Is it possible to influence safety in the building sector?: A literature review extending from 1980 until the present

SummaryThe available literature on construction safety is not very optimistic about the chances of evidence-based safety in the construction industry exerting a positive influence. Many articles indicate that the structures and processes that are designed to ensure safety in the industry are poor. Safety management systems do not work, or are limited, the business processes executed are fragmentary, it is not clear who is responsible for safety and parties lower in the construction hierarchy tend to be saddled with the consequences. Safety detracts from the primary production process and is seen as a bureaucratic burden. But there are some positive developments as well. Lists of prevalent accident scenarios and central events are available and information is published on barrier failures. What is missing is a reliable exposure gauge of the relative importance of scenarios and the identification of pivotal events. The more clearly the cause-effect chains of accident processes can be recorded, the more specific the measures, solutions and interventions can be when it comes to avoiding or reducing the effects of accident scenarios. Audit methods have also been developed, such as the Safety Index, which can be used to not only negatively but also positively assess safety. Finally an approach that can best be described as ‘frappez toujours’ seems to yield noticeable results. In such cases it does not really matter what safety intitatives are taken. Simply highlighting the issue is a factor that can, in itself, have an effect.

► Literature is not optimistic on evidence based safety in the construction sector.
► The hierarchy of accident causes, barrier failures is known for decades.
► Safety management in construction is laborious, with the exception of some examples.
► Project design phases have potentially opportunities for safety, but are hardly used.
► Methods are developed to assess accident scenarios prospectively.