So the farmer wants to sell his pig at market for the best price. He wants the pig to be as heavy as possible. So he weighs the pig to see how heavy it is. Next day he weighs the pig again. He writes each weight down in a notebook. He creates a database on his i-pad (very 21st century farmer this one). He plots a graph of weight against time and toys with line graphs and bar charts to find the jazziest way to present the data. He weighs the pig each and every day to track its progress.
Only thing is, if you want a pig to get fatter you have to feed it, not weigh it.
I came across this anecdote in conversation with my brother-in-law a few weeks ago.
“How’s work going?”, I asked.
“Ah, it’s alright”, he said. “Too much pig weighing though”.
“What?” I said; “I thought you worked in the aerospace industry?”
…and so he explained the story above and the fact that in his large international company there was a fixation with performance metrics and pig weighing.
And that got me to wondering about the use of performance metrics in health and safety management. And all of the various ingenious and delightful ways that we (…when I say we, I really mean I, but lets depersonalise this and stay objective for a moment) …that we use to measure health and safety performance, from the lowly lost-time accident incidence rate, to frequency rates, severity rates, number of near-misses reported, number of risk assessments reviewed, number of safety briefing held to schedule;… to the full-on quantitative scored audit with marks out of 100 for compliance (or conformance if you prefer) to standard. Not to mention climate survey tools and stress survey questionnaires. There are others. Your company or organisation probably uses a few. If your company is large then it probably uses a lot. And if you are studying NEBOSH Diploma Unit A then you will have read about the use of performance metrics and statistic as you work through some of the elements. In particular elements A2 and A3.
And yes I accept the age-old saying that “if you can’t measure it, you can’t manage it”.
But there is an alternative view. Well perhaps complementary view is a better way to put it. And it is one that I would subscribe to. If you want to know how well your organisation is doing at health and safety management, get up off your rear end; switch the PC monitor off; walk out the door and go and have a look! Talk to people. Watch them work.
That’ll tell you a lot about how good your safety management system is.
Now before we get distracted there are two points that I wanted to make here.
The first is that in larger organisations, in particular, people can get very disgruntled about the amount of data collection and recording that they have to do. Everything has to be recorded so that the stats can be collated, manipulated, plotted and analysed. And this gives rise to a myriad graphs and numbers that are then thrown back at managers in briefing packs and weekly / monthly / quarterly performance reports (delete as appropriate). The end result; information overload and an inability to see the wood for the trees.
The second is that when you measure something there will inevitably be a desire to see an improvement in that particular performance metric. The graph of accident incidence rate should show a gradual downward slope; “that’s good, well done”. An upward slope; “that’s bad, that is most definitely not well done”. This leads inexorably to expectations, targets and objectives. And the problem with them is that as soon as any target is announced the sharp-eyed manager, keen to impress, will start to work towards that target, because that is how their performance is being measured.
And with that comes a world of pain. Or rather I should say a world of perverse decisions and management practices that in fact work against the aims and aspirations of the organisation but allow achievement of target. If you want a classic example of this in action look at the some of the target driven perverse management practices that happened historically in the NHS and for all I know still happen today Straight Statistics.
Set a target of ten near miss reports per department and what do you get; line managers wandering around on a Friday afternoon looking for trivial safety issues to write down on a near miss report form because they haven’t hit the target yet.
Set a target of five lost-time accidents in the year and what so you get; a worker signed off on holiday for a fortnight the day after they broke their fingers because “that doesn’t count; he’s on holiday and therefore it ain’t lost time”. Incidents get covered up and hidden so that the target can be hit.
So don’t measure it unless you really need to know.
Keep the metrics to a minimum.
Take care with the way performance metrics are used and in particular great care with the setting of targets.
And finally, always check that what you think the metric is showing you is in fact the case. Always check for ground truth. This expression comes from the world of remote sensing. The imaging satellite is telling you something about the surface of the planet. You need to check that location to verify that what it is telling you is correct. Only by calibrating in this way can you start to trust what the satellite is telling you.
I once worked for an operations manager who ran a factory from his office. He sat in his office and his management team brought information to him, much of it in the form of data and some in the form of descriptions about the way things were. He went in the factor twice a year. And promptly hit the roof every time because how he had imagined it to be was not how it was. Reality did not match his expectation. Check for ground truth.
So get off your seat and go take a look.
Jim Phelpstead BSc, PhD, CMIOSH
Working in health and safety for 18 years Jim is a long-standing RRC Associate Tutor, who loves the great outdoors.