can be very useful in some circumstances. Important though knowledge and advisory services are, in many cases they are far more powerful in leveraging policy and institutional change and promoting consensus and implementation when associated with lending and underpinned by the discipline of timelines and of commitments on both sides. Equally, the operational experience of lending preparation and supervision keeps the Bank's advice practical and focused. The Bank Group would be less effective in pursuing the policy dialogue and promoting reform in emerging market economies without the IBRD lending window.

WHAT DO WE ACTUALLY LEND FOR?

Potential development impact remains the critical criterion for Bank engagement, whether with policy advice alone or with advice supported by lending. Emerging market economies -- even those with strong credit ratings -- are riddled with policy and institutional imperfections, some localized in particular geographic regions, some concentrated in particular sectors, some more pervasive. World Bank assistance helps them reduce these second-generation policy and institutional imperfections that impede private sector development and that keep public sector agencies from achieving peak performance. We work with countries on the legal and regulatory framework for privatization and private sector development, for the financial sector, and for infrastructure. We work at the state/provincial level in federal systems to promote fiscal sustainability and efficient social programs and safety nets. We help urban and local governments undertake municipal finance, zoning, and other subnational policy and institutional reforms geared to improving the provision of local public goods and services and the efficiency of economic activity. Policy-based loans support such efforts; and International Finance Corporation (IFC) and Multilateral Investment Guarantee Agency (MIGA) transactions with second-and thirdtier companies, small-and medium-scale enterprises, and others complement them. But Lerrick ignores all this.

He also ignores the fact that since the 1980s, Bank lending has come to focus less on the hardware and the bricks and mortar of development -- the tangibles that are so often more effectively provided by the private sector -- and more on the software and the policy and institutional reforms needed for a healthy and stable private sector, institution building, and social development. In telecommunications, for example, the Bank has not financed investment in new capacity for several years, relying instead on the private sector, often in collaboration with IFC and MIGA. However, the Bank actually finances more projects in the sector today -- focusing on helping governments build regulatory capacity and design mechanisms for private participation.

Similarly, the power sector accounted for 20 percent of IBRD lending in 1980. Today, reflecting the much increased private sector participation, it accounts for only 2 percent of IBRD lending. And even that 2 percent is different, focusing much more on public policies and institutions -- regulatory frameworks, tariff setting, environmental issues -- than on machinery and equipment. In India, for example, where there are decades-old fiscal problems rooted in the power sector, we are helping reform-minded states dismantle the bankrupt State Electricity Boards, privatize generation and distribution, and put in place new pricing and regulatory policies. At the other end of the spectrum, IBRD lending for education, health, nutrition, safety nets, and so on has expanded fourfold -- from 6 percent of lending in 1980 to 26 percent in 1999. (Figure 3 illustrates these trends.) Another rapidly growing area is the financial sector, for which the share of IBRD lending also has expanded fourfold over the period. At the same time, the substantive content of these operations has evolved from funding credit lines, which proved to be unsustainable absent the proper regulatory and prudential environment, to supporting financial sector reform and capital market development.

[Fig. 3 OMITTED]

LOW-INCOME COUNTRIES

For the low-income countries eligible for highly concessional IDA resources, Lerrick charges that we fail to heed our own research findings on the importance of the country policy and institutional environment for aid effectiveness. This is untrue. IDA allocations are based on an assessment of the quality of countries' policies and institutions. During the FY 1997-99 period, annual average IDA commitments to the good performers were $11 per capita, whereas poorly performing countries received less than $3 per capita -- about a fourfold difference between the two groups.

Meanwhile, our research consistently points to the importance of institution-building for aid effectiveness. Yet Lerrick's recommendation that we deliver project assistance to low-income countries exclusively by paying international service providers directly for basic social services and infrastructure would put this institution-building function at risk. To be sure, capacity building is among the most challenging tasks we face as a development community. But address it we must.

Finally, Lerrick recommends that we transform IDA credits into grants, and provide them also to a number of IBRD countries. This recommendation is simply not realistic financially. At a time of severely constrained foreign aid budgets, it is highly doubtful that donors would be able to provide and sustain the needed level of funding. It is difficult enough to secure the resources to fund current IDA programs. And requirements for donor funding would increase substantially under this proposal, as IDA would lose the reflows it currently receives from the repayments of credits and probably also the contributions it receives from IBRD net income, even as "demand" for IDA programs would increase with the proposed extension of eligibility to countries with higher per capita incomes.

