SFAIRP Gone Mad?

A few years ago – around the time the Health & Safety at Work Act was coming into effect here in New Zealand – a friend and I met for a beer. He started to complain to me about his employer’s latest policy. This workplace was a light industrial outfit, with an upstairs office, and – relevantly here – two staircases.



The employer had designated one staircase strictly for ascending traffic, and the other for descending. The idea was to avoid falls due to people bumping into each other. My friend, telling me this, was fuming at the silliness of it all (he even broke out the classic cliché of “health and safety gone mad!”). From the sounds of it, his colleagues felt much the same.

The company’s reasoning, I learned, was “Well we’ve thought of it, and we have the means to do it, so now we must do it – right? Isn’t that what ‘So Far As Is Reasonably Practicable’ means?”

This is indeed a bit of an oddball case of SFAIRP principles. More usually, we satisfy our prescriptive obligations, align with good practice, and exploit all the cheap and effective options. Then, as the scrape marks start to appear in the bottom of the barrel, we must break out the formal tools of SFAIRP assessment to decide how much more to do. In practice, this level of scrutiny is often reserved for the most drastic actions: potentially very effective, but also very onerous. This example with the stairs, however, is a bit different: what becomes of those ideas that are trivially easy and cheap to implement, but also… just not very good?


In heavier industry, I’ve seen a few clients go on a similar journey with low-pressure leak detection in pipelines. It’s certainly not too onerous – if there’s already a pipeline boundary shutdown valve, all it might take is a new pressure transmitter and a trip. But I’ve seen those same companies come to realise that it’s not a silver bullet, and in a lot of cases, would simply never be triggered at all, even in the worst leak scenarios.

Returning to my friend’s work stairs, let’s set out the factors we need to consider. We’re all familiar with stairs, so in this case our lived experience is a fair guide. Presumably we must say the likelihood of an accident is low – very specifically, falling down the stairs due to two people bumping into each other, which is all that the one-way traffic system sets out to address. For sure, I’ve known people who’ve fallen down stairs, but at least anecdotally I’ve never heard of it happening to anyone in this way. It is, however, a potentially serious event – if they both take a tumble, that could be two casualties with concussions or broken bones.

Screenshot 2022-09-01 145030Ad: Introduction to Bowties E-Learning From $130 USD




Some stairways implement a “keep left” rule, so that might be worth considering. Of course, we should also encourage good stair safety in general – don’t carry an open coffee, for example – but since this does not relate to people bumping into each other, we will set it aside for now. I don’t believe there’s any known example of one-way stairs (naturally, fire exits and the like are usually used in one direction – but to my knowledge, going the other way is never prohibited. And before you say it, no I’m not counting escalators!) By and large, this is usually just left to people’s own situational awareness.

Finally, and strictly subordinate to the considerations above, we may look at the cost. On its surface, the one-way stairs are certainly easy and low-cost. All it would take is some printing and a few minutes of somebody’s time to post some ‘Up Only’ and ‘Down Only’ signs. A Stairs Champion could remind people in passing. Once the rule is in effect, a few added seconds of travel time, a few times a day, will certainly not bankrupt the business.

But the true cost here goes deeper. Now, HSWA (NZ) and the Model WHSA (Australia) make no distinction between tangible and intangible costs in their definitions of “reasonably practicable”, but I would argue the intangible costs deserve some attention. These are perhaps best reflected in my friend’s cynicism towards this whole situation, his cry of “health and safety gone mad”. The company had gained a flimsy control against one very niche type of staircase fall, but in exchange, had turned their safety systems into something the staff now saw as a joke. This can sow an insidious doubt. “Well, this rule is stupid, so which other rules around here are stupid? Do our health and safety people even know what they’re doing? Maybe I’d better just ignore the rules and rely on my own common sense instead…”. The trust in the safety system is broken, and the willingness to participate in, and contribute to, that system starts to fall apart.

The pipeline example from earlier most likely does not live in the same camp. Low-pressure leak detection might not work very well in some situations, but it doesn’t invite the same scorn as an absurd staircase rule. It’s not a ridiculous idea; it’s just an idea that doesn’t work sometimes. Thus, it may still be worth it.

More generally, we must always keep an eye on the controls we put in place. It’s easy to add another rule to the pile, but this can turn into so much fluff and clutter. It’s trickier to build an elegant safety system that is simple, memorable, effective, and carries the trust and enthusiasm of the people who use it. To achieve that, there comes a time for restraint.

Was my friend’s company right to implement their one-way stairs rule? The price may have been near zero, but I would say the cost was grossly disproportionate to the benefit. What do you think?