Safety coaching vs safety policing: why the approach matters
- Jul 15, 2025
- 7 min read
Updated: Apr 14
There are two ways to respond when you see someone doing something unsafe at work. You can punish them, or you can coach them. Both aim to prevent the unsafe behaviour from happening again. But only one of them actually works long term.
This isn't a philosophical debate. The difference between a coaching-led safety culture and a policing-led one shows up in incident rates, near-miss reporting, worker engagement, and (ultimately) whether people go home safe at the end of the day.
What "safety policing" looks like
Most organisations don't set out to create a policing culture. It tends to develop gradually, often with good intentions. A serious incident occurs, management responds with stricter rules and harsher consequences, and over time the entire safety programme becomes focused on catching people doing the wrong thing.
In a policing culture, safety officers are seen as enforcers. Their job, in the eyes of the workforce, is to find violations and hand out consequences. Research from the CBIA describes these as "forced" safety cultures, where health and safety officers are perceived as police-like figures whose constant enforcement creates fear, not engagement.
The problems with this approach are well documented. When people are afraid of punishment, they stop reporting near misses. They hide minor incidents. They don't speak up when they see something unsafe, because the reporting system feels like it exists to get people in trouble rather than to learn from what went wrong.
The Agency for Healthcare Research and Quality (AHRQ) identifies non-punitive responses to error as a defining feature of high-reliability organisations. Industries like aviation have built strong safety cultures around continuous improvement precisely because they moved away from individual blame and towards systems thinking. Workplace safety is still catching up.
What safety coaching looks like
Coaching starts from a different premise entirely. Instead of asking "who did the wrong thing?" it asks "what happened, why did it happen, and how do we prevent it from happening again?"
In a coaching culture, supervisors and safety leaders respond to safety events with curiosity, not consequences. They gather the team, review what occurred, and have a constructive conversation about the conditions that led to the event. The focus is on learning and improvement, not punishment.
This isn't about being soft on safety. It's about being effective. Gallup's meta-analysis of more than 82,000 business units across 230 organisations found that workplaces with high employee engagement experience 70% fewer safety incidents compared with those in the bottom quartile. Engagement doesn't come from fear. It comes from feeling valued, trusted, and involved.
Research from Frontiers in Psychology breaks safety leadership into three dimensions: safety controlling (rules and monitoring), safety coaching (role modelling, participation, and motivation), and safety caring (trust and concern for wellbeing). The study found that all three contribute to safety outcomes, but the coaching and caring dimensions are what drive proactive participation, where workers go beyond minimum compliance to actively contribute to a safer workplace.
This matters because compliance alone doesn't prevent injuries. The most dangerous moments at work are the unscripted ones: the unusual load configuration, the blocked sightline, the unfamiliar shift pattern. Rules can't cover every scenario, but a workforce that has been coached to think critically about risk can adapt to any of them.
Why policing drives underreporting
One of the most damaging consequences of a punitive culture is that it destroys the data you need to actually improve. If workers are afraid of being blamed for near misses, they simply stop reporting them.
The National Center for Biotechnology Information (NCBI) published research showing that 42% of safety incident reports in one national database were used to censure others, and that fear of being held personally accountable is a primary driver of reluctance to use reporting systems. Separate research in the same field confirmed that underreporting remains a significant challenge, often driven by fear of punitive actions, reputational damage, or legal consequences.
When near-miss data dries up, organisations lose the leading indicators that predict serious incidents. They're left relying on lagging indicators (injury rates, lost-time incidents, workers' compensation claims) which only tell you something after someone has already been hurt.
A coaching culture reverses this dynamic. When workers trust that reporting an event will lead to a constructive conversation rather than a disciplinary one, reporting rates increase. And with better data comes better decision-making about where risk actually sits and how to address it.

