Of loans and results: Elements for a chronicle of evaluation at the World Bank

Human Organization, Fall 1998 by Bare, Jean-Francois

Key words: evaluation, World Bank, anthropology, history, methodology

A Textbook Case

Evaluation procedures are currently inseparable from development policies and projects, and more broadly, from what are somewhat vaguely grouped under the heading of "public policies." It is, however, necessary to underscore that the very creation of institutions devoted to evaluation (taken in the most neutral sense of the word, that is, that of a diagnosis of the adequacy of results with regard to aims) is historical in nature. For instance, the World Bank, founded in 1944 in its first version as the "International Bank for Reconstruction and Development;' was not endowed with an independent - and therefore credible - evaluation department until 1975 (the Operations Evaluation Department or OED, which is directly answerable to the president and later to the board of directors). Political analysts have, moreover, pointed out this general evolution. An American political scientist specializing in social program evaluation problems noted:

In the United States toward the late 1960s it was practically impossible to find reports on the realization of programs. This notion of "realization" was virtually unknown and federal analysts assumed that once a program was underway it would function exactly the way the analyst had expected it to. (Levy 1984:30).

However, he adds: "Of course the reality proved quite different since each of the supervisors recast the program in the shape of his own views" (ibid.). In other words, the institutional means of public action cannot be dissociated from the results.

The practice of evaluation generally gives rise to difficult problems, not unlike those inherent to the social sciences and in particular to causal analysis in history. For instance, the process of determining whether the failure of a rural project was due to its intrinsic organization or to unrelated variables such as market fluctuations in the price of agricultural products during project implementation (Chelimsky 1995, Baneth 1996).

The problem really gets thorny, though, when evaluation raises questions about variables deriving from the aid institution itself, although it may seem to be one of the logical factors inherent to the process (Bare 1991). This is currently happening (November 1996) at the World Bank, despite its reputation as a monolithic and opaque institution. This occurs within the framework of an approach confronted with the necessity of integrating the historical dimension into an anthropological view in the broadest sense. The goal that is relentlessly repeated from the highest echelons is a shift from an "approval culture" or a "lending culture" to a "results culture."

One may, of course, be tempted to deride a "revolution" that in some respects sounds like reinventing the wheel. But it would be forgetting, as certain self-appointed critics of multilateral institutions often do, that the World Bank is a bank; for a bank, a loan is a result. When the bank is paid back, that is a good result, but the term result connotes that idea. This brings us to the core of the debate on evaluation, which calls up many necessary comments on terminology or, if preferred, problems of interpretation.

By choice, the World Bank only provides low-interest or interest-free loans via its subsidiary, IDA (International Development Association), for the poorest countries. The founding philosophy of the Bank is totally against a handout system of aid, epitomized by free loans. To criticize the World Bank for being a bank is, as Nietzsche puts it, like criticizing a lion for being a lion. It likes to present itself as a "cooperative bank" in that access to loans is proportional to capital contributions. It finds most of its resources on the international capital market, and is bound to lend at rates compatible with what it borrows.

This does not make it structurally distinct from commercial banks; the difference, on the other hand, lies in a far more cautious ratio between commitments and its own capital (Bretaudeau 1987). If the World Bank goes bankrupt, only central banks can refinance it. Its action is thus inseparable from multiple legal procedures, having to do with bank guarantees - pledges, co-financing, conditional loans - and linking it to "client countries." That these variables are often neglected in ritual critiques wielding, as Georg Lichtenberg (1742-1779) says, a knife without a blade, for which the handle is missing, must therefore be borne in mind.

This period of willful transition to a "results culture," marked by the creation of a Social Development Task Force by current president J. Wolfensohn, thus provides for interested parties an extraordinary observatory for the logical and methodological questions that arise with evaluation procedures.

Instead of tackling these questions head-on and from an abstract perspective, we prefer first to consider some of the events that have marked a "ponderous" context - in de Jouvenel's sense of ponderous decisions -- through the accounts of "chroniclers:" oral accounts by the institution's collaborators and written accounts by journalists specialized in the field.