Managing “a little bit unsafe”: complexity, construction safety and situational self-organising

This explored shared the shared understandings held by UK construction workers towards safety by applying complexity theory. It involved interviews, document review, and site observations.

The authors first explains that construction projects have the traits of complex systems, with networks of interrelationships, multiple companies/subcontractors and long supply chains.

Further to the multi-nodality of construction projects, they frequently involve technologically complex interactions which are tightly coupled. The projects are “carried out within the spatially complex construction site environment that is in constant flux, resulting in non-linear interactions and unexpected consequences” (p2521).

However, typically, the methods employed to navigate complex sociotechnical environments in construction “focuses far more on technological project components than the social complexity of their interactions” and that construction has “continued reliance on work breakdown structures (WBS) and subsequent critical path analysis (CPA) for construction planning which are only able to acknowledge interrelationships in terms of sequence as associated with construction technology” (p.2521); rather than the influences of social interactions.

It’s argued that the Safety Management Systems (SMS) of many construction companies struggle to address systemic aspects of safety. While SMSs bring “together processes for management they are again limited by the simplistic conceptualisations of what they are trying to describe, unable to manage complexity in practice” (p2521).

It’s also argued that modern SMS approaches in construction may struggle with contradictions between work-as-imagined (as specified in SOPs, plans or according to managers) versus work-as-done (work as it’s actually performed in the field). They argue that while it’s tempting to simply call any local variations to work as done as drift or violations, these “are, nevertheless, essential “work-arounds” through which the overall functioning of the system is maintained” (p2522).

However, the local workarounds, while helping to satisfice local goals may also create unexpected and contradictory goals elsewhere; increasing risk and inefficiencies.

They also discuss literature around self-organisation and HRO/collective mindfulness which I’ve skipped.

In all, it’s noted that “Critical to safe working is the bringing together of the often contradictory demands of safety and productivity in the context of a complex system that also presents that space with design and planning shortcomings, and a failure to always appreciate how work is or can be done in practice” (p2522).

Results

First, there was no single “coherent, accepted understanding of what safety is on construction sites” (p2525). Safety was seen from perspectives of danger and when it becomes unsafety; seen as something that could be fluid and flexible or categorised into a safe/unsafe dichotomy; as a practice in and of itself; inherently linked to construction work; or something managed through enforcement or engagement.

Construction workers knew that their work was dangerous and that their understanding of safety didn’t always “fit with contemporary safety management practice” (p2526).

Safety, broadly, was accepted as part of construction and because it was accepted as inherent - it was belittled somewhat as a hindrance to getting work completed. Production pressures heavily influenced perceptions of safety, where safety was rarely considered outside of production pressures.

One safety discourse was a bureaucratic one, promoted by laws, SMS practices & which generally saw things as a binary ‘safe’ or ‘unsafe’. Workers’ experience was more fluid, influenced by changing factors at any moment during the day. For them, things could be “just a little bit unsafe”. Safety and unsafety could emerge at any point and time and was, therefore, socially constructed by workers through shades of grey rather than the black/white from legislation.

Because of the fluidity of risk & acceptance accidents are part of work, the alignment with zero harm as a target was problematic as it directly contradicted workers’ experience of reality.

They readily accepted that construction was a “complex mess” with competing demands that influence accidents. They saw the necessity to adapt: therefore, site rules were bendable; rules didn’t quite get broken, just “bent a bit” and was so necessary and banal that it was barely discussed.

People expected to get caught and punished for bending rules as they believed work couldn’t be accomplished as prescribed without it.

The authors discuss whether construction is ready to become adaptive. They argue that, given all its challenges, it’s perhaps already one of the most adaptive: where workers are aware of its complexity and the need to be creative to complete work successfully.

In contrast, the classic safety structure used within construction is said to be struggling with this complexity.

Link in comments.

Authors: Sherratt, F., & Ivory, C. (2019). Engineering, construction and architectural management.