Speed cameras: Are we interested in evidence?
Discussion
I notice that no-one has asked why speed cameras lead to more deaths, it is just assumed that drivers brake and lose control. This has happened, but won't be the only reason.
As an engineer, I want to know when, who and why?
Are the deaths occurring when the camera van is on site, or when it's not?
I did ask the speed camera partnership, but they had no information on that.
They said, if a collision occurs when they are enforcing, they do not log that.
Who is being killed?
Is it drivers, pedestrians, motorcyclists, other?
and why?
Check the collision reports to see how contributory factors have changed.
Here's a potential senario.
A 30mph road has a school.
The road is busy at school time and:
1) drivers are adjusting their speed for conditions (and are therefore slow)
2) this is also when collisions occur.
The road is quiet at other times and:
1) drivers are adjusting their speed for conditions (and are therefore faster)
2) collisions rarely occur (few other road users to collide with).
So most drivers vary between 9mph and 41mph.
The speed camera van starts operating, but only when speeds are high (ie outside of school times).
Drivers start diverting a bit of their attention to staying within the speed limit.
They are therefore adjusting their speed slightly less for conditions and more to the speed limit.
They do this at all times because there's a possibility they will be prosecuted.
So now most drivers vary between 13mph and 29mph.
So the speed camera van could lead to more collisions, even when it's not on site.
This could be tested.
What is the change in variance of speeds after speed cameras start operating?
What we mustn't do is jump to conclusions,
so, after finding out what happened, an evidence led approach then asks when, who and why?
courty said:
OP, seems as if you have good evidence that speed cameras should be hidden/not signed, unmarked vans etc.?
Yes providing "good evidence" did take me a lot of time and effort, so thank you.We still don't know why the cameras resulted in more deaths and injuries, though.
Your proposal might reduce some of the problem or it might make things even worse.
Before embarking on another potentially dangerous policy, we need to first ensure that we know exactly what is happening:
and that requires scientific trials.
Video: https://www.youtube.com/watch?v=-GqOm-keyss
Edited by Dave Finney on Monday 27th November 19:23
rscott said:
Dave Finney said:
rscott said:
vonhosen said:
Dave Finney said:
vonhosen said:
Dave Finney said:
vonhosen said:
Dave Finney said:
vonhosen said:
If the allegation is that an unintended consequence of the visible camera themselves, is an adverse effect on driver behaviour/safety ...
It's not an "allegation", it's evidence.And the evidence suggests that speed cameras are causing an increase in death and serious injuries.
Have you seen the video, or read the report?
No spike where the camera is then.
Be careful what you wish for.
And perhaps we can also agree that that means we need to run scientific trials to prove what effect they're having.
Only then should we decide what policies to pursue next, along with the evidence required to prove them.
It's good to see agreement every now and again
.The fatal & serious collisions have increased in "particular isolated areas" compared to elsewhere.
Those "particular isolated areas" are called "speed camera sites".
The speed cameras started operating, and fatal & serious collisions rose above the previous mean rate.
My report also checks speed camera sites, against all the other roads.
There is a lot more detail in my report than I could fit into my video.
My report: https://speedcamerareport.co.uk/08_mobile/
Or is it only at a few camera sites?
1. What happened to accident rates across the entire TVP area ( not just camera sites). If the overall rate increased by more than that of camera sites, then it suggests cameras are working.
2. How did traffic levels through the camera sites change?
Accident rates should be as percentage of vehicles through the sites to be meaningful comparisons. A road near me has seen a doubling of accidents over the last 4 years, but there have also been 3 new housing developments of about 750 houses built, which has trebled traffic volumes, so the number of accidents per journey has fallen.
and good questions:
1. Yes, my full report compares the accident rates at the camera sites to the accident rates across the entire TVP area.
2. Don't know. I did ask but the speed camera partnership said they never took those speed readings.
You're right, there are a host of factors influencing accident rates across the entire TVP area,
such as traffic volumes, cost of fuel, weather, Police reporting practices, choice of transport mode etc etc.
And these factors should influence accident rates at the camera sites in proportion.
My video doesn't mention "trend" (as they call it), but it is compensated for in my full report.
I use the same concept as official reports,
but I developed a new more accurate method.
My new method has since been adopted and recommended by official road safety researchers.
See graphs 8.2, 8.3 and 8.5 at the end of this page (the 3 graphs that show "Proportions"):
https://speedcamerareport.co.uk/08_mobile_report/
Both of my new methods (for trend and site selection effects (RTM)) are here:
https://speedcamerareport.co.uk/rtm-regression-to-...
