Accidents go DOWN as speed limits go UP...
Discussion
Canadian video that seems very well researched and is based on state research (and research facts that are ignored by other state depts so as they can issue tickets...)
https://www.youtube.com/watch?v=2BKdbxX1pDw
Still not sure what a "pocket book" is mind...a wallet?
https://www.youtube.com/watch?v=2BKdbxX1pDw
Still not sure what a "pocket book" is mind...a wallet?
Foliage said:
Yeah but the chance of death goes up as speed goes up, and I think that's more important than the chance of accidents, id rather see lots of low speed non-injury accidents than an increase in fatal accidents.
That is a fair point mate and maybe an argument for decreasing limits everywhere? Views are very entrenched on these subjects but what is needed is more transparency, more honesty and better logic - especially if it results in less frustration, fewer accidents and better progress. Imagine that
Ken Figenus said:
Views are very entrenched on these subjects but what is needed is more transparency, more honesty and better logic - especially if it results in less frustration, fewer accidents and better progress.
Foliage said:
Yeah but the chance of death goes up as speed goes up, and I think that's more important than the chance of accidents, id rather see lots of low speed non-injury accidents than an increase in fatal accidents.
Sort of.Is it better to have someone paying attention at 40mph when a child runs out, who then applies the brakes hard and slows dramatically. OR someone who is paying little/no attention to the road and mows the child down at 30mph ?
Foliage said:
Yeah but the chance of death goes up as speed goes up, and I think that's more important than the chance of accidents, id rather see lots of low speed non-injury accidents than an increase in fatal accidents.
Very fair point but can you confirm that would actually be the case? Accidents are often described as a pyramid, with fatalities at the top. reduce the base size, reduce the top size?The figures show accidents reducing- maybe that includes fatal ones.
RobinOakapple said:
Ken Figenus said:
Views are very entrenched on these subjects but what is needed is more transparency, more honesty and better logic - especially if it results in less frustration, fewer accidents and better progress.
*may only apply in Canada
Jasandjules said:
Sort of.
Is it better to have someone paying attention at 40mph when a child runs out, who then applies the brakes hard and slows dramatically. OR someone who is paying little/no attention to the road and mows the child down at 30mph ?
Is that not to do with drivers paying attention, and not speed? Is it better to have someone paying attention at 40mph when a child runs out, who then applies the brakes hard and slows dramatically. OR someone who is paying little/no attention to the road and mows the child down at 30mph ?
Yi had a hire car this week and it came with a new "safety system" I've not come across before. It is a speed limiter. You set it to whatever speed you like and then you cannot exceed this speed. Talking to some people at work the general response was great set it to 30 and keep your foot down! I'm not sure this is a safety feature I particularly want on the roads! But "speed kills" so stopping you going over 30 must make you safe, even if it then encourages people to drive with the throttle hard down and not worrying about their speed as they're now "safe" as they cannot go over 30!
*cracks knuckles* Right...
(Sorry, I recently did a module at uni on this and have spent a significant amount of time looking at theory of accident statistics and causation as part of my job, so you are all about to be subject to a diatribe. Y'all gonna lern summthin.)
The poster above refers to accident triangles. These are a commonly used theory in industry that says, basically, that frequency of accidents decreases as severity increases. This means that when stacked up, they form a triangle. Common categories are:
So with this model for accidents we can start to play around with things. The first thing you can do is look at the stats and see if it meets with the model. Yup, there are mercifully few fatalities on the roads these days, but lots of crashes, some quite severe. Then there are uncountable numbers of parking knocks and scrapes and as said above, god knows how many near misses.
Now over time (i.e. decades) we can watch how the triangle changes shape, and think about why it's changing shape. Over time, the roads have got safer in terms of fatalities, despite increasing total journey miles being driven. That is largely down to technological improvements - the introduction of seatbelts, airbags, ABS/EBD/ESC are probably associated with step change reductions in severity of accidents. It has to be said that there has been less improvement in the human factors side of things - driver training and education is very similar really to the methodologies that were used decades ago when compared to the technological progress in car design.
