The right to refuse is the sharp end of the internal responsibility system, and the part employers most often mishandle — not by denying it, but by treating it as a grievance rather than as a procedure with statutory steps and a clock.
What it means in practice
- The threshold is belief with reason, not proof. The worker does not have to be right; they have to have grounds. Whether the danger was real is decided by the investigation, not before it.
- The worker stays available. They remain at a safe place near the workstation, on pay, until the matter is resolved. Sending them home is where reprisal complaints usually start.
- There are two stages. The employer investigates first, in the presence of the worker and of a worker member of the joint committee or a representative. If the refusal stands after that, the regulator is called and an inspector decides.
- The work cannot simply be handed to somebody else. Where another worker is asked to do it, they must be told, in the presence of a committee member, that the work is the subject of a refusal and why.
- Reprisal is prohibited and separately enforceable. Discipline that follows a refusal will be examined on the employer’s evidence, not the worker’s.
What it is not
It is not universal in scope. Police, firefighters, health care and some correctional work face limits where the danger is inherent to the job, and the wording of those limits differs by jurisdiction.
It is not a strike or a work-to-rule. It is individual, tied to a specific task and a specific hazard, and it ends when the hazard is dealt with.
And it is not the same instrument as the European right to withdraw. France’s droit de retrait and Italy’s equivalent let a worker stop work in the face of a grave and imminent danger, but neither comes with a two-stage investigation procedure, a committee member at the employer’s side, or an inspector’s ruling to close it. Same right on paper, a different machine underneath.