Have you tried contacting the Highways department to see if they have any speed and volume surveys for the areas in question?
You demand the highest standards of evidence.
You reject as "meaningless" all official reports because they don't remove the effect of site selection or compensate for trend (let alone not accounting for traffic volumes).
Whilst my report presents the highest standard of evidence possible given the data available, you reject this also for no traffic volumes.
You say every single road safety report is "meaningless".
Excellent.
We both agree that current levels of evidence are simply not acceptable.
We both demand the highest standards of evidence.
We both therefore insist that speed cameras must only be operated within scientific trials.
bearman68 said:
I've read quite a bit of this thread. I'm an engineer. I cannot be anything other than supportive of the figures DF has put forward, and would endorse the view that if you want something to change, you first have to measure it properly.
It's called the scientific principle. I suspect that's where the problem starts, because it appears many of our ruling class have no idea about science. It's something that seems ever more apparent.
I hereby nominate you for best post of my thread!It's called the scientific principle. I suspect that's where the problem starts, because it appears many of our ruling class have no idea about science. It's something that seems ever more apparent.
Thank you.
Video: https://www.youtube.com/watch?v=-GqOm-keyss
heebeegeetee said:
bigothunter said:
Dave Finney said:
bearman68 said:
I've read quite a bit of this thread. I'm an engineer. I cannot be anything other than supportive of the figures DF has put forward, and would endorse the view that if you want something to change, you first have to measure it properly.
It's called the scientific principle. I suspect that's where the problem starts, because it appears many of our ruling class have no idea about science. It's something that seems ever more apparent.
I hereby nominate you for best post of my thread!It's called the scientific principle. I suspect that's where the problem starts, because it appears many of our ruling class have no idea about science. It's something that seems ever more apparent.
But why should a Latin scholar be concerned with the tedium of scientific principle? In his world, ideology is the driving force not evidence or pragmatism.
This mindset is defining and controlling (ie cameras) our speed limits.
https://www.walesonline.co.uk/news/politics/discon...
Politics before science...
I find it impossible to believe that the presence of a speed camera/van can cause a crash so severe that it results in fatal or serious injury. I'm basing this on my own experience, which i believe is reasonably extensive. I'm no angel and have been done for speeding, but i simply can not imagine how speed enforcement could cause me to crash. If people are braking excessively at sight of a camera, are you going to measure the effect of the camera or the bad driving, as i think only bad driving could cause a crash?
The UK is a nation where every day billions of miles are travelled on roads daily, by some 40+ million people each day.
I'm all for measuring whatever you want, but for how long, at what cost, and what precisely are we measuring?
Or have i just got this all wrong?
it - speed cameras, visibility barriers, switching off street lights, changing the speed limit etc
measure? - their effect on injury rates.
Every safety device has positive benefits and negative side effects,
so surely you agree that they need to be tested to ensure that they prevent more injuries than they cause?
And that this should be required by law?
eg, a smoke alarm must:
1) reliably detect and warn of fire.
2) not itself cause a fire (electrical faults are a common cause of fire).
3) not leak, explode, radiate nuclear particles, etc.
4) not lead to risky behavour that results in more deaths than are prevented.
That 4th one is interesting.
Everything that makes us feel safer, reduces our attention to that danger.
eg, suppose that people with smoke alarms tend to light candles and smoke in bed, whereas those that don't don't?
Then suppose that this resulted in smoke alarms leading to more deaths than were prevented.
Would you say: "it's their own fault, let them burn?"
Or would you say: "we need to find out exactly what is happening, and fix it."
With road safety, fatalities occur due to a wide range of factors and mostly when drivers are not speeding.
It is not possibly to change peoples speed without changing other factors,
and increasing fatalities at speed camera sites will not just be people braking and losing control.
I have a page on this on my website:
https://speedcamerareport.co.uk/effects-of-speed-c...
Video: https://www.youtube.com/watch?v=-GqOm-keyss
heebeegeetee said:
I find it impossible to believe that the presence of a speed camera/van can cause a crash so severe that it results in fatal or serious injury. I'm basing this on my own experience, which i believe is reasonably extensive. I'm no angel and have been done for speeding, but i simply can not imagine how speed enforcement could cause me to crash. If people are braking excessively at sight of a camera, are you going to measure the effect of the camera or the bad driving, as i think only bad driving could cause a crash?
The UK is a nation where every day billions of miles are travelled on roads daily, by some 40+ million people each day.
Sorry, just realised what you meant and it's an important point.The UK is a nation where every day billions of miles are travelled on roads daily, by some 40+ million people each day.
You're saying: It seems impossible for speed cameras to cause a serious crash (other than people braking and losing control).