What I am getting at here is that there is a distinction to be made between the top and bottom of the pyramid, and doing something which attacks the top will not necessarily affect the bottom. Similarly, finding a way to eliminate tens of thousands of the low level incidents may do nothing to stop a fatality. This makes sense when you think about it though - if reverse parking sensors were made a legal requirement, it seems likely that the number of parking scrapes occurring would go down, but this will do nothing to stop a high speed fatality occurring on a motorway.
The caution here is to be careful of forgetting that different types of accidents have different causation. Also that accident causation is a very complex subject in and of itself. Speed limits are a very blunt instrument that is applied to driving in an attempt to prevent higher level accidents by limiting the maximum potential consequence as expressed by the kinetic energy stored in the moving vehicle. The faster you go, the harder you can hit the tree basically. However, speed limits will only coincidentally reduce total accident numbers in those accidents where speed is a causal factor (rather than being a decider of consequence severity). Statistics have shown that speed is not the most common factor in accident cause. (I don't have them to hand, but it comes up fairly regularly and I believe Dept. of Transport did a study a while back? Google it.)
What it does come down to more often, rather than speed, is the human factor. Driver error of one form (and yes, inappropriate speed is also a human error, barring Toyota rogue accelerator pedals) or another is the main cause. What I am getting at here is that there will not be any step change in accident safety until that human error is addressed. Unfortunate as it may be for enthusiast drivers, we may be seeing the start of this with the Tesla self-drive mode and Google's autonomous cars. Taking the driver out of the equation will be the step change that completely changes accident statistics and does for driving what airbags and EBD have done from the technical side.
Anyway, coming back to speed limits. Changing speed limits will never directly correlate with changing accident stats for all the reasons above and more. Firstly, there is not a one to one relationship between speed limit and the actual speeds driven. Secondly, a speed limit set at a single fixed point cannot take into account variable conditions of road, car or driver. Thirdly, politicisation and monetisation of transport and speed limit policy will mean that there will never be a pure science based approach to limit setting. The interaction of the general public, the media and the political machine is almost perfectly set up to make bad decisions when it comes to risk-taking and risk-based decision making. The bias will almost always be towards perceived conservatism which means that raising limits will always be a very, very tough sell.
(Sorry, I recently did a module at uni on this and have spent a significant amount of time looking at theory of accident statistics and causation as part of my job, so you are all about to be subject to a diatribe. Y'all gonna lern summthin.)
The poster above refers to accident triangles. These are a commonly used theory in industry that says, basically, that frequency of accidents decreases as severity increases. This means that when stacked up, they form a triangle. Common categories are:
- Fatality - At least one death from accident
- LTI - Lost time Incident, an accident leading to a severe injury meaning that an individual is away from work for several (often 3) days
- Medical/Reportable - An incident leading to injury requiring medical treatment more involved than first aid or requiring seeing a doctor.
- First Aid - An incident leading to... yeah you get the idea.
- "Near Miss" - An incident which had the potential to be an accident but the consequences weren't realised. Maybe the car stopped just before hitting something, or there was no tree there to hit, so the car spun off into a grass verge, etc.
So with this model for accidents we can start to play around with things. The first thing you can do is look at the stats and see if it meets with the model. Yup, there are mercifully few fatalities on the roads these days, but lots of crashes, some quite severe. Then there are uncountable numbers of parking knocks and scrapes and as said above, god knows how many near misses.
Now over time (i.e. decades) we can watch how the triangle changes shape, and think about why it's changing shape. Over time, the roads have got safer in terms of fatalities, despite increasing total journey miles being driven. That is largely down to technological improvements - the introduction of seatbelts, airbags, ABS/EBD/ESC are probably associated with step change reductions in severity of accidents. It has to be said that there has been less improvement in the human factors side of things - driver training and education is very similar really to the methodologies that were used decades ago when compared to the technological progress in car design.
