@ david wilson
“Even if a driver driving at the 85th percentile speed on an unregulated road has the least accidents, that can easily be substantially due to their relative position in the speed distribution, rather than their absolute speed.”
That was an arbitrary thing to say, seems like you’re now clutching at straws. It of course far more likely they have fewer accidents because they are better drivers – afterall, they must be great if they can go faster than average whilst having fewer accidents than average.
“They're less likely to be hit from behind than a slow driver. They don't have the danger of being the fastest vehicle on the road. They're probably not in a slow or awkward vehicle. They're probably not a desperately nervous driver, etc”
They’re more likely to be cut up by a slow driver. They don’t have the danger of being the slowest vehicle on the road. Slow drivers probably aren’t in slow or awkward vehicles (practically all unlimited vehicles can exceed speed limits).
What I did there was to give inverted claims; this proves the factors you gave are logically redundant. The last one would support my point anyway.
“…likely slower than before, and one would hope that those speeds are generally unlikely to attract a ticket.”
“Hope”? Or, like I already explained, a reasonably set limit would also result with less tickets being issued (especially to those with the lowest accident risk), as well as reducing the amount of deviant driving.
“Someone who felt impelled to carry on at the speed they would do if there wasn't a limit even when other drivers slowed down *and* who wasn't capable of noticing cameras seems a fairly unlikely candidate for 'best driver on the road'.”
I think those least likely to have accidents are those most able to spot cameras due to their increased general awareness (well they must be more aware if they have less accidents even though they go faster – right?) Regardless, it doesn’t take away from my point: use of the least risk drivers is a damn good guide for setting reasonable speed limits.
“In fact, if you're really looking at *public perception*, if someone announced the accident rate at fixed camera sites fell by 45% (from H7), and didn't make any clarification, I think most people would think of that as 'compared to if nothing had been done'. For that person, the 23.5% figure (from H8) would best fit what they were thinking.”
Jeepers creepers! Are you being serious? You are the only person I know who has interpreted such claims in such a nonsensical manner (and I’ve argued this with a lot of people). For you to make such an interpretation would require the assumption that all such reduction claims are factoring RTTM, and always were even before any RTTM study had been done, let alone quantified; that is of course just silly!
Here are a few examples of the claims, taken from UK speed camera web sites:
“49% reduction in the number of people killed or seriously injured (KSI casualties) at camera sites compared to the baseline figure” [www.stratford-dc.gov.uk]
“the collision statistics for roads across the Wiltshire and Swindon area, including camera sites indicated a 30.25% reduction in the number of people killed or seriously injured when compared with the baseline data [required by the Department for Transport before Safety Camera enforcement began]. [safetycameraswiltshire.co.uk]
“Figures were produced at the presentation to Scrutiny to show that a reduction in casualties at camera sites has occurred, with the mobile deployment in Rutland showing a 77% reduction in killed and seriously injured [and a 36% reduction in personal injury casualties overall] when compared to the baseline.” [rutland.gov.uk]
“(3 Year before site establishment compared to most recent 3 Year period)” [cmis.derby.gov.uk]
“There were 50% fewer collisions in which people were killed or seriously injured at Warwickshire camera sites in 2003 compared to the annual average 1994-98.”
“63% reduction in the number of people killed or seriously injured (KSI casualties) at camera sites compared to the baseline figure.” [warwickshire.gov.uk]
“The report shows that on average at the camera sites in the pilots the number of people killed or seriously injured fell by 47%, compared to the average over the previous three years.” [dft.gov.uk]
The ‘baseline figure’ is the cluster of accidents which allowed the cameras to be placed (as per the placement policy).
These examples are typical and the list is far from exhaustive. These are as clear as day; I’ve not selectively quoted convenient parts of the text. How about you list such claims from UK government/partnership websites clarifying as you would interpret, then we could then compare how many we can dig up so we can prove this point. It is obvious that only table H7 gives the data in the necessary context; to use table of H8 would be to compare apples and oranges.
Regardless of how you decide interpret it, this is the way it is. What was it about what you said about people grabbing hold of the highest figure they see: “It's certainly true that some people pick up on and promote the larger figures, though how much of that is down to people bending the truth for their own ends, “. The words Hoist and Petard come to mind.
To summarise table H7 (for everyone who hasn’t read the document):
For the sites used in the RTTM analysis, there was an observed average of a 54.5% drop of KSIs at the camera sites. The analysis showed that the portions of this drop is split (in absolute terms, the sum totalling 54.5%) as follows:
- Scheme effect: 10.4%
- RTM effect: 34.8%
- Trend effect: 9.3
Points to note:
1) The RTTM effect is obviously not the camera.
2) The trend effect is derived from the overall national reduction of casualties, this resulting from all safety initiatives, such as better vehicle safety, post crash care, road engineering etc. The few cameras available at that time of analysis play no significant part in this factor (to argue otherwise will only confirm that speed cameras indeed do kill).
3) The scheme effect is the camera (obviously), but it is also any other genuine safety measures which happens to be within the camera site such as: a new pedestrian crossing/barrier, cycle lane, redesign of that section of road etc. Urban camera sites can be up to 5km in length, so you would expect a few other measures in there. Yeah, it is good to have combinations of safety features, but would all of the KSI drop at eh site be wholly thanks to only the camera? If you don’t think so then you understand what I call ‘bias on selection’.
Now, with the figures given, do you care to tell us how much of the drop of KSI is actually due to the camera? From that can you now answer the question:
Do you agree that the claims of speed camera effectiveness is greatly exaggerated? (by at least a factor of 5)[let’s not forget that even the report with its RTTM analysis does not factor in the effect of ‘bias on selection’, itself greatly significant; we can discuss this after we conclude RTTM if you wish]
“It's perfectly possible for someone to read the whole report and conclude that cameras are worth having, and presumably at least some people at the camera partnerships will have done that. For someone not set against cameras on principle, the report doesn't seem particularly likely to change their mind.”
Of course! However, an author can guide the reader to a desired conclusion if only partial information is given and other critical factors are withheld and aren’t acknowledged, such as in this very report. If instead the report detailed the potentially huge effect of ‘bias on selection’, the shift away from traf-pol (even though we drive more), and the loss of national trend since the rollout of the speed camera policy, then I think most people ‘not set against cameras on principle’ are very likely to change their mind - don’t you think?
I won't be surprised if you don't even try to reply before this forum closes.