Assuming we know why collisions occur, without the proper evidence, may be the starting point for many of the mistakes made in road safety.
Take a simple example.
A driver is not speeding but doesn't know it.
Because he is below the limit, his car is not very loud.
He sees a camera ahead so looks down to check his speed.
A pedestrian is looking at his phone.
The road is quiet and he walks into the road without looking properly.
The chances of both those happening in the same place at exactly the same time is remote, billions to one.
But there are billions of oppertunities for that to happen every day.
Speed cameras increase the chance of that combination occurring.
Plus there are other senarios.
Also, the vast majority of even fatal collisions occur when drivers are not speeding.
So we are looking at very unlikely combinations of events (a tiny number), but multiplied by a huge number of potential conflict points.
So the positive benefit of speed cameras may be much lower than we would naturally assume,
and the negative side effects could be much larger.
heebeegeetee said:
I'm sorry, it's but a moment to check a speed, in no way can this be blamed for failing to see a pedestrian.
It's not about blame, it's about the result. Are more people dead?I've given 2 other senarios, but there are more:
https://speedcamerareport.co.uk/effects-of-speed-c...
We can't fix any problem until we recoginise that the problem exists.
Hence the need for scientific trials.
Why are you so opposed to an evidence led approach?
bearman68 said:
The original post was that the data related to accidents, and the claims that speed cameras prevent accidents is at best misleading, and at worst downright fraudulent.
WHY accidents appear to go up in speed camera zones is a matter of further research. Understanding the real reasons is what leads to road safety improvements.
Unresearched opinion on driving standards, the presence or otherwise of trees, street furniture, automation and the like is interesting, (to a point), but not the subject of the thread. (By all means chat about it, but it is superfluous waffle in an interesting stats based discussion).
If we are genuinely serious about improving road safety, rather than just reducing road speeds, we should be looking at evidence based approaches. It might well find that increased automation actually causes safety concerns rather than address them. Human behaviour is a funny old thing, and the law on unintended consequences is strong.
Absolutely agree.WHY accidents appear to go up in speed camera zones is a matter of further research. Understanding the real reasons is what leads to road safety improvements.
Unresearched opinion on driving standards, the presence or otherwise of trees, street furniture, automation and the like is interesting, (to a point), but not the subject of the thread. (By all means chat about it, but it is superfluous waffle in an interesting stats based discussion).
If we are genuinely serious about improving road safety, rather than just reducing road speeds, we should be looking at evidence based approaches. It might well find that increased automation actually causes safety concerns rather than address them. Human behaviour is a funny old thing, and the law on unintended consequences is strong.
It's surprising that, even now in the 21st century, there are many who still reject an evidence led approach.
In a democracy, therefore, it should perhaps not surprise us that the authorities can get away with failing to do the research that needs to be done.
heebeegeetee said:
...difficulty in envisaging someone crashing heavily simply because they've seen a speed camera...
You're right, accidents generally (not just in road safety) don't occur due to a single "cause", it's a combination of several "contributory factors". If any one of the factors had not happened, neither would the accident. In fact some of the factors may occur time and again without an accident because other factors were not present.This is probably true of the largest factor (by a long way) in serious crashes: "Failed to look properly".
I've given two scenarios already for how the speed cameras might have caused serious crashes, without anyone speeding, so try looking at it from the other side.
The vast majority of even the most serious crashes (fatalities) occur when drivers are not speeding so it is likely that, at many sites, speeding was not a factor in any of the fatalities in the 3 years before the speed cameras started operating.
So whatever was causing the fatalities will still be there and they will continue.
Now you've got huge numbers (millions?) of vehicles passing that would not have crashed but the speed cameras have changed these drivers behaviour (as indeed is the stated aim of the speed camera operators).
If you change the behaviour of a group that will have zero crashes, what happens?
It can only go one way.
And that is what the data is telling us did happen.
Check out speeding: https://speedcamerareport.co.uk/speeding/
I have the speed camera partnerships response to my report, along with my answers.
It's quite technical, but does show how detailed the evidence has to be, especially when official policy is challenged.
Here: https://speedcamerareport.files.wordpress.com/2023...
It's quite technical, but does show how detailed the evidence has to be, especially when official policy is challenged.
Here: https://speedcamerareport.files.wordpress.com/2023...
heebeegeetee said:
1. You did, but didn't explain how they can cause crashes. A driver maintaining his speed does not cause crashes; or to put it another way, any driver who can not maintain a speed without crashing can not be allowed to drive.