What I am getting at here is that there is a distinction to be made between the top and bottom of the pyramid, and doing something which attacks the top will not necessarily affect the bottom. Similarly, finding a way to eliminate tens of thousands of the low level incidents may do nothing to stop a fatality. This makes sense when you think about it though - if reverse parking sensors were made a legal requirement, it seems likely that the number of parking scrapes occurring would go down, but this will do nothing to stop a high speed fatality occurring on a motorway.
The caution here is to be careful of forgetting that different types of accidents have different causation. Also that accident causation is a very complex subject in and of itself. Speed limits are a very blunt instrument that is applied to driving in an attempt to prevent higher level accidents by limiting the maximum potential consequence as expressed by the kinetic energy stored in the moving vehicle. The faster you go, the harder you can hit the tree basically. However, speed limits will only coincidentally reduce total accident numbers in those accidents where speed is a causal factor (rather than being a decider of consequence severity). Statistics have shown that speed is not the most common factor in accident cause. (I don't have them to hand, but it comes up fairly regularly and I believe Dept. of Transport did a study a while back? Google it.)
What it does come down to more often, rather than speed, is the human factor. Driver error of one form (and yes, inappropriate speed is also a human error, barring Toyota rogue accelerator pedals) or another is the main cause. What I am getting at here is that there will not be any step change in accident safety until that human error is addressed. Unfortunate as it may be for enthusiast drivers, we may be seeing the start of this with the Tesla self-drive mode and Google's autonomous cars. Taking the driver out of the equation will be the step change that completely changes accident statistics and does for driving what airbags and EBD have done from the technical side.
Anyway, coming back to speed limits. Changing speed limits will never directly correlate with changing accident stats for all the reasons above and more. Firstly, there is not a one to one relationship between speed limit and the actual speeds driven. Secondly, a speed limit set at a single fixed point cannot take into account variable conditions of road, car or driver. Thirdly, politicisation and monetisation of transport and speed limit policy will mean that there will never be a pure science based approach to limit setting. The interaction of the general public, the media and the political machine is almost perfectly set up to make bad decisions when it comes to risk-taking and risk-based decision making. The bias will almost always be towards perceived conservatism which means that raising limits will always be a very, very tough sell.
Rovinghawk said:
Nice to see a thought-out post. I hope the RSI from the typing subsides quickly.
I'm just doing that to procrastinate from having to do the EVEN MORE long winded and convoluted environmental risk assessment of a subsea oil field expansion off the coast of Norway... 
Would far rather have an argument with forum folks about speed limits than with Norwegian venture capitalists about why their selected field architecture concept is insufficiently well characterised to get it past the regulators. Sigh.Rovinghawk said:
Ryvita said:
Stuff.
Nice to see a thought-out post. I hope the RSI from the typing subsides quickly.Although raising speed limits is a tougher sell than lowering them, I think we should do what is best rather than what is easiest.
That's where the whole "inappropriately low limits" argument fails, who decides what is appropriate and what isn't?
RobinOakapple said:
That's where the whole "inappropriately low limits" argument fails, who decides what is appropriate and what isn't?
Well a starting point in the debate might be to agree that inappropriately low limits exist. I presume you'd agree that having a 1mph speed limit on all roads would inappropriately low?RobinOakapple said:
who decides what is appropriate and what isn't?
There are a large number of stakeholders to something as wide ranging as speed limit policy. Everyone in the country is affected at least on some level by it, from its affect on the driving experience, through to noise and pollution issues, and the effect that it has on things like haulage prices and therefore indirectly inflation and high-street pricing.So who should get a say? Everyone? It is arguable that that is what does happen, in that we live in a democracy (No, no. Stop it. That's a different argument. Not now.) and our elected representatives and the civil service work for all of us. However, as we know, there are significant pressures placed on this process by the lobbying industry both positive and negative which means that some people may feel that they are under-represented by the process.
