SafeSpeed is a UK based group which styles itself as a road safety organisation. It campaigns against the current UK policy of enforcing speed limits as a means of improving road safety, especially by means of using speed cameras. Safespeed's case is: "you can't measure safe driving in miles per hour"

SafeSpeed is run by Paul Smith, who ran the project as a hobby from his home in Scotland for some time. In 2003, following a period of illness in which his self-employed computer engineering business ran down, he took it up full time at some significant personal cost. The group has a vocal following, recently raising £15,000 from donations to ward off closure due to debts incurred in Smith's early months of illness.

Although Smith has always used the word "we" when discussing SafeSpeed on the website, it is very much his idea and his organisation, and SafeSpeed's high profile is a tribute to Smith's persistence. Smith is also a member of the motorists' pressure group Association of British Drivers (ABD).

The group has links with libertarian groups. One prominent supporter is Dr. Alan Buckingham, a specialist in family life and relationships at Bath Spa University College and a contributor to The Centre for Independent Studies.

Recently the accuracy of the cameras in use has been called into question by ITV, BBC and the Daily Mail, after conducting tests. SafeSpeed have claimed that there are accuracy issues for some time.

Paul Smith also has outspoken views on issues such as disabled parking.

Several motorists’ groups claim that there is no observable correlation between camera use and speeding fines, and improvements in road safety. SafeSpeed go further, claiming that use of cameras actively reduces safety. Key claims include:

SafeSpeed’s primary contention is that since the advent of speed cameras, the message “speed kills” has become the dominant road safety message, but that this is misplaced. They contend that speeding is a minor problem at worst, and that setting a safe speed for the conditions is more important. There is a clear implication that this speed will often be above the speed limit. Smith has summarised this as “speed kills, kills”.

SafeSpeed makes a number of claims in support of this, chief among which is their headline-grabbing assertion that "one third of roads fatalities are now caused by speed cameras".

The "one third" claim originates from a levelling off of the downward trend in road fatalities which has existed since the 1970s. Smith attributes this to the advent of speed cameras in the early 1990s: the figure itself is arrived at by extrapolating the change in number of fatalities between two selected years to predict the fatality figure had that year-on-year trend continued, and comparing that with the observed fatality figures.

As a subsidiary point, SafeSpeed states that the historical decline of fatality rates means that government claims that the accident rate is improving under cameras are meaningless as the rate has been improving anyway.

SafeSpeed claims that the safety benefit from slower speeds has been overshadowed by greater risk from the side effects of speed enforcement. Drivers, fearing that a speeding conviction will have disastrous consequences for them, may become paranoid and be distracted from driving even if they are not actually speeding. Also, they may "panic brake" at the sight of a camera.

Smith and other SafeSpeed members have calculated that one second's distraction from the road will increase the speed of any impact by about 10mph owing to the lost braking opportunity. They claim that this is an enforced distraction demanded by the presence of a camera, unlike other distractions which are discretionary.

SafeSpeed argues that the use of maximum legal speed limits reduces the driver's ability to judge what is an appropriate speed for the situation, and that the widespread failure of motorists to drive at an appropriate speed in many circumstances, such as in fog on the motorway or down residential streets where children are playing, is due to the law placing a legal maximum of the speed at which they may drive.

More and more data is presented in car, and functions such as ABS and Brake Assist help complete this illusion that everything is "done for you".

The government claims that casualties at camera sites have fallen by large amounts, usually between 35% and 65%, depending on the site. SafeSpeed claims that the way sites are selected, this kind of fall is inevitable and does not necessarily represent a real improvement, since it is due only to regression to the mean. Because of the way camera sites are selected (a relatively large number of accidents must have recently occurred at that site), casualties at that site would be expected to fall back to their previous level even if no action were taken.

Effects such as regression to the mean and accident migration are well documented in the road safety literature.

One of the key facts in favour of speed enforcement and limit reduction is that at 40mph over 90% of pedestrians struck by a car will die; this falls to around 50% at 30mph, and around 10% at 20mph (Ashton & Mackay, .1979).

SafeSpeed have criticised this as applying only to the impact speed, which in normal circumstances is much less than the free travelling speed before the incident. They calculate that the average impact speed is 12mph, and that only 2% of pedestrians (and 0.5% of children, often featured in campaigns) who are hit by cars die, across all types of road. Emergency braking slows a vehicle by about 15-20mph per second, and a driver will usually have at least 1-2 seconds to take emergency action.