2. How do you know a group will have zero crashes? Who is monitoring that?
1. No. In terms of speed, it's a driver that adopts a safe speed for the conditions, adjusting where needed, that "does not cause crashes".2. How do you know a group will have zero crashes? Who is monitoring that?
And, by not allowing adjustment for conditions some of the time, speed cameras do interfere in that process.
2. I'm not doing anything radical here, simply trying to explain basic safety engineering.
We know that serious collisions increased after speed cameras started operating, but we don't know why.
We could look at the evidence but that requires the collision investigation reports.
Only the Police, the speed camera partnership, the councils or the DfT can do that, but they haven't yet.
So we can only really consider what might have occurred, with reference to the known contributory factors to collisions.
Imagine a site where collisions are going to occur.
Speed cameras start operating at the site.
We can divide the instances of people driving through the site into 3 groups.
1) Those that will be speeding and involved in a serious crash.
2) Those that will not be speeding but will be involved in a serious crash.
3) Those that will not be involved in a serious crash.
Group 1 may be very small (maybe 1 every 3 years).
Speed cameras will be able to influence this group so some of their crashes are avoided.
Group 2 tends to be larger (maybe 3 every 3 years).
This group will crash anyway, despite the speed cameras.
The 3rd group is massive (maybe millions every 3 years).
The problem is that the speed cameras may change the behaviour of most (or all?) of this group ever so slightly.
If you change the behaviour of this 3rd group that will have zero serious crashes, what happens?
Serious crashes can only go one way.
And that is what the data is telling us did happen.
The overall result of whatever is happening, is an increase in fatal and serious collisions:
Video: https://www.youtube.com/watch?v=-GqOm-keyss
heebeegeetee said:
1. How do speed cameras interfere in that process? (of adopting a safe speed for the conditions)
I expect to have speed and law enforcement, even as a driver who likes to speed.
If you like to exceed speed limits, but speed enforcement is preventing you, then either:I expect to have speed and law enforcement, even as a driver who likes to speed.
A: you are capable of adjusting your speed for the conditions but, due to the possibility of speed enforcement, decide to stop doing that sometimes and instead just drive at the speed limit.
In this case you are switching between driving "safely for the conditions", and "obeying the speed limit".
So the speed cameras are interfering in your process of selecting a safe and appropriate speed, even if you think the effect is only minimal.
B: you are not capable of adjusting your speed for the conditions to ensure that you and everyone else is safe.
In which case, you are a dangerous driver who is likely to crash somewhere, and perhaps one of those who are increasing the deaths and serious injuries at speed camera sites.
heebeegeetee said:
2. I would take a simpler view and break it into 2 groups: those that have serious crashes and those that dont.
You are right there are those 2 basic groups, but more groups are created when enforcement devices are introduced.Just as breathalysers create 2 groups: "drunk" and "sober",
speed cameras create 2 groups: "speeding" and "not speeding".
You can't just ignore that.
Edited by Dave Finney on Friday 1st December 14:33
Hey people, can we please try to be constructive?
Rather than argue, I suggest we need more evidence and more testing.
First we need to know who was injured, when and why at the speed camera sites.
We can't do that, only the road safety professionals can.
Once we have that evidence, we can then propose how to fix the problem (and there's more options than just hide the cameras or stop using them).
But what is MOST important, is that whatever changes are made, they MUST be run within scientific trials.
We can't allow another policy to potentially create an even greater problem than we already have!
Is it even possible to persuade the authorities to start using an evidence led approach?
Rather than argue, I suggest we need more evidence and more testing.
First we need to know who was injured, when and why at the speed camera sites.
We can't do that, only the road safety professionals can.
Once we have that evidence, we can then propose how to fix the problem (and there's more options than just hide the cameras or stop using them).
But what is MOST important, is that whatever changes are made, they MUST be run within scientific trials.
We can't allow another policy to potentially create an even greater problem than we already have!
Is it even possible to persuade the authorities to start using an evidence led approach?
768 said:
In theory, yes, of course. It could happen tomorrow. And it should.
The problem is their road safety model.All crashes involve speed, whether it's bumping the gatepost at 2mph, or causing death at 100mph.
So they created a model that says: "if speed is reduced by 1mph, crashes reduce by 5%", and they've made all their policies and statistics fit this model.
Their model is simple, plausible, beautiful.
It allows them to do anything they want in road safety, with no accountability.
If their intervention reduced speeds, then the model says collisions must have reduced and lives must have been saved.
The issue, though, is that when their own data is analysed properly,
we find that speed cameras do reduce speeds,
but they increase fatal and serious collisions.
This doesn't just dent their model, it seriously damages it, perhaps fatally.
And scientific trials?
