I have a problem with speed cameras. After registering the Goatmobile several times and never having had to fork out for traffic violations, I've just discovered three speeding fines, all courtesy of the anonymous grey boxes. Each time, I was only just above the posted limit, each was on a dead-straight dual carriageway and on all occasions there was virtually no other traffic on the road. On one occasion, the posted limit was 120kph and I was doing 100kph. But the camera says otherwise, and cameras never lie. Ask anyone who can use Photoshop. As they say in 1950s cop shows, “It's a fair cop, Guv. You got me bang to rights.”
I say ‘virtually’ no other traffic. Twice, after the camera flashed, I was overtaken by a vehicle travelling at near the speed of sound, and on one of those occasions the high-speed overtake was actually an undertake on the breakdown lane. Naturally this didn’t get caught by the camera. It might have been caught by an officer of the law had there been one present. But luckily, the majority of traffic enforcement can nowadays be dealt with by machines: the camera and the ATM. The policeman’s job has been reduced to sitting in an air-conditioned office, taking the money and rubber-stamping the receipts.
The pro lobby obsesses about ‘safety cameras’, citing evidence that shows that in the illustrative village of Netherby Poppleworth there were five fatal accidents in the three years preceding installation of Mr Gatso's device, and only two in the three years following.
Yet there were no fatal accidents at all in the five years before that.
‘Regression to mean’ is a known statistical phenomenon that, in a nutshell, says that extreme results are unlikely to be repeated. In the case of traffic accidents, a sudden rash of them (that causes uproar from the locals and demands that the authorities “do something”) is probably an isolated peak value. Once you hit the peak there’s only one way to go. When regression to mean is included in the accident reduction calculations, the resulting rate reduction is nowhere near as impressive as the headline figure.
According to an old copy of Autocar magazine, the British killed a horrifying 8000 people on the roads in 1963. Introduction of anti drinking and driving laws caused a substantial drop in death rates. Similar drops occurred following compulsory fitting of seat belts, then compulsory wearing of seat belts, and improvements to roads and cars have contributed to progressively lower and lower fatality rates. For many years, the British have killed around 3500 people a year in road accidents, despite increases in vehicle numbers and population. This appalling figure, albeit one of the lowest accident rates in the world measured by miles driven, vehicles numbers or population, has remained largely unchanged since the introduction of speed cameras. Yet prosecutions, which in the UK includes points on the licence, at the hands of Mr Gatso continue to increase exponentially.
“What do I do with all these points?”
“You collect them. If you collect 12 in a three-year period, you get a bicycle.”
“But the number of accidents in Netherby Poppleworth has decreased!” Of course! Drivers slow down for the village and then speed up again afterwards. Maybe a driver will take additional risks because he was delayed in the village and will have his accident elsewhere. I can only speculate as to the number of accidents that occur as a direct result of a speed camera. Mr Mercedes gets flashed and then stamps on the brakes at 150kph in the overtaking lane. Why? Can he brake faster than the speed of light or something? He then gets rear-ended by Mr Volvo who wasn’t really expecting the car in front suddenly to stand on its front wheels.
Speeding through built-up areas is daft. Rumble strips and prominent signage is as effective as a dayglo box on a stick to get drivers to slow down. At least it for me. It’s a darned sight more effective than a camouflaged grey box or an anonymous vehicle hidden behind a bush. But rumble strips don't extract money, do they? I would be interested to learn what the effect on speeds, fines and accidents was during the GITEX exhibition in Dubai a couple of years ago when Hewlett Packard's advertising involved giving each camera on Sheikh Zayed Road a fluorescent orange overcoat.
The supreme irony of my having to pay a pile of speeding fines this year is that I’m actually not a speed-crazed nutter (any more). Being older and wiser has slowed me down from my erstwhile days aboard a 1000cc Kawasaki. And it’s the effect of age and change of vehicle that have slowed me down, not the unwelcome plague of cash registers that have appeared on the roads over recent years.
I shall start to be more in favour of enforcement cameras when I learn of one that can detect bald tyres, lack of insurance, and drunken driving. The Mk I Police Constable is actually rather good at doing this. Detecting offences, that is; not drunken driving.
Traffic accident figures for Dubai exist, and are published on the Dubai Police. Records of numbers of registered vehicles, population and accident rates and types are available for 1993 to 2002. For the same figures up to 2005, Dubai Road and Transport Authority (the RTA, in a fit of irony) has published a booklet. Several things are demonstrable from the published statistics. At the risk of Lying With Said Statistics:-
- Dubai tends to kill around 20 people annually per hundred thousand population;
- The rate of fines per vehicle was between 2.0 and 2.5 per year up to 1999. It then dropped to 1.5 but has been gradually increasing since then;
- The number of fines per unit population was steady at 0.5 to 0.8, but since 1999 has been creeping up;
- Although fatal accidents per number of vehicles has halved, this is offset by a doubling in car ownership rates.
Conclusion: However it is measured, increases in the quantity of traffic violations detected and presumably punished does not bring down the accident death rate. And certainly as far as speeding is concerned, which constitutes around 60% of all reported traffic offences, these are detected almost entirely by camera.
Smart cameras exist (not in the UAE, for which I am grateful) that can read number plates. In principle these seem like a good idea: stolen, unregistered, tax-expired vehicles and fake plates can be readily identified and a police patrol scrambled to apprehend the occupants of the suspicious vehicle.
The system can prove where your car was at any time, and can therefore be used to prove its average speed on a particular section of motorway. This is only a small step away from Big Brother asking why your car was parked outside the Unicycle Nightclub last night, why your car was parked overnight outside a particular apartment, and what it was doing near the Barracuda hotel and beach resort. I’ll consider issues relating to the “If you've done no wrong you have nothing to fear” platitude in a different post.