PERFORMANCE TURNAROUND

A touchstone of the Bank's internal reform program has been its focus on product quality and development effectiveness. And the results are now showing up. Indeed, in the late 1990s, despite the growing complexity of the development agenda, there were major gains in quality and impact. Lerrick relies heavily on the analysis of the 1992 Wapenhans Task Force (of which I was a member), as if nothing had changed since then. Contrary to his assertions, however, recent data from the Bank's independent Operations Evaluation Department show substantial improvement in project outcomes, especially for adjustment lending. For operations evaluated in FY98-99, 81 percent of funds lent -- and 93 percent of funds lent for adjustment -- were in operations rated satisfactory or better by OED (Figure 4). Even by the very stringent test of likely sustainability, the respective scores were 66 percent and 84 percent; conversely, only 6 percent of overall disbursements, and 2 percent of adjustment disbursements, went for operations whose sustainability OED rated unlikely. Investment lending's lower scores in part reflect the fact that most of these operations date from the pre-reform era; this factor imparts a negative bias to scores weighted by number of projects rather than by dollars disbursed. The full impact on project outcomes and sustainability will only show up in the future, since average project life for investment operations is five to eight years. But even then, perfect outcome scores are unlikely, and even undesirable: given the inherent riskiness of the development business, they could mean the Bank was avoiding key challenges.

[Fig. 4 OMITTED]

COSTS TO TAXPAYERS

The multilateral development banks have proven to be very cost-effective intermediaries. A dollar of paid-in capital supports five dollars in loans, which are repaid and then loaned out again. The income from those loans can support more loans or can be used for grants to support other priorities of shareholders. In the case of IBRD, for example, some $11.4 billion in paid-in member contributions since 1946 have supported total lending of about $340 billion. With this financial structure, the budgetary costs to shareholders are essentially limited to their annual contributions to me concessional loan windows to help the poorest of the poor.

Lerrick, however, argues that "costless donor participation" in the MDBs is a myth, estimating the full costs at some $22 billion annually. But that estimate greatly overstates major cost items. It includes a hugely overestimated risk-premium charge on the "callable" capital that shareholders have promised to pay if IBRD needs it to meet its debt and guarantee obligations. With prudent financial policies and $28 billion of shareholder equity already on the balance sheet, the risk that IBRD will have to call on any of that additional capital to pay off its borrowings is minute.

Lerrick's analysis of costs also is inappropriate in that it imputes interest to cumulative donor contributions to IDA, which is akin to calculating interest on past charitable contributions. The right cost measure is the annual donor contribution, which amounts to roughly one-half of the calculated interest costs. Lerrick does have a point, conceptually, in counting interest forgone on shareholder equity as a cost. But he ignores the fact that IBRD earns significant financial returns, which, over the years, shareholders have chosen to reinvest in IBRD and to use for other development purposes, such as HIPC debt relief. In other words, for their investment of capital, shareholders have implicitly received financial compensation, which they have used to fund aid programs.

BOTTOM LINE

Where does all this leave us? With a World Bank that operates as an honest broker for emerging market economies in addressing their remaining policy and institutional bottlenecks en route to graduation from the Bank. With a World Bank that plays a broader developmental role in low-income countries, supported by IDA concessional funding and the HIPC facility. With a World Bank that uniquely combines a global perspective on sectoral issues with a local perspective in country programs all over the world, bringing to the table research findings, policy advice, and finance in the service of poverty reduction. With a World Bank that is midway through an important reform program designed to modernize its operations and to scale up their quality and impact.

In setting the agenda for the next phase of reform, we welcome the current debate about the appropriate division of labor between the Bank and the Fund on the one hand and the Bank and the other MDBs on the other. Getting this fight will be an important contribution to the design of the new development architecture. And we certainly are open to new ideas about how the development community in general, and the World Bank in particular, can best carry forward its mission. But Lerrick's recommendations are not the answer. If implemented, they would do more harm than good, and the world's poor people would be worse off. They are not the way to save the children....

(1) "The World Bank's Wrong Portfolio" (January/February 2000) and "Save the Children" (March/April 2000).

Joanne Salop is Director of the World Bank's Operations Policy & Strategy Group.