Safe Speed Forums

Observer



Intersting. My take is entirely based on the opposite (almost) premise. Speed cameras are forcing people to slow down their free travelling speed where the 85th percentile (or even lower percentile for that matter) is significantly above the speed limit. People who may previously have travelled at, say, 65 - 70 through an area now camera-infested will be doing the 40 or 50 mph instead. This may well reduce the overall number of KSI accidents on the stretch of road - or it may have no effect. However, whereas before they were all eyes on the road and reducing impact speeds in incidents/accidents by braking, making the majority SIs instead of Ks, there are now a worryingly significant number who, although travelling slower, are doing so by diverting attention to the speedo, and NOT slowing down in the second or so that leads to an accident - greater collision speed = K instead of SI.

Might be true in some cases, but I think the more obedience at the cost of less attention theory is more likely. There's probably some of both going on, but if we think that the cameras are raising concentration levels among the any groups we should, as you said, expect a lower number of crashes in total. Can the BiBs tell us if that matches what they see on the roads? Are you guys encountering fewer numbers of crashes, not regarding type or seriousness. If not then the more-obedience-but-less-attention theory fits better IMO.

_________________
Make everything as simple as possible, but not simpler - Einstein

The bifurcation of the K and the SI chartlines and the change in trend behaviour of RTCs since cameras were introduced is something I've been looking at (in a non expert capacity ), along with the reduction in vulnerable road user fatalities to forward this theory.

It's difficult to assess a trend when actually dealing with collisions as they arrive in a fairly ad hoc way - I dealt with 3 yesterday. I might not get a sniff of another for a week or more. So on a practical level I can't answer your question.

But the national collision stats are quite good for indicating collision trends.

Looking at a couple of charts using the road collision data from the above DfT spreadsheet, the trends could actually reveal something.



This chart shows the number of collisions and the number of collisions per BVKM compared to the figure at 1994 (100%). This tends to show that after 1994 there was still a reduction in collisions per BVKM, and also a numerical reduction despite the increase in vehicle miles.





This chart is perhaps even more indicative of a possible beneficial effect, as many people do not believe that cameras had much effect on peoples psyche until around 1997.

If you look at the collision data relative to 1997 (100%) then it suggests a rise in collisions, (or a drop in reduction per BVKM) up to 1997. After 1997 there was a steady drop again. Is it feasible that this is a beneficial camera effect?

Am I misreading these figures or analysing them incorrectly?

But some food for thought in any case.

_________________
Fixed ideas are like cramp, for instance in the foot, yet the best remedy is to step on them.

Ian

That's nice work Ian, but with 'injury collisions' it's really very hard to know that the data is meaningful. For a start the TRL estimated that injury collisions were under-reported by a factor of 2.6 (from memory) so any change in the level of under-reporting might dramatically influence the figures.

We might even be seeing a loss of trust in the Police showing up as a reduction in the reporting of injury collisions. (And for that, I'd blame the cameras...)

The bottom line is that any conclusion based on injury stats is dangerous... That's why I concentrate on fatalities to measure overall effects. It's the only stat worth trusting.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

That info only covers collsions where someone is hurt, and as such it doesn't do anything to support or dent either theory.
AIUI the idea put forward by IanH and Observer is this (and correct me if I've misread you guys):

• There is a fairly inflexible number of fatals due to nutters, which speed cameras can do little or nothing to affect. This explains why cameras have not done much to improve fatals.
• Beyond that the cameras bring a benefit by "downgrading" the severity of other crashes so that a proportion of SIs become minor injuries etc. But since the fatals are fairly inflexible very few are being "downgraded" to SIs.

The first problem I have is that there is necessarily a presumption that a large proportion of fatals were caused by nutters at the point when cameras were introduced. It's not impossible, but it seems an unlikely coincidence that non-nutter fatals had been reduced so much just at the point cameras came along. The only other explanation that occurs to me if we assume the effect is there is that we have more nutters than we used to who are offsetting the benefit in that area. Can we measure the nutter number in any way?

The next thing is that if we assume the theory is correct that a proportion of SIs should be being reduced to minor injuries, then it implies that a proportion of minor injuries are being reduced to damage only collisions, and presumably a proportion of damage only collisions are being elminated. I've realised that asking the coppers if they see fewer collisions of all types wasn't going to be useful as there's bound to be a number of bumps and scrapes that they never attend with or without cameras - sorry, my fault . But we need to know if there has been an overall reduction in the number of all collisions, including non-injury ones. If the total number of collisions has gone up or even remained the same (accounting for increased mileage of course) I'd say that supports the compliance over concentration theory.

_________________
Make everything as simple as possible, but not simpler - Einstein

Might we perhaps be seeing the effects of an increased propotion of lack-of-attention collisions?
Lack of attention will result in a higher proportion of high collision speeds, which are more likely to be fatal.

Cheers
Peter

Yes I agree with much of that Paul, but I'm trying to find a reasonable explanation for the greater reduction in SIs than Ks over the last few years.
The only reasonable answer I can find is an increase in proportion of the uninfluencable driver, you know the usual list of suspects.
I believe that many of them, drink/drug drivers, boy racers etc., suffer a significantly higher ratio of Ks to SIs than the normal law abiding driver.
Because they drive older less protected cars, they have less ability to defend themselves in the event of a serious RTC due to impairment, they are less likely to have seatbelts on, and are more likely to drive aggressively making a fatal RTC much more likely.
It could be that a 'COAST' driver driving a quality car has a K to SI likelihood of 1:20, whereas the boy racer or drunk in an older car wearing no seatbelt might have a K to SI likelihood of 1:5.
It would be feasible that an increase of 30 or 40% in numbers of uninfluencable drivers over the last 10 years could explain this trend diversification.

_________________
Fixed ideas are like cramp, for instance in the foot, yet the best remedy is to step on them.

Ian

That's certainly what I believe, and it's a good fit for all the facts I know and can verify/trust.

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Okay, but is there any evidence that suggests that sort of increase has actually taken place? And how would we know it's an increase in one type of uninfluencable driver rather than another? We need to be able to tell them apart if we want to get an idea which theory best describes what's going on.
Exactly. That seems simpler, and Occam's razor and all that... It's what I meant by compliance over concentration - that for some motorists compliance with limits has taken priority over concentration on driving safely.

_________________
Make everything as simple as possible, but not simpler - Einstein

I completely agree about the out of reach drivers (although I don't think they're out of reach of good roads policing).

I'm not at all convinced that the reduction in SI is real. I think it's most likely to be a reporting / recording artefact. If it was real, would it really affect all road groups and all road user groups?

http://www.safespeed.org.uk/serious.html
http://www.safespeed.org.uk/serious2.html

And don't foirget we have an increasing number of 'lucky' nutters as well as an increasing number of unlucky nutters - that means more SI from the nutters we both believe in.

I really did try hard to make a spreadsheet model based on an increasing proportion of high risk drivers. I failed to produce anything convincing. How are you getting on?

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Not too well if truth be told.

Although I need to find an afternoon to myself to sit down and think.

_________________
Fixed ideas are like cramp, for instance in the foot, yet the best remedy is to step on them.

Ian