Project Labor Agreements and Bidding Outcomes: The Case of Community College Construction in California
This is a study of the effects of using Project Labor Agreements (PLAs) in the construction of community college projects in California. We divide the study into two parts.
The first part is a case study of seven projects built by the College of Marin, three with PLAs and four without PLAs. The upshot of this case is that the PLAs in comparison to the nonPLAs attracted a similar number of bidders, came in at a slightly lower price point compared to the engineer’s estimate, had about the same or fewer construction problems and trained more young, local workers due to the social justice component of the PLAs. We also find that local contractors were eager to bid on both PLA and nonPLA projects while bidders coming from afar preferred to bid on either the PLA or nonPLA projects but not both.
The second part is a statistical study of 88 community college PLAs and 175 community college nonPLAs representing $501 million in PLA work and $206 million in nonPLA work. controlling for when and where these projects were built, and how large each project was, we found that the PLA projects had slightly more bidders compared to nonPLA projects. We also found that PLA low-bids came in slightly lower compared to nonPLA projects. From these results, our conclusion is that PLAs do not reduce the number of bidders nor do they raise costs on California community college projects.
In June 2004, bond measure C passed in Marin County, California, providing $249.5 million to modernize the facilities of the local community college, the College of Marin. The modernization of the College included the construction of 7 new buildings, 3 of the projects were completed under a Project Labor Agreement (PLA) and 4 were not. All construction occurred between 2008 to 2015 providing a useful opportunity to compare bidding and construction on similar PLA and non-PLA projects
The PLA included common stipulations including sections outlining grievance procedure, management rights, and work rules. Like many PLAs, the College of Marin PLA included a social justice component encouraging the hiring of local workers, veterans, and disadvantaged workers, such as those with a criminal record. The PLA also stipulated that contractors were to hire students enrolled at the College to work on the project.
All seven new buildings were finished on time. A study of the first two PLA projects by Dannis, Woliver, and Kelley, Attorneys at Law concluded that “the two PSA [Project Stabilization Agreement—a synonym for a PLA] projects had fewer problems than some non-PSA projects.” The College’s satisfaction with the two PLA projects approved in 2008 led the College to assign a third project to be administered under the PLA in 2013.
Initially, each project was completed under budget. However, alterations following completion of two of the four nonPLA projects imposed cost overruns leading to final amounts that exceeded their original budgets. Nonetheless, it appears the cost overruns were related to architectural design errors rather than faulty construction.
Five College of Marin students were hired on PLA projects. Each student was trained by a different trade—sheet metal, carpenters, electricians, laborers, and plumbers. A recent study of apprenticeship training concluded that apprentices that complete their programs earn about $300,000 more over their work-lives compared to workers without apprenticeship training. One student, Julian Stone stated: “My whole life I’ve wanted to be a carpenter….The PLA project gave me the opportunity I needed to get my life together and going in the right direction”
In all cases, the lowest bid (excluding subsequent cost-overruns in two cases mentioned above) came in under the engineer’s estimate. For the four nonPLA projects, the sum of the lowest bids was $38 million or about $10 million per project. The sum of the engineer’s estimates for these four nonPLA projects was $50 million or about $12.25 million per project. The average number of bidders was 9.5 per project, and the average nonPLA project came in at 79% of the engineer’s estimate.
In the case of the 3 PLA projects, the sum of the lowest bids was $66 million or about $22 million per project. The sum of the engineer’s estimates for these three PLA projects was $88 million or about $29 million per project. The average number of bidders was 7.3 per project and the average PLA project came in at 75% of the engineer’s estimate.
On average, those contractors who bid only on nonPLA projects were located 51 miles from the College of Marin’s Kentfield Campus. Those who bid only on the College’s PLA projects were located 63 miles from Kentfield. However, those contractors who bid on both PLA and nonPLA projects at the College of Marin were located much closer to the Kentfield Campus—on average they were found about 25 miles from the College of Marin.
This “U” shaped relationship seems to reflect that those contractors interested only in bidding on nonPLAs or only on PLAs were willing to look far afield for such opportunities. Those interested specifically in College of Marin projects, regardless of whether they were PLAs or not, were located closer to the Kentfield Campus in the first place.
We supplement our case study of the College of Marin with a statistical analysis of 88 PLA and 175 nonPLA community college projects representing $501 million in PLA work and $206 million in nonPLA work. Built in 10 California community college districts over the period 2007 to 2016, using statistical analysis controlling for when and where these projects were built, and how large each project was, we found that the PLA projects had slightly more bidders compared to nonPLA projects, but that this difference was not statistically significant. Our findings rejected the hypothesis that PLAs reduced the number of bidders compared to nonPLA projects.
In a second statistical analysis of low bids on 105 projects where the engineer’s estimate was available, controlling for when and where the project was built, and how large the project was envisioned to be based on the engineer’s estimate, we found that PLA low-bids came in slightly lower compared to nonPLA projects, but that this difference was not statistically significant. Our analysis rejected the hypothesis that PLAs raised the cost of projects relative to the engineer’s estimate compared to nonPLA projects.