21 Sep 2012 Added Eric Bridgstock’s Critique
Reports by the authorities and Partnerships are seriously flawed and their claims of casualty reduction are not remotely compatible with accident causation data.
In September 2006 the DfT published new and more detailed accident causation findings (C.22) showing that 4% of slight, 5% of all, 7% of SI, 8% of KSI and 12% of fatal accidents involve speeds above the speed limit even as one of the (usually many) causal factors. “Going to fast for the conditions” (but implicitly within the limit scored substantially higher while “Driver error” of various sorts showed higher still at 66%, 67%, 60% and 64%. Far higher – but immune to speed cameras. (later reports have shown slightly higher numbers, but this does not affect the points I make here.)
These figures should have nailed once and for all the lie we have been told for years in the attempt to justify speed cameras – “the one-third lie” that “one third of fatalities are due to “speed” – a word routinely used to imply “speeding” without actually saying it. Yet report after report, Partnership acounts year after year, routinely claimed casualty reductions due to cameras far in excess of these figures - even when speeding was far from eliminated and when, as the DfT accepts as it logically must, even eliminating all speeds above the speed limit would not eliminate all casualties in which speeding is a factor, because the many other factors would remain. In any case as cameras cover no more than the odd % of our road network the benefit in national terms could only ever be minimal, yet for years road safety policy has been heavily concentrated on speeding, a relatively minor causal factor.
Of many examples of errors or deliberate misrepresentation over many years, and continuing to this day, the single most important has been at first to ignore completely and later minimise the contribution of Regression to the Mean (see next section). In laymans’ terms they claimed credit for accident reductions which would have happened anyway, without the cameras – and indeed routinely did across the country where there were no cameras in any case – see “Analysis of 6m Accidents”.
The first speed cameras were installed in 1992 and grew only modestly in numbers without much detailed analysis as far as I am aware, until around 2000. Then, on the basis of the Eight Area Trial, planned for 2 years but acclaimed after only 1 year to be such a success that it be applied across the country as the Hypothecation Scheme, numbers grew very rapidly to reach some 6,000 by 2006.
However the results of the Eight Area Trial were very seriously flawed, for these reasons:
(TO BE CONTINUED, INCLUDING THE 2nd and 3rd, Reports and the updated RACF Report of 2013 that halved the 2010 estimate)