Retreating again to the theoretical world, policy should be set in a scientific way based on a cost-benefit process. Some all powerful entity should amass all the necessary evidence on the economic, societal and environmental impacts both positive and negative and should impartially select the set point at which the negatives are minimized and the positives maximised, applying appropriate weighting to societal preferences. The closest we get to that in the real world is when someone like the Highways Agency or maybe an independent like the Transport Research Laboratory does a study.
There is a complexity to this though which I think the discussion often misses. People make statements about speed limits which often conflate safety, the setting of speed limits (and the associated cost-benefit arguments) with the discussion about punishment for rule breaking. I firmly believe that in an ideal world all speed limits should be set in the manner described above, by an independent calculative scientific process as outlined above. At that point, all incorrectly set speed limits have gone. They are all set right. This will not magically make everybody happy though, because we are not all perfectly rational machines. Even knowing that those limits are now set correctly we will not always abide by them, because we have made a decision to take on additional risk for a perceived benefit.
So, a side note here on the theory of behavioural safety. Behaviour models are a different thing from the triangle stuff above and it's a whole other kettle of fish. I'll have to ask you to go Google Maslow's Hierarchy of Needs in your spare time (Quite interesting stuff if a bit old now. Incidentally, the updated version has "4G signal" at the top of the triangle... why is it always triangles?) but the one I've always liked is the motivational mapping for safety behaviour which will allow us to further the discussion here.
Consider an action which can be completed in a safe or an unsafe way. How can we affect the behaviour of people to complete this action in the safe manner? So there's a direct link to what speed limits do here.
The behavioural safety models state that when seeking to affect a behaviour you can only really do two things. Reward or Punish. The knowledge of the existence of these actions will then be factored in to how the individual chooses to undertake the action. There is huge variety and it can be anything from money in gifts or fines, or something more nebulous like social approval or disapproval (See Mazlow mentioned above for how these things drive us).
The effectiveness of these actions depends on a number of factors. It has been shown that the most effective are:
- Certain (as opposed to uncertain) - If you know you will always be observed and action taken accordingly, versus infrequent checking, for example.
- Immediate (as opposed to delayed) - Humans are forgetful creatures and if the reactionary effect is delayed we associate it less strongly with the initiating event.
- Negative (as opposed to positive) - Generally, fear of punishment is more effective than desire for reward.
We have our perfectly set speed limits. We have a hypothetical driver who can choose to drive compliantly (which because the limit is set right, is presumed to be safe - Yes, I know limits are an inherently inflexible concept, changing conditions etc.) or non-compliantly. What could we do?
- Immediate, Certain, negative - All cars fitted with GPS speed trackers activating seat mounted electroshock. Amount of people speeding - Just the masochists
- Uncertain, immediate, positive - Cars fitted with a device which has a lottery-level chance to spit out a £50 note for every mile driven safely. Amount of people speeding - Probably a few, and more likely to be the rich. (Expect this in a conservative manifesto soon...)
- Uncertain, delayed, negative - Roving bands of inspectors penalise non-compliance with fines and incarceration. Amount of people speeding - Well, this is what we have now right?
Anyway, where was I going with this. Yes, Speed limit setting and compliance are separate issues, don't conflate the two. It doesn't matter what the limits are set at and whether they are set right or not, some people will still, some of the time, break them. Behavioural modification and appropriate setting can minimise that though and the true goal of safety in road management should be the consideration of both of those aspects in how you achieve the highest overall result. Considering either in isolation will not work properly.
Further credit questions:
Variable limits - We see them in a negative light because they exist currently in a form which only goes downwards. If there was no standard limit and it was continually varied according to condition with no upper limit (good visibility, dry day, low traffic - on you go, 120mph or more) would people be much more prepared to accept them?
Lowest common denominator - Speeds can come up only when the slowest, least able participant is capable. Would we accept the loss of classics as a price for higher limits? What about classed licensing for cars as done with motorbikes?
The No Limits argument - What level of personal responsibility and competence would have to be guaranteed for people to accept the risk of no speed limit. If that's not possible, could it be if we had autonomous cars?
Edited by Ryvita on Thursday 3rd December 15:52
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff