Sensitivity of Cost Overrun Propensity and Severity to Highway Project Bundling

• Published in: Journal of management in engineering 2024-03, Vol.40 (2)
• Main Authors: Qiao, Yu, Nafakh, Abdullah, Alinizzi, Majed, Fricker, Jon D., Labi, Samuel
• Format: Article
• Language: English
• Subjects: Bundling; Construction contracts; Contingency; Cost control; Cost reduction; Highway construction
• ISSN: 0742-597X; 1943-5479
• DOI: 10.1061/JMENEA.MEENG-5425

This paper addresses the confluence of two issues associated with transportation construction: cost overruns, which represent a threat to cost savings, and project bundling, which represents an opportunity for cost savings. The paper hypothesizes that there exists a significant relationship between project bundling practice and the occurrence propensity and severity of cost overrun, and that the nature of this relationship varies across the different project types. Recognizing that current literature lacks adequate evidence regarding the nature of this specific relationship, this paper addresses the question using econometric modeling techniques. The study objective is to quantitatively assess project bundling impacts on the propensity and magnitude of cost overrun and how these impacts vary by project type. The methods used include binary logit and logistic regression specifications. The data were from a highway construction contract database involving 8,782 projects (consisting of 36 project types grouped into six work categories) and 5,180 contracts. The paper finds that, generally, project bundling does affect cost overruns, and that there exists variation of this relationship across the different types of projects. The outcomes of this study can help guide agencies on which project types could be bundled to reduce the occurrence propensity or magnitude of cost overrun as well as the bundling combinations for which cost savings may not be realized. As bundling practices become increasingly prevalent, public agencies interested in adjusting their financial estimates to accommodate possible cost overruns can replicate the study methodology provided in this paper. With the bundling and cost overrun relationship functions that could be developed using their local data, agencies can carry out, in a more reliable manner, the task of contract contingency sum estimation for bundled projects.