Measuring success in battle by counting casualties
Discussion
I've just been reading a book on performance measurement in companies, but one section struck me as being very relevant to road safety. In "Transforming Performance Measurement", Dr Dean Spitzer comments on employee safety measurements:
What is the relevance to driving? Well look at the highlighted text. Measuring failure is what we have -- KSIs, exceeding speed limits -- while accidents are heavily influenced by external factors ("chance" ) that mean that dangerous behaviour rarely results in accidents (look at that number 330!!). I think this fairly reflects the beliefs we have of the relationship of so-called dangerous activity to resulting accidents.
So what should be done? The book talks briefly about measuring (and therefore encouraging/reinforcing) safe behaviour, but it doesn't give the detail I'd like -- it's just 1 page in nearly 300 after all. But consider a road safety strategy that focused on encouraging safe driving, rewarding observed safe driving rather than penalising dangerous driving (which would be handled by providing advice and guidance, at least initially). Imagine if normal drivers welcomed the sight of a police car because they would then have the opportunity to gain safe driving points that led to financial rewards of some kind (eg. insurance-related) or that could be offset against penalty points.
The strongest argument against our current regime is, IMHO, the "330" figure. Who know exactly what the figure would be for road safety? But I'm sure the figure isn't "1". This supports the evidence that we see from insurance companies, ie. that points on your licence have little correlation with increased risk of making a claim. That's because the link is so tenuous as to be essentially random. And also our own experience that we see thoughtlessness and danger many times a day but it rarely leads to accidents because of others' foresight and awareness. That foresight and awareness needs to be rewarded rather than taken for granted.
Dean Spitzer said:
As Daniel Patrick O'Brien explains, most safety measurements "measure past efforts, loss events, problem areas, and past trends. They are totally dedicated to how things were...". Safety measurement tends to focus on injury statistics (such as lost-time accidents) and safety rule compliance issues (safety violations). Almost all accident data are failure-based measurements such as: the number of injuries/deaths, number of lost work days, number of spills, cost of accidents, number of safety violations, and so on. To make matters worse, accident statistics tend to be incomplete, because some accidents aren't reported because of peer pressures or because they don't result in lost-time injuries.
Trying to manage safety by counting accidents is like trying to fight a battle by measuring the number of casualties. Measurement used to show how well, or poorly, you did is going to be of little help for improving things. Furthermore, accident rates are due to chance as much as any other factor. This is because there are lots of unsafe conditions in workplaces and people engage in many unsafe behaviours that do not result in accidents. Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident! When an injury does occur, it is often due to bad luck. Companies with low injury rates may actually have big safety problems; they are often just lucky.
He then goes on to describe an alternative approach called "safe behaviour", where safe behaviour is measured rather than dangerous behaviour, because what is measured is often achieved and it is under the control of the employees whereas accidents are often due to "bad luck".Trying to manage safety by counting accidents is like trying to fight a battle by measuring the number of casualties. Measurement used to show how well, or poorly, you did is going to be of little help for improving things. Furthermore, accident rates are due to chance as much as any other factor. This is because there are lots of unsafe conditions in workplaces and people engage in many unsafe behaviours that do not result in accidents. Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident! When an injury does occur, it is often due to bad luck. Companies with low injury rates may actually have big safety problems; they are often just lucky.
What is the relevance to driving? Well look at the highlighted text. Measuring failure is what we have -- KSIs, exceeding speed limits -- while accidents are heavily influenced by external factors ("chance" ) that mean that dangerous behaviour rarely results in accidents (look at that number 330!!). I think this fairly reflects the beliefs we have of the relationship of so-called dangerous activity to resulting accidents.
So what should be done? The book talks briefly about measuring (and therefore encouraging/reinforcing) safe behaviour, but it doesn't give the detail I'd like -- it's just 1 page in nearly 300 after all. But consider a road safety strategy that focused on encouraging safe driving, rewarding observed safe driving rather than penalising dangerous driving (which would be handled by providing advice and guidance, at least initially). Imagine if normal drivers welcomed the sight of a police car because they would then have the opportunity to gain safe driving points that led to financial rewards of some kind (eg. insurance-related) or that could be offset against penalty points.
The strongest argument against our current regime is, IMHO, the "330" figure. Who know exactly what the figure would be for road safety? But I'm sure the figure isn't "1". This supports the evidence that we see from insurance companies, ie. that points on your licence have little correlation with increased risk of making a claim. That's because the link is so tenuous as to be essentially random. And also our own experience that we see thoughtlessness and danger many times a day but it rarely leads to accidents because of others' foresight and awareness. That foresight and awareness needs to be rewarded rather than taken for granted.
Edited by Peter Ward on Tuesday 21st August 22:24
Spitzer gets the prize for stating the bloody obvious; engage more often in something that presents a hazard and you have more chances of having an accident.
"Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
"Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
Edited by shack55 on Tuesday 21st August 23:50
shack55 said:
Spitzer gets the prize for stating the bloody obvious; engage more often in something that presents a hazard and you have more chances of having an accident.
"Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
If his statement is obvious, then perhaps we should be thinking of why it's not acted on more often."Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
Edited by shack55 on Tuesday 21st August 23:50
He is saying that the focus should be on increasing safe behaviour rather than reducing unsafe behaviour. The same thing? No, there's a fundamental difference. One is positive and the other is negative. A positive focus is more productive, he suggests, than a negative one, while rewarding good behaviour is more effective than penalising bad behaviour (carrot vs stick). I imagine that one reason for this is the natural response to the "that could be dangerous" warning is "I've done it 329 times before and I've never had a problem".
This should not be reduced to a discussion on speed, but consider this. If a speed camera provided the means to reward people for driving below the speed limit -- perhaps even graduated rewards -- then people would be desperate to have as many cameras as possible. They would be delighted to see them, and would welcome covert speed traps because they knew that they could receive unexpected rewards. That's positively rewarding good behaviour rather than penalising bad behaviour.
If that's obvious then it's a shame we don't do more of it.
Bing o said:
peter ward said:
But consider a road safety strategy that focused on encouraging safe driving
It's called NCB?Not a strategy as such, but I trust insurance companies more than the taxernment...
Perhaps we should be asking - not for a change in strategy - we all want safer driving - but better research into what is safe.
I think it's the rule of threes again for driving - when you drive unsafely, you have less time to respond to surprises, which come along every now and again. Reactions save you most of the time and you get a "near-miss". However, when dealing with a near-miss, if one other thing goes wrong, you're having an accident. You swerve to miss the door opening and hit the cyclist coming the other way; you brake sharply for the person who steps out onto the zebra and get collected by the arctic behind you...
Peter Ward said:
shack55 said:
Spitzer gets the prize for stating the bloody obvious; engage more often in something that presents a hazard and you have more chances of having an accident.
"Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
If his statement is obvious, then perhaps we should be thinking of why it's not acted on more often."Research has indicated that, on average, a worker would have to engage in an unsafe behaviour 330 times before it resulted in an accident!" Spitzer and friend.
Conversely then, engage in fewer unsafe behaviours and the chances of having an accident reduce, engage in no unsafe behaviours and the chances of having an accident are massively reduced; they won't return to 0, they are accidents after all.
"Measurement used to show how well, or poorly, you did is going to be of little help for improving things."
I disagree; this sort of measurement highlights unsafe behaviour and hence improvements can be made; little use is very pessimistic, perhaps Spitzer is showing the quality of his writing and thought on the matter, this is a short passage so saying the analysis is shallow may be harsh but shallow it seems to be at best I'm afraid.
I see the relevance to driving being this: reduce unsafe behaviour, reduce the incidence of accidents.
Is that what he is saying? I rather think it is.
Those who choose to walk closer to the edge are more likely to fall over it.
Edited by shack55 on Tuesday 21st August 23:50
He is saying that the focus should be on increasing safe behaviour rather than reducing unsafe behaviour. The same thing? No, there's a fundamental difference. One is positive and the other is negative. A positive focus is more productive, he suggests, than a negative one, while rewarding good behaviour is more effective than penalising bad behaviour (carrot vs stick). I imagine that one reason for this is the natural response to the "that could be dangerous" warning is "I've done it 329 times before and I've never had a problem".
I've been saying on here for ages that the big problem with driving errors (as a learning tool), is that they don't result in collisions often. The trained/experienced observer will see the potential that one could have resulted though. A lot of people on here don't worry about potential, they seem to think it's only a problem if it actually results in a collision, near misses don't count. Unfortunately the driver 'who gets away it' begins to see the risk associated with their errors as a diminishing one (if they saw it as a risk to start with that is) & then when it does go all wrong they can't accept that they have contributed to the collision, because they've driven that way for years & haven't had a collision. That's why they stand their blaming each other.
Peter Ward said:
This should not be reduced to a discussion on speed, but consider this. If a speed camera provided the means to reward people for driving below the speed limit -- perhaps even graduated rewards -- then people would be desperate to have as many cameras as possible. They would be delighted to see them, and would welcome covert speed traps because they knew that they could receive unexpected rewards. That's positively rewarding good behaviour rather than penalising bad behaviour.
If that's obvious then it's a shame we don't do more of it.
And what happens with NCB ?If that's obvious then it's a shame we don't do more of it.
Those who build it up pay less, or do they ?
No, in fact those who don't/can't build it up just don't bother with insurance because they can't afford it & when they have collisions those with NCB pay it through increased premiums to cover for it.
Edited by vonhosen on Wednesday 22 August 11:22
vonhosen said:
And what happens with NCB ?
Those who build it up pay less, or do they ?
No, in fact those who don't/can't build it up just don't bother with insurance because they can't afford it & when they have collisions those with NCB pay it through increased premiums to cover for it.
Let's not forget that UK NCB mostly stops at four years.Those who build it up pay less, or do they ?
No, in fact those who don't/can't build it up just don't bother with insurance because they can't afford it & when they have collisions those with NCB pay it through increased premiums to cover for it.
No point in being a better driver once you reach that level.
More sensible countries run NCB up to twenty years or more.
Then there is a sound commercial reason to be a better driver.
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff