A permit is signed off. A supervisor has reservations but does not raise them. A team knows the job is running late, the client is waiting, and the next shift is already under pressure. No one intends to compromise safety. Yet this is precisely where the question, can culture affect safety, becomes operational rather than theoretical.
In high-risk work, culture influences what people notice, what they challenge, what they report and what they decide when procedure, time and production pressure collide. It does not replace technical controls, competent people or sound engineering. It determines how reliably those controls are used in real conditions.
Can culture affect safety? Yes, through behaviour under pressure
Safety culture is often discussed as if it were a set of values on a wall. That is not where its effect is felt. Its real test is behavioural: whether a project manager stops work when a control is unclear; whether an engineer admits uncertainty before a decision becomes irreversible; whether a contractor believes the same standards apply to them as to the client team.
Culture affects safety because it shapes the social cost of doing the right thing. In a healthy environment, pausing a task, escalating a concern or asking a basic question is treated as evidence of professional judgement. In a weak environment, the same actions can be read as inconvenience, lack of confidence or failure to cope.
That distinction matters. Most serious events are not caused by people setting out to be reckless. They develop through small adaptations: a check shortened because it has been done many times before, an unresolved interface issue accepted for one more shift, a near miss left unreported because no harm occurred. Culture determines whether those adaptations are surfaced and corrected, or normalised.
Leadership sets the practical standard
Senior leaders can state that safety comes first. Their teams will judge the claim by what happens when safety has a cost.
If operational reviews focus relentlessly on schedule recovery while assurance concerns receive limited attention, people receive a clear signal. If leaders thank someone for escalating a problem but later criticise the delay it caused, they create an equally clear signal. The written policy may be unchanged. The working culture is not.
Leadership credibility comes from consistency between stated expectations and the decisions leaders make under pressure. This includes the questions they ask. “Are we safe to proceed?” is weaker than “What has changed, what are we assuming, and who has independently checked the control?” The latter requires evidence and makes uncertainty discussable.
It also includes how leaders respond to bad news. A defensive response teaches people to manage the message. A disciplined response asks what was known, when it was known, what made escalation difficult and what needs to change in the system. Accountability remains necessary, particularly where standards have been deliberately ignored. But blame used as a substitute for learning makes future warning signs less visible.
The gap between policy and practice
Complex organisations commonly have mature management systems and capable safety professionals. Their risk is not always the absence of process. It is drift between the process as designed and the process as performed.
This gap widens when teams are stretched, interfaces are unclear or leaders are distant from the practical constraints of delivery. A procedure may be technically correct but difficult to apply in the conditions people actually face. If people cannot safely challenge that mismatch, workarounds emerge quietly.
The answer is not to abandon discipline in favour of flexibility. It is to understand where the operational system is creating predictable pressure, then make it possible for people to raise and resolve it early. Strong cultures combine high standards with accurate reporting of reality.
Speaking up is a safety control, not a sentiment
Organisations often measure reporting volumes and assume that more reports automatically indicate a stronger culture. The picture is more complicated. A rise in reporting may show growing trust. It may also reveal a deteriorating operation. A fall may indicate improvement, or it may indicate resignation.
The useful question is whether people can speak up at the point of risk, to the person who can act, without excessive personal exposure. That requires more than a reporting channel. It requires supervisors who listen without reflexively defending the plan, managers who close the loop and leaders who remove recurring barriers.
Psychological safety is sometimes mistaken for being agreeable. In safety-critical work, it means something more demanding: the ability to raise a concern, challenge an assumption or admit an error while maintaining professional standards. Teams need constructive challenge, not false harmony.
This is particularly significant across client, contractor and supply-chain boundaries. Different organisations may bring different histories, incentives and levels of confidence in challenging authority. If expectations are not made explicit, the most junior or commercially exposed party can become the least likely to raise a concern. That is a material control failure, not merely a relationship issue.
Where cultural drift becomes visible
Cultural problems rarely announce themselves as cultural problems. They appear in patterns that operational leaders can observe: repeat findings, late escalations, recurring rework, poor handovers, unresolved actions and decision bottlenecks. They also appear in language.
Phrases such as “we have always done it this way”, “there is no point raising it” or “just get it over the line” should concern leaders. They indicate that informal norms may be overriding formal expectations. The issue is not the isolated phrase. It is the shared belief behind it.
A credible assessment therefore needs to go beyond engagement scores or a one-off workshop. It should examine behaviour across levels and functions: how decisions are made, where challenge stops, whether teams understand accountabilities, and what changes when the workload rises. Quantitative data can identify patterns, while facilitated conversations and operational observation explain why those patterns exist.
TEG uses behavioural insight through TEGinsight to help leaders map the behaviours influencing performance, rather than relying on broad cultural labels. That distinction enables action. “Trust is low” is too vague to manage. “Supervisors do not escalate emerging issues because previous escalations were left unresolved” points to a specific leadership and operating problem.
Improving safety culture without creating another initiative
Safety culture work fails when it is detached from the way work is planned, governed and reviewed. Posters, campaigns and generic training can raise awareness, but awareness alone does not change the decisions made at 02:00 during an abnormal operating condition.
Start with a defined operational risk. It may be weak shift handover, inconsistent contractor challenge, delayed stop-work decisions or poor learning from near misses. Establish the behaviours that currently sustain the problem, the conditions that trigger them and the leadership actions required to change them.
Then make the new standard observable. Instead of asking leaders to “be more visible”, define what visibility means: attending pre-job briefs, asking about control effectiveness, following up reported constraints and feeding outcomes back to teams. Instead of telling managers to encourage challenge, assess whether they invite dissent before decisions are made and whether they respond to it constructively.
Measurement matters because culture change can otherwise become a matter of opinion. Use a balanced view. Leading indicators may include the quality and timeliness of escalations, closure of corrective actions, evidence of challenge in reviews and consistency of supervisory behaviours. Lagging indicators still matter, but they tell leaders what has already happened. The aim is earlier visibility of drift.
There are trade-offs. Excessive reporting requirements can overwhelm teams and obscure the most significant issues. Over-standardising every conversation can turn judgement into compliance theatre. The objective is not a perfect score or a frictionless workplace. It is a culture where people recognise risk, apply disciplined judgement and act before pressure becomes failure.
Safety is shaped in ordinary moments
Culture affects safety long before an incident investigation identifies it. It is shaped in planning meetings, handovers, informal conversations and the response to the first inconvenient concern. The organisations that manage this well do not rely on good intentions. They make expected behaviours clear, equip leaders to hold the line under pressure and test whether those behaviours are present in the work.
The most useful next question for any leadership team is not whether its safety values are well written. It is whether a capable person, facing real delivery pressure today, would feel able to say: “Stop. We do not yet understand this risk.”