This is also the backbone of the claim that distraction can be more deadly than speed. They argue that 10-50% of collisions being fatal is not a figure to strive towards, as the government seems to suggest we should, since far fewer are currently fatal at present.

Smith lodged at least one (unsuccessful) formal complaint against public information films on this theme.

In common with most opponents of speed enforcement, SafeSpeed assert that speed cameras cannot assess dangerous driving or catch dangerous drivers.

SafeSpeed argues that while pro-camera lobby often claim that excessive and inappropriate speed is a major contributory factor in road accidents, they fail to quantify or adequately acknowledge the difference between excessive speed and speed above the limit.

SafeSpeed also notes that drivers detected by cameras are not stopped there and then, where they might be by police.

The Transport Research Laboratory is an internationally recognised authority on traffic and traffic safety. Smith denounces several TRL reports as "flawed" or even "fraudulent", but he has given great credence to an interpretation of a study conducted in the 1990s for a number of police forces, investigating an experimental system for accident reporting system for a number of police forces, documented in TRL report 323.

The results attributed 15% (sometimes stated incorrectly as 7.3% or lower) of collisions to excessive speed, and this was seized upon by SafeSpeed and others, as undermining the claim that speed is a widespread causative factor in accidents. It has been widely cited by SafeSpeed as evidence that speed is only a minor problem, and that much of that "excessive speed for the conditions" was in any case within the speed limit.

Since the downgrading of traffic policing in the police performance measures in the early 1990s, there has been a considerable drop in the number of dedicated traffic officers. SafeSpeed attribute this drop largely to speed cameras.

They cite concern that visible speed cameras, which are used on less than 5% of the road network, mean that many motorists feel that there is little need to comply with speed limits on those sections of road where cameras are not used, and at camera sites simply slow down before speeding up again. Police speed traps are not bound by the “3 coffins rule”, and can be set up anywhere. They usually detain offenders at the site, rather than allowing them to continue unhindered.

Counties where cameras are not in regular use, such as Durham have achieved reductions in fatalities through targeting key causes of death and injury such as “boy racers” and "born again bikers".

Opposition to SafeSpeed and its claims is widespread, and begins with its claim to be a road safety group. No other British road safety group is known to accept or support SafeSpeed’s agenda.

Critics have argued that Safespeed is much more about speed than safety, and is part of a "culture of speed" which also promoted by motoring magazines and TV programs such as Top Gear, presented by Jeremy Clarkson. It is argued that by stoking the controversy over cameras, SafeSpeed and others contribute to increasing the excessive focus on speed, rather than reducing it. Other organisations advocating safer speed choices emphasis that speeds must be both safe and legal.

The statement that what is shown to be incorrect will be removed from the website, is both contentious and disputed:

• it reverses the burden of proof, which normally requires that in order to be accepted a claim must be proven rather than being made and then challenging others to disprove it
• it requires in some cases an impossible proof of a negative
• in many cases (e.g. "one third of fatalities") SafeSpeed applies a higher standard of proof to opposing evidence than its own claims can meet
• where a claim has been proven to be based on false premises (the calculated 12mph impact speed, which is arrived at by using an empirical relationship at speeds and in conditions outside its stated applicability), it has not been removed.

There may be some merit in the suggestion that the speed limit suggests that a given speed is safe for the road, regardless of conditions, but this is explicitly contradicted by the Highway Code and there is little verifiable evidence to support it, except for ignorance of the code claimed by some drivers (a striking situation given that all new drivers are tested on their knowledge of the Code, which is also freely available online). While it is true that the safe speed for a road under given conditions may often be much lower than the speed limit, there is a no significant informed dissent from the finding that both probability and severity of collisions rises with increasing speed.

Many road safety campaigns are registered charities or limited companies. This appears not to be the case with SafeSpeed: for example, their funding appeal makes no distinction between personal and corporate debt. Supporters have no apparent guarantee that the money will be used for the stated purpose.

The various claims made by SafeSpeed and outlined above have also been vigorously rebutted.

The slogan “speed kills” has been used since at least the 1970s if not earlier. Speed enforcement goes back to the earliest days of motoring; the UK’s largest motorists’ organisation was founded to warn members of speed traps, and libertarian motorists have opposed regulation since the beginning of motoring (as satirised in Kenneth Grahame’s Mr. Toad in 1908). There is no evidence to link the slogan, or the idea that speed kills, with the advent or growth of cameras.

