Comparison of Cost and Time Performance of Design-Build and Design-Bid-Build Delivery Systems in Florida

TLDR

The selection of a construction delivery system is a critical decision for public highway agencies, yet no statistical study has compared costs and durations across transportation projects while controlling for agency differences. This study aims to identify whether design‑bid‑build or design‑build yields lower costs and shorter schedules for Florida highway and bridge projects, supported by a rigorous statistical analysis. Data from the Florida Department of Transportation were analyzed using Levene’s test, independent samples t‑tests, Welch’s t‑test, and the Mann‑Whitney U test to compare the two delivery methods. Results show design‑bid‑build projects achieved significantly lower costs across all tests, while design‑build projects performed no better in schedule and sometimes even slightly better, though differences were statistically insignificant. Other agencies can replicate the analysis to assess whether similar outcomes occur in their contexts.

Abstract

The choice of construction delivery system is one of the most important decisions that a public highway agency will make with regard to a highway, bridge, or transit construction project. Although several research projects have been conducted on the subject of comparing costs and project duration, there has been no study that statistically analyzes these factors specifically for transportation projects, much less one that eliminates the largest variable of all: different public agencies. The objective of the research reported in this paper was to determine which of two delivery methods, design-bid-build (DBB) and design-build (DB), delivers highway and bridge construction projects at the lowest cost and in the shortest time period, and to back up the findings with a rigorous statistical analysis of the data. Data were taken from the databases of the Florida Department of Transportation (FDOT) and arithmetically and statistically compared through Levene's test, the independent samples t-test, the Welch unequal variances t-test, and the Mann-Whitney U test. In fact, this paper's unique contribution is its rigorous statistical analysis of data gathered directly from FDOT, comparing the performance of the two delivery systems in Florida over a finite period of time. In this comparison, DBB projects performed significantly better in terms of cost according to all statistical and arithmetic tests. DBB projects did not compare as favorably in terms of duration. Some tests showed a statistically insignificant advantage to the use of DB over DBB, whereas other tests showed no difference between the two. Other public agencies can replicate the research and determine whether the same findings hold true in their situation.