The question of bonuses has been around since 1890. It became popularised through the term ‘agency theory’, used from the 1920s onwards. The term described how an individual could incentivise an agent to act in the best interests of the investor – and recoup maximum profits. The assumption was, the more an agent was paid, the more incentivised they would be.

Where the task is obvious and the target clear, bonuses will work. However, when it comes to safety, this requires the individual to be aware of contingencies. For example, are they conscious of where the infrastructure is getting weak, but such weakness is not yet obvious and may not require immediate attention? Safety is a philosophy, a way of life. How can a bonus deal with that?

It is therefore concerning that BP is considering paying bonuses for safety when that is a basic requirement of a managers job. BP needs to analyse how safety is regarded within its organisation – that is the real issue. The notion that someone will not take a safety concern seriously until they get their bonus is absurd.

The Kingston University research has merits. What the financial crisis has shown us is that bonuses can work. Each of the individual transactions that bankers conducted were probably well handled. However no one was thinking about the limited assets they were trading, whose value inflated to the point where no one wanted to invest in them.

Poor strategy cannot therefore be improved by bonuses, and nor can safety.