Savannah Norman is a senior at Duke University studying Public Policy, Spanish, and Ethics and Society with a focus in human rights. Her thesis combines all of her academic interests and crosses over the Public Policy and Spanish departments as a double-honors thesis. She is particularly interested in participatory international development in a Latin American context, having traveled and worked in Latin America for a year before coming to Duke. She has supplemented this hands-on experience with other research surrounding post-conflict societies in Central America. She speaks Spanish and Portuguese, which supplement her regional interests. Savannah is interested in working in project management for international development after she graduates.

Honors Thesis:

Assessing the Evaluation Methods of the Millennium Challenge Corporation’s Latin American Compact Projects

Faculty Advisors: Dean Judith Kelley and Professor Gustavo PT Furtado

Abstract: International development agencies have, for decades, worked to remedy global development challenges. One of these agencies is the Millennium Challenge Corporation (MCC). This paper explores the MCC’s Latin American programming via its evaluations. It specifically assesses the technical soundness of the MCC’s Honduras, Nicaragua, and El Salvador evaluations, as well what the findings and soundness of evaluations mean for future projects in the region. Through document analysis, this paper found that the average evaluation was technically sound, as it employed consistent, appropriate, and objective metrics. However, evaluation structure varied according to author and proprietor agency. External evaluations did not directly contradict the findings of MCC-sponsored evaluations. Evaluations were not inclusive to the intended recipients of MCC programming, beneficiary groups. Interviews with the authors of evaluations confirmed these findings. The implications of findings include the importance of culturally competent, inclusive, and multi-faceted development processes that assimilate learning from prior programming. Findings are applicable to development and evaluation processes in Latin America.