Publications / CSCE/CRC 2025 - Montreal, Canada
Project bundling is widely believed to reduce construction costs through economies of scale and administrative efficiencies. However, these cost savings are generally perceived based on qualitative analysis and engineers' experience rather than robust statistical analysis. This study analyzes over a decade of project data from the Tennessee Department of Transportation (TDOT) to assess the impact of project bundling on Cost per Lane Mile (CPLM). The dataset includes project information for 349 bundled contracts and 1,193 unbundled contracts for maintenance projects across the state. It provides a) descriptive analysis, b) t-test for state and region level data for all project types, and c) Mann-Whitney U test for state and region level data for each treatment type. The results present a complex picture of the relationship between bundling and CPLM. The descriptive statistics shows lower CPLM for bundled projects across the state and in various regions. A statewide one-tailed t-test shows that bundling significantly reduced CPLM. However, regional-level t-tests produced mixed results, with region 2 showing statistical significance, while other regions yielded weaker or insignificant effects. The Mann-Whitney U test revealed that only two out of seven treatment types exhibited significant cost reduction associated with project bundling. These findings shows that the impact of bundling is not consistent across various treatment types and geographic regions. It also highlights the need of more in-depth analysis to identify factors that significantly impact the cost savings while bundling.