These would expose the effect of their interventions and the road safety officials would become accountable for their actions.
This takes them from the safety of their beautiful model, into a scary new world (the real world).
Make no mistake, scientific trials would be a root and branch reform of road safety,
but if they don't do them, more people will be dead.
If they do run scientific trials, lives WILL be saved,
and that life could be yours, whether you support current policies or not.
Video: https://www.youtube.com/watch?v=-GqOm-keyss
heebeegeetee said:
Dave Finney said:
And scientific trials?
These would expose the effect of their interventions and the road safety officials would become accountable for their actions.
This takes them from the safety of their beautiful model, into a scary new world (the real world).
That took me by surprise somewhat, then on reflection I realised we're already in your "real world" Dave, where fault can be passed onto others, when people make mistakes there must be someone else to blame.These would expose the effect of their interventions and the road safety officials would become accountable for their actions.
This takes them from the safety of their beautiful model, into a scary new world (the real world).
Those in charge of improving that system state that their aim is: "vision zero", ie no deaths at all.
Before they can even start to achieve that, they at the very least need to ensure that they don't make the system worse.
NFT said:
Hey OP,
While this is going a little off topic, is there statistics for the majority of people caught being on or close to threshold's, and the number of multi-offence (e.g also no seatbelt, using phone, not in proper control of vehicle etc..) and serious speeds is a low percent?
Wow, a rare post about evidence, THANK YOU !!!While this is going a little off topic, is there statistics for the majority of people caught being on or close to threshold's, and the number of multi-offence (e.g also no seatbelt, using phone, not in proper control of vehicle etc..) and serious speeds is a low percent?
I haven't really researched what speeds people got NIPs for, or what other offences and the combinations of offences were obtained by camera vans.
The number of drivers with speed camera points must be available on line though, along with NIP rates for no seatbelt and using phone etc.
I do have evidence of the "photo to penalty" rate, and it's around 60%.
The 40% that avoid speeding penalties might include:
Drivers of stolen cars, foreign drivers, foreign registered vehicles, criminals, wealthy people with specialist lawyers, incorrectly registered vehicles, etc.
What is interesting is that this 40% might be the very drivers they actually need to prosecute (as they may be both more likely to speed and more likely to crash).
So a possible reason why fatal and serious collisions have not reduced at speed camera sites is that they are failing to prosecute the drivers that are actually causing the crashes.
In that case, hiding the cameras would not help.
This is why we need to know who, when and why people are being killed at the camera sites,
and need scientific trials to test any new policy that is tried.
Page with speeding offence data: https://speedcamerareport.co.uk/effects-of-speed-c...
Spreadsheet here: https://speedcamerareport.files.wordpress.com/2023...
Cat said:
Dave Finney said:
The number of drivers with speed camera points must be available on line though, along with NIP rates for no seatbelt and using phone etc.
It is difficult to take your research seriously when it appears that you don't understand what a NIP is. They are not applicable to the offences of not wearing a seatbelt and driving whilst using a mobile phone. Cat
Cat said:
What do they call what for seatbelt or mobile phones? Certain offences require a warning in terms of section 1 of the RTOA. A NIP is one way of giving that warning. Seatbelt and mobile phone offences don't require a s1 warning so there will be no "NIP rates" in relation to them.
Cat
That's fine, if not NIP rates, what would you like to call them then?Cat
Video: https://www.youtube.com/watch?v=-GqOm-keyss
Cat said:
A term that actually means something in relation to those offences? There are no NIPs for seatbelt and mobile phone offences so referring to NIP rates for them is nonsensical.
What do you suggest?Speed camera van man says:
"Officers are there to make sure you are wearing a seatbelt and are not using your mobile phones behind the wheel. Anyone caught breaking these laws will be prosecuted."
Surely, if there is an intention to prosecute, the drivers need to be sent a notice to tell them?
And we could call that a NIP?
Here: https://www.cambridge-news.co.uk/news/uk-world-new...
Video: https://www.youtube.com/watch?v=-GqOm-keyss
Cat said:
Dave Finney said:
Surely, if there is an intention to prosecute, the drivers need to be sent a notice to tell them?
Only for certain offences. Seatbelt and mobile phone offences do not require a NIP. https://www.rac.co.uk/drive/advice/legal/fixed-pen...
and NIPs:
https://www.roadtrafficlaw.com/guides/notice-of-in...
So it's FPN rates, not NIP rates for seat belt and mobile phone. We have both learnt something!
Video: https://www.youtube.com/watch?v=-GqOm-keyss
Gassing Station | Speed, Plod & the Law | Top of Page | What's New | My Stuff