Predictive factors for inter-agency partnership success

Abstract

Building partnerships among development cooperation agencies strengthens their collective capacity. However, 2 partnerships out of 10 fail to achieve their intended objectives. Ex-ante evaluation of future partnerships should help attain a higher rate of success. To do this, agencies need a better understanding of the factors behind successful or unsuccessful partnerships. This paper draws from the first stage of a research project aimed at systematically identifying and classifying such factors by importance in a semi-automated manner, to streamline partnership planning and evaluation tasks. The principal result we report here is a dataset containing 750 factors of influence, visualized along three axes: operations and country context, governance and management, and project quality, and grouped into ten clusters. Evaluators and planners can readily use the factors and complete method to focus multi-stakeholder discussions and information gathering, as well as to learn to use Natural Language Processing and Machine Learning approaches on project design and evaluation document corpora. Future research can overcome the limitations of the method by standardizing the factor set through ontological and taxonomic work, as well as incorporating automated context and behavioral analyses.