How coaching works in practice
Coaching isn't a single technique. It's a set of consistent behaviours from supervisors and safety leaders that, over time, reshape how the entire team relates to safety. Some of the most effective coaching practices include:
Reviewing safety events as a team, using blurred or anonymised footage where available, and asking the group what they see, what could have gone differently, and what changes would help. This shifts the dynamic from "someone is in trouble" to "we're solving a problem together."
Running toolbox talks that are interactive rather than scripted. Workers remember conversations far better than they remember a safety officer reading from a clipboard.
Recognising safe behaviours, not just correcting unsafe ones. When a team member makes a smart decision under pressure, calling it out reinforces the behaviour across the whole group.
Sharing data openly. When workers can see the trends (where events are happening, whether things are getting better or worse, what changes have made a difference) they become invested in the outcomes.
Finch Consulting's research highlights that coaching empowers workers to identify and address safety concerns themselves. Instead of waiting for an instruction or a rule change, coached workers develop the critical thinking skills to handle unforeseen situations. In dynamic environments like warehouses, manufacturing plants, and logistics operations, that adaptability is invaluable.
Where technology fits
Technology can amplify either approach. If it's implemented as a surveillance tool that watches workers and punishes them for violations, it will entrench a policing culture. If it's implemented as a coaching tool that generates data for constructive conversations, it can accelerate culture change.
This is the distinction that matters when evaluating computer vision AI platforms. A system that detects safety events is only as valuable as the response it triggers. At inviol, the platform is designed around a coaching-first philosophy. When the AI detects an event, the intended response is a team coaching conversation, not a disciplinary process.
Faces are blurred so individuals aren't identified. Events are classified by type and severity, with high-priority events surfaced first so supervisors can focus their coaching time where it matters most. Over time, the data builds a picture of trends and patterns that informs not just individual coaching sessions but broader process improvements, like adjusting delivery schedules to reduce pedestrian-vehicle conflicts or redesigning traffic routes based on heatmap insights.
The result is a feedback loop: better data leads to better coaching, better coaching leads to safer behaviour, and safer behaviour shows up in the data. inviol customers see an average 67% risk reduction through this approach.

The regulatory case for coaching
In New Zealand, the Health and Safety at Work Act 2015 (HSWA) requires PCBUs to engage workers in health and safety matters and to have clear, effective, and ongoing ways for workers to suggest improvements or raise concerns. WorkSafe NZ's position is explicit: when workers are involved, engaged, and represented, it leads to healthier and safer work, and it boosts business performance because workers suggest practical, cost-effective ideas and are more likely to put them into use.
A policing culture undermines these obligations. If workers are afraid to speak up, you aren't meeting your duty to provide reasonable opportunities for participation. A coaching culture, by contrast, creates the environment where genuine engagement happens naturally.
In Australia, similar principles apply under the model Work Health and Safety laws, which require consultation, cooperation, and coordination between duty holders and workers. The intent is the same: safety improves when workers have a genuine voice, not just a rulebook to follow.
Making the shift
Moving from a policing culture to a coaching culture doesn't happen overnight. It requires visible commitment from leadership, consistent behaviour from supervisors, and a willingness to reframe how the organisation responds to safety events.
The good news is that the shift produces results quickly. ProAct Safety reports that organisations implementing leadership safety coaching have experienced accident reductions of up to 55% in the first year, with some achieving even greater improvements over time. Gallup's research confirms that even incremental improvements in how leaders communicate, coach, and recognise safe behaviour lead to measurable improvements in frontline safety outcomes.
Start by changing the conversation. The next time a safety event occurs, resist the urge to find someone to blame. Gather the team, review what happened, and ask: what can we learn from this?
That's the difference between policing and coaching. One punishes the past. The other builds a safer future.
Ready to see what a coaching-first safety platform looks like? Book a demo and we'll show you how inviol turns safety events into coaching conversations that drive lasting behaviour change.

Frequently Asked Questions
What is the difference between safety coaching and safety policing?
Safety policing focuses on catching and punishing unsafe behaviour, relying on fear of consequences to drive compliance. Safety coaching focuses on understanding why safety events occur, having constructive conversations with teams, and building the skills and knowledge that prevent events from recurring. Research consistently shows coaching produces better long-term safety outcomes than policing.
Why does a punitive safety culture lead to underreporting?
When workers fear being blamed or disciplined for reporting safety events, they stop reporting. Research shows that fear of punitive actions is one of the primary barriers to honest safety reporting. This means organisations lose the leading indicator data they need to identify and address risks before someone gets hurt.
How does employee engagement affect workplace safety?
Gallup's meta-analysis of over 82,000 business units found that workplaces with high employee engagement experience 70% fewer safety incidents compared with those with low engagement. Engaged workers are more attentive, more likely to follow safe practices, and more willing to speak up about hazards, all of which directly reduce risk.
Does New Zealand law require worker engagement in safety?
Yes. The Health and Safety at Work Act 2015 (HSWA) requires PCBUs to engage workers in health and safety matters and to provide clear, effective, and ongoing ways for workers to participate in improving safety. WorkSafe NZ has stated that workplaces where workers are involved and engaged achieve better health and safety outcomes.
How does technology support a coaching approach to safety?
Computer vision AI platforms like inviol detect safety events and provide the data and footage needed for constructive coaching conversations. With features like face blurring, event classification by severity, and trend reporting, the platform is designed to support learning and improvement rather than surveillance and punishment.


