Contractor Safety Prequalification: Current Practices and Prospective Models
Authors: Peter Philips and Norman Waitzman
Date: April 2008
Summary
The authors were tasked to investigate existing prequalification processes through literature review, to perform an analysis of the effectiveness of current safety and health prequalification processes, to identify critical attributes of prequalification processes that are most likely to enhance contractor safety performance, to explore the attributes of a model for contractor safety prequalification and to explore the methods for testing the effectiveness of safety prequalification of contractor.
Conclusion:
The report from the University of Utah provides extremely valuable insights into the safety prequalification process and critical issues to be considered in developing a template for safety prequalification. The University of Utah study results indicate that some measures of past performance – historical lost workday injury rates and EMRs – are useful predictors of current contractor rates for lost workday cases. In addition, some measures of current capability had predictive value. The existence of a company safety director and on-site safety representatives were positive predictors of favorable performance. Behavior-based safety programs, which are based on peer-to-peer observation and feedback, also seemed to predict positive results. These potential predictors represent a balance between both traditional outcome or trailing metrics and leading indicators; data that should be available for many firms for incorporation into a contractor safety prequalification system.
Findings:
According to the lead researchers Philips and Waitzman, host employers are able to manage the safety and health of their own workforce while in a sense they must buy the safety of their contract workers. They do this through the prequalification process and by selecting contractors that are likely to perform safely on the job.
Their extensive literature review shows that safety pre-qualification can be used as part of a strategy to reduce the liability concerns that have hindered promotion of safety in the host-contractor relationship. But, to be effective, safety prequalification needs to be integrated as part of a larger strategy that encourages the extension of an effective safety culture from the host firm to contract workers.
The authors noted that expense is a consideration since in prequalifying contractors for safety host employers must not only pay to assess the contractors they hire; they must also pay to assess contractors they choose not to hire.
Philips and Waitzman found that two categories of information can be used are necessary to assess the potential safety performance of a contractor on the job. These are as follows:
1. Past safety experience: In an analysis of contractors' past performance as a predictor of future safety outcomes, the authors found that past lost workday injury rates strongly and tightly predicted current lost workday injury rates. Past worker's compensation experience modification rates (EMRs) were also good predictors of current lost workday injury rates, although not as good a predictor as past lost workday injury rates. A multiple regression model including both factors tested on the history of 373 contractors showed that both predictors had independent value in predicting current contractor safety outcomes and can be used together to help screen contractors.
However, when it comes to case severity the opposite was true. The seriousness of past injuries seems to influence the amount of attention paid to safety. A multiple regression analysis of 114 contractors showed that when the average number of days lost per lost workday injury case is doubled in the past, the contractor's current lost workday injury rate will be 7% lower (assuming all other things are equal). It appears that contractors learn from past mistakes, especially serious mistakes. The authors also tested the effect of the number of past fatalities on current lost workday injury rates, finding that each additional past fatality lowered the contractor's current lost workday injury rate by 7%. The problem with using fatality data is that fatalities are rare events. A larger sample is needed to give a more precise idea of how past fatalities reform the ways of contractors.
2. Current capacity: The authors also examined several potential measures of current safety capacity (based on a multiple regression analysis of 526 contractors). The presence of a behavior-based safety program, all other things being equal, lowered the contractor's predicted lost workday injury rate by about 64%. Having a company safety director, compared to those who do not, lower the predicted injury rate by 72%. Having a site safety representative, compared to those who do not, lower the predicted lost workday injury rate by 44%. Return to work policies, light duty policies and accompaniment to medical treatment policies did not affect lost workday injury rates in the sample studied. Doctor accompaniment policies seemed to reduce minor injury reports. Contractors with these policies have a 43% lower restricted duty injury rate.
Overall, the authors believe that past safety performance seemed to be a much stronger predictor of current safety performance than assessments of current capacity.
The authors identified factors that should be considered in developing a safety prequalification model. These include the need for pre-bid risk assessment which allows the host employer to determine if safety prequalification is needed to address the risk. Pre-bid risk assessment also allows the host to clearly articulate the risk in the bidding process so that the contractor can properly factor safety into their bid. Another item is the size of the bidding pool. For example restricting bid awards to contractors with appropriate safety qualifications may have an economic impact on the bidding process because it restricts the pool of available bidders. Ideally every bidding pool will have the optimum number of bidders to meet the host employer's need. The authors found that achieving realistic costs for safety prequalification is critical in developing a prequalification program. Interviewed host employers were almost unanimous in stating that contractor safety culture was an important indicator of the future safety performance of the contractor. However, assessing contractor safety culture in a safety prequalification system can be difficult and expensive. It is clear to the authors that extensive safety prequalification is needed for higher risk projects. The report discusses different methods for evaluating the results of the prequalification effort and the impact of having contract worker prequalified and how this might be accomplished.
Deliverable:
Contractor Safety Prequalification: Current Practices and Prospective Models