It is argued that even if a minority of drivers are distracted by speed cameras (by implication, due to unwillingness either to obey the speed limit or bear the consequences of speeding), the lower general speeds which result from enforcement have a greater effect. It is problematic to claim that an inanimate object such as a speed camera can cause a crash; clearly inappropriate driving is the true cause, the camera is at most a catalyst. If "high visibility" cameras are genuinely distracting perhaps covert speed enforcement should be much more widely used, or speeds enforced by means of automated Intelligent Speed Adaptation systems.

Opponents of enforcement say that say hidden cameras would be even more distracting than visible ones as drivers would see them late and panic brake to an even greater degree, but panic-braking for speed cameras - which implies greater regard for speed than safety - is not behaviour which is widely attributed to careful drivers, and reducing enforcement on the basis of such behaviour is effectively to tailor enforcement of the law for the benefit of criminals.

The “one second problem” is also disputed. First, measurements indicate that the time to look at the speedometer is probably closer to 1/3 or 1/2s. Second, the distance covered in this time is less than the difference in combined thinking and braking distance between, say, 40mph and 30mph.

Cameras are set to allow a margin of error, usually 10%+3mph (i.e. 36mph in a 30mph limit). Fixed sites are notified in advance. For the distraction problem to exist at all, the driver has either ignored the advanced warning of enforcement, perhaps preferring to remain at the top of the margin of error rather than the limit itself, or has failed to see both that and the camera until too late. And if the “distracted” driver were doing 30mph instead of 40mph the question would be moot anyway. Either way, the problem clearly lies with the driver not with enforcement - a driver who is comfortably within the prosecution guidelines will have no reason to be distracted, and ample time to react.

Finally, as part of the driving test every driver is required to drive safely while remaining within the speed limit –if they drive too slowly they will fail for “failing to make satisfactory progress”, if they exceed the limit they will fail for speeding. New drivers can be disqualified after accumulating six penalty points instead of the usual twelve, an indication that the Government is aware that the novice driver is more of a risk. Experienced drivers should have no difficulty in keeping to a speed which is both safe and legal.

The “one third” claim has a number of widely publicised problems, some of which were documented in a report in the consumer magazine ‘’Which?’’

• The fatality trend has been declining approximately exponentially, with the rate of change decaying as it tends towards an asymptotic constant level. To project a straight line as SafeSpeed does is therefore invalid.
• Road fatalities are subject to stochastic noise; with a fatality rate of approximately 3,000 per year from 15-20 million cars, the rate is low enough that individual events (a year with bad fog, for example) can influence the figures significantly. It is therefore invalid to extrapolate from differences between years.
• Documented changes in reporting mechanisms at around the time claimed mean that the figures before and after 1993 cannot be directly compared.
• The Killed or Seriously Injured (KSI) rate, which is less subject to fluctuation, exhibits a more consistent downward trend.
• The fatality trend for car drivers is also still downward, the levelling off of the overall trend is largely due to an increase in the number of motorcyclist deaths (possibly an effect of increasing motorcycling due to congestion and fuel costs).
• The fatality trend is worst for motorways and cross-country roads (where cameras are rare) and best for urban minor roads (where they are more common).

Even if one were to accept the claim that the road fatality trend has become worse since 1990, there are other problems with attributing this wholly to speed cameras:

• Mobile phone use has grown during this period, and studies show that this is detrimental to driving performance.
• There has been a significant growth in the number of large SUVs on the roads, and these have been shown to be more likely to kill others (especially pedestrians) in collisions.
• There has been a sustained period of economic growth, which is historically correlated with more road deaths.
• Car designs have changed, with thicker door pillars causing larger front-quarter blind spots. Smith has campaigned on this issue. This could explain increased fatalities among motorcyclists, due to so-called SMIDSY collisions.
• Fatality trends do not correlate with the numbers of cameras; radar detectors were first introduced in the mid 1970s but numbers of cameras rose most sharply between 2001 and 2003 (at which point the fatality trend had already flattened out and started to decline again, conicidental with a loss of growth in the economy).

SafeSpeed discounts these and other factors, but there is no verifiable evidential basis on which to do so – this exemplifies the difference in evidential standard applied by SafeSpeed to supporting versus conflicting evidence.

This argument is founded on the premise that without limits, drivers routinely choose a safe speed. It could justifiably be argued that speed limits exist precisely because they don’t. This dispute is It is, however, fundamental to disputes about speed enforcement generally. On one side are those who believe that they should be free to choose a speed they feel is appropriate, on the other are those who demand that drivers should select a speed which is both safe ‘’and’’ legal, as an admittedly crