in agricultural extension and advisory systems in developing countries. The World Bank directly shapes the development path of many borrower countries through its research and through structural adjustment programs that restructure national economies or specific sectors (including agriculture). Yet Bank agricultural lending has decreased steadily over the past 60 years; currently it constitutes less than 10% of IBRD and IDA lending. The very mixed effects of these trends and shifts in financing on AKST and on innovation in the agricultural sector have been assessed in the 2008 World Development Report on Agriculture (World Bank, 2008). Internal as well as external analysts over the last 15 years have recommended that the trend be reversed.

     Like other development actors, the World Bank has evolved over the decades in response to different drivers, external pressures and internal experiences (Stone and Wright, 2007). According to one narrative, the Bank initially perceived its central role as the transfer of capital from rich countries to poor ones. The bulk of its portfolio lay in infrastructure projects developed by engineers. In the 1970s, Bank management concluded that infrastructure development alone was not sufficient to eliminate poverty and so Bank agricultural economists focused on "poverty alleviation." In the 1980s, macroeconomists, who played a leading role in designing investment projects at the Bank, viewed the debt crisis as evidence that sectoral development efforts could not succeed in the presence of major macroeconomic imbalances. Powerful interests in industrialized countries (where commercial banks feared that the loans they had made to developing country governments were at risk), pressured their government representatives in the Bank, and in the IMF to intervene. Accordingly, Bank management promoted structural adjustment programs as a condition of its lending. In the 1990s, the Asian economic crisis demonstrated that a narrow macroeconomic perspective was not appropriate for the pursuit of a sustainable development agenda, and the role of social sciences was gradually recognized. Changes in the hierarchy of professional disciplines within the Bank did not come about smoothly. Struggles eventually led to greater inclusion of the social sciences (Kardan, 1993); Ismail Serageldin, a Bank vice-president, spelled out why non-economic social scientists were not listened to earlier and delineated key intellectual challenges that remained to be faced (Cernea, 1994).

     Political economic, anthropological and ethnographic analyses have also assessed the role of the Bank (Wade, 1996, 1997, 2001, 2004; Ferguson, 1990; Harris, 2001; Mosse, 2004a; Broad, 2006; Bebbington et al., 2007). Simple causal linkages between external event, internal analysis, policy formation and subsequent implementation have been questioned (Mosse, 2004b). Evidence suggests that the Bank through its principal research unit has constructed, defended, maintained and promoted a neoliberal paradigm, despite changing contexts and emerging empirical evidence that challenge this paradigm (Broad, 2006). Organizational dynamics and international political economy have consistently shaped policy statements produced by the Bank over the period, while organizational culture-the everyday imperative to disburse loans and move projects through the pipeline, the internal incentive structure, hiring, staffing and subcontracting decisions and, importantly, power relationchapter

ships within the Bank and between it and other actors- have been more decisive determinants of the outcomes of Bank interventions than its policy statements (Liebenthal, 2002; Mosse, 2004a; Bebbington et al., 2007). The empirical evidence indicates a need for more political economy and social science-based analyses of the World Bank's institutional behavior, culture, internal and external power relations and dynamics, and outcomes in terms of equitable and sustainable development.

     The positive role of the Bank in the provision of financial resources to AKST includes loans to many governments in support of public agricultural research and extension institutions. Such support usually accompanied commitment to institutional reforms of these institutions. For instance, in Mali a Bank loan permitted the creation of research user committees at the level of regional research centers. These committees were designed to give a voice to farmers in the selection of research topics and in the evaluation of the usefulness to them of research results. The initiative gave some space for educated farmers with more resources to participate in research agenda-setting. In India, a large loan made in the late 1990s promoted significant reforms in the large Indian public sector agricultural research establishment, which had become quite bureaucratized. In Brazil, the volume of the loan made in the late 1990s was relatively modest; it was used by the then new leaders of EMBRAPA, the national research institute, to facilitate institutional reforms. The impact of Bank supported projects have been assessed and documented by the Bank's own Operation Evaluation Department (OED), a quasi-independent body which, while providing a degree of critical analysis, admittedly often reflects the dominant ideology in the Bank. Accordingly, the final and critical evaluation by OED of the T&V agricultural extension system, long promoted by the Bank particularly in Africa, was published only in 2003 (Anderson et al., 2006) i.e., long after the shortcomings of T&V had been emphasized by its critics; thus internal institutional learning and reform has been slow.

     In 1992, the Bank joined 178 governments in committing itself to Agenda 21, a global effort to articulate the link between environment and development issues. An internal World Bank review of progress towards environmentally sustainable development found that it had failed to integrate environmental sustainability into its core objectives or to forge effective cross-sectoral linkages between environmental and other development goals (Liebenthal, 2002). External assessments similarly found that the Bank had not lived up to the expectations of its Agenda 21 commitment (FOE/Halifax Initiative, 2002). A four-year Structural Adjustment Participatory Review Initiative, in which the World Bank participated, reported that the effects of the Bank's structural adjustment programs on the rural poor over the past 20 years had been largely negative (SAPRIN, 2002). External analyses likewise found that these programs tended to drive the evolution of AKST towards high external input models of production, while the pressure of debt repayment schedules in turn prevented governments from investing in poverty-oriented multi-sector sustainable development programs at home (Hammond and McGowan, 1992; Danaher, 1994; Korten, 1995; Oxfam America, 1995; Clapp, 1997; McGowan, 1997; Hellinger, 1999; SAPRIN, 2002).