No multilateral institution suffers more from an identity crisis than the World Bank. Globalisation has brought private capital to the doorsteps of most developing countries, seriously undermining the bank's central function - development lending. Not surprisingly, when the United States Congress recently appointed the Meltzer Commission to pronounce on the future architecture of international financial institutions, the commission recommended that the bank exit the business of lending, except in Africa.

In The World Bank: Structure and Policies , Christopher Gilbert and David Vines come to the bank's defence. Armed with 11 essays written mainly by authors from within the bank and with an introduction by its recently departed chief economist, Joseph Stiglitz, they try to offer a markedly gentler picture of the institution than the one conveyed by the commission, dealing with the bank as an institution and with the effectiveness of its lending. But they fail to persuade.

In their main essay, Gilbert and Vines make much of turning the bank into a "Knowledge Bank" to give it a new identity. But their detailed discussion leaves little doubt that they do not seek major reform. They identify three key functions that they would like the bank to perform: lending, research and the development agency function, which is defined to include aid, loans with conditionality and global public goods. These being precisely the functions the bank performs currently, the authors' main task is to defend their pursuit by the institution.

The main argument offered favouring lending relies on a syllogism. The first premise is that the market fails to provide education, health and infrastructure facilities adequately in developing countries. The second is that due to weak powers of the state, governments fail to take corrective action, which the bank can take more effectively. Besides being paternalistic, this argument ignores two key facts. First, the bank has little chance of succeeding in countries where national governments are weak. Second, countries with weak governments can scarcely absorb the $18 billion currently lent. It is not an accident that 70 per cent of the bank's lending goes to 11 countries that are also large recipients of private capital and are thus perceived to have credible governments.

ÁñÁ«ÊÓƵ

A separate essay by Lyn Squire, a former World Bank research director, offers a defence of the bank's research function, on which it spends $25 million annually. Squire's main argument is that the creation and dissemination of development research is an international public good, meaning that no country will pay for its creation in adequate amounts for the fear that its benefits will spill over to others. Frankly, this argument has as much validity in practice as the infant-industry argument did when it was invoked by developing countries to protect and subsidise industrialisation in the 1960s and 1970s. One is hard pressed to explain why so many private consulting firms are able to sell their research to country governments if spill-over is a serious problem. And why does this problem not afflict developed countries?

Among all of the bank's functions, the one that is most easily defended is aid. As Gilbert and Vines correctly point out, this function is directly related to the poverty alleviation objective, on which there is little disagreement. But even here one must answer why it cannot be delegated to regional banks, at least in Asia and Latin America, where these banks function effectively as recommended by the Meltzer Commission. The authors do not address this issue.

ÁñÁ«ÊÓƵ

A key question concerns the effectiveness of lending. During the first three decades of its existence, the bank's lending was almost exclusively project-based. In the past two decades, conditionality-driven, policy-based loans have acquired greater importance. In their essay, Shantayanan Devarajan and Vinay Swaroop point out that when funds are fungible, a donor project may exhibit a very high return but it may be financing something with a very low return at the margin. To overcome this, they rightly recommend that aid be tied to an overall public expenditures programme that provides adequate resources to crucial sectors.

Fungibility of funds would seem to partially explain the conclusion in the essays by Craig Burnside and David Dollar, and Jonathan Isham and Daniel Kaufmann that aid is effective when the overall policy environment is good, but not otherwise. Burnside and Dollar find this in their analysis of the effect of aid on growth and Isham and Kaufmann of the rates of returns on projects. Thus, ultimately, overall policy environment would seem to be the key to development.

Sadly, current president James Wolfensohn has shifted the bank's focus away from what are known to be good policies based on 50 years' experience to what he calls the comprehensive development framework (CDF). Accordingly, projects and policy loans are to be evaluated via a CDF matrix that lists different aspects of development such as corruption, legal system, education, health, environment and even culture across columns and different participants including government, all donor agencies, civil society in all forms, and private sector at home and abroad across rows. As Roy Hopkins et al point out in their excellent essay on conditionality, priorities imposed by the CDF may conflict with priorities based on policy quality. This is not idle speculation: recently, changed priorities have led the bank to finance research on pre-Hispanic cultures in Honduras and museums in Mauritania and to spend its scarce resources on organising conferences of religious leaders.

While this volume fails to offer a persuasive defence of the bank's current identity or to suggest a substantially new one that will protect it from its critics, it contains a rich and up-to-date evaluation of its activities.

ÁñÁ«ÊÓƵ

Arvind Panagariya is professor of economics, University of Maryland, United States.