Exploring social capital debates at the World Bank.

I. INTRODUCTION

This article is located at the intersection of two literatures. The first relates to the study of development institutions and the practices that surround the production of discourses and their translation into material practices of intervention. The second addresses social capital in development studies, and specifically the role of the World Bank in promoting the concept of social capital. These two literatures are related and have made important contributions to critical development studies in recent years. Through an analysis of the genesis, evolution and effects of social capital debates at the World Bank, this article builds on and challenges each of these bodies of work, in an attempt to contribute to critical approaches in development studies. It argues that 'getting inside' development institutions is important for understanding how and why certain discourses emanate from them, for interpreting the significance of these discourses, and for understanding the indeterminate relationship between discourse and practice. More specifically, it also argues that only by understanding the embodiment of the social capital discussion in particular agents and its place in a longer history of arguments about social development inside the World Bank, can its relative significance --and insignificance--be appreciated. We suggest that external readings of these debates--because of their methodological choices--have simultaneously overstated the operational significance of these debates, understated the political significance of actual programmatic changes stemming from them, and mis-specified causal relationships in their production.

The article proceeds as follows. It opens by engaging these two literatures, and on that basis elaborates a framework for analysing the production of discourse within development institutions. Section III examines the micro-sociology and politics of knowledge pertaining to social development within the World Bank, with a specific focus on the emergence of social capital debates and their entry into the World Development Report 2000/01. Section IV then explores the relationship between the production of discourse and the practices surrounding its country-based operations. Section V elaborates implications of the argument made.

II. ENCOUNTERING ANTI-POLITICS MACHINES

The Production of Discourse

Critical development studies in the 1990s were marked by two books: James Ferguson's The Anti-Politics Machine [Ferguson, 1990, 1994] and Arturo Escobar's Encountering Development [Escobar, 1995]. Each provided devastating critiques of development, and more particularly of the ways in which not only official development agencies but also much development theory elided any discussion of the political economy of underdevelopment, or of the confrontational politics that can characterise efforts of the excluded to challenge existing models, policies, concepts, and constellations of power. Each also laid particular importance on the role of texts produced by these development organizations in this erasure of politics, and in presenting development as a technical intervention to be effected by expert institutions on objects of development (that is, people) whose ostensible problems had also been in large measure constructed by these institutions. Indeed, both studies depend considerably on discursive analyses of printed texts, though Ferguson [1990] also draws on his ethnographic work at the interface between these more formal discursive practices and the recipient society (Lesotho).

Yet while such analyses may reflect how the texts of development institutions can be and are read from the outside, they say little about the actors, dynamics, and political processes underlying the production of such texts. There are quite legitimate methodological and logistical reasons for this: getting inside the institutions producing these texts to conduct ethnographies of their production is immensely difficult. This does, however, leave hanging the question of how interpretations might differ were it possible to get inside these black boxes. It also leaves open the question of contingency--that is, of whether the organisations might, under particular conditions, have produced different texts, different understandings of development, and different forms of intervention.