UC Santa Cruz

International development nongovernmental organizations (INGOs) are a particularly good example of how transnational accountability commitments present challenges to producing credible accounts to diverse and at times competing stakeholder groups. This dissertation examines accountability at the intersections of power and knowledge production. I argue that the concept is necessarily a political and epistemic practice regulating human action through demands for its representation. Three assumptions support this treatment of accountability: the concept of accountability is applicable exclusively to human action; it assumes an agent has the freedom to choose and to self-regulate; and that an agent can produce an account that serves as credible evidence in verifying his actions are appropriate. The legibility and credibility criteria of accounts are examined through two distinct approaches to development. The "Modern Development Enterprise" is shaping upwards accountability expectations through funders' emphasis on aid effectiveness and the ability to demonstrate measurable "results". Funders are not only interested in verifying the effectiveness of past development interventions, they are also looking to secure development models which can assure their future investments will continue to maximize results for their money. "People-Centered Approaches" to development place an emphasis on the participation and empowerment of poor people, which has implications for knowledge-based accounts and claims of "downward accountability" by development agencies and INGOs. This focus on human agency, as both the means and the ends for development, drives specific methodological requirements for producing knowledge about poor-peoples' experiences and measuring people-centered interventions. Finally, the measurement practices of two INGOs are examined to demonstrate how they must negotiate multidirectional accountability to their stakeholders through methodological pluralism. Examples are drawn from the same food-security program in Bangladesh to illustrate how these specific and varied methodological requirements for measuring the work of development activities plays out in the INGOs' monitoring and evaluation practices. The preliminary evidence suggests that calls for multidirectional accountability may in fact ultimately take INGOs further from the sources of their credibility as development actors--their contribution as unique, value-based organizations and the ability to be responsive and accountable to those stakeholders who are the least powerful.