A real-time evaluation framework for maximizing the likelihood of success on large public infrastructure projects

Abstract

This thesis is motivated by the broad objective of identifying ways to maximize the likelihood of success on large public infrastructure projects. The specific objective of the thesis is to develop a conceptual framework for performing real-time evaluations during the implementation stage of these projects. This framework would provide a thoughtful and systematic approach for identifying threats and opportunities that relate to accomplishing long-term goals such that corrective or prescriptive actions can be taken on a real-time basis. As a starting point for the research an argument is presented to explain why large public infrastructure projects might fail to realize levels of success commensurate with expectations. It is often argued that project proponents, in their effort to gain support create unrealistic expectations, and that expectations should be lowered. However, lowering expectations may suggest an excessive pessimism about the potential of large-scale projects to be important elements in bringing about change. This thesis focuses on the possibility that large projects can be successful in terms of delivering expected results, and where they have failed it is often because of -- inadequate long-term/system-wide planning throughout the life of the project, and especially during the implementation stage. We propose that real-time evaluation related to accomplishing project goals can help to facilitate long-term/system-wide planning during the implementation stage of these projects. To provide a frame of reference on evaluation, this thesis presents a discussion on the theory and practice of evaluation. From this it is possible to extract some general principles for performing real-time evaluations. These principles are used to guide a discussion on some mechanisms that approximate real-time evaluation on public infrastructure projects. Out of this discussion, the factors that suggest how real-time evaluations could be performed effectively in the context of public infrastructure projects are identified and a conceptual real-time evaluation framework proposed. The central components of this framework consist of a trustee for the project goals and a supporting mechanism designed to perform real-time evaluation consistent with these project goals. This conceptual framework is then applied in the context of the Tren Urbano rail system currently under construction in San Juan, Puerto Rico. This effort reveals that while different elements for performing real-time evaluation exist in the Tren Urbano context, there are limitations to their effectiveness partly because they are structurally fragmented. A consolidated structure for conducting real-time evaluation consistent with the conceptual framework is presented. Finally, some recommendations for large public infrastructure projects in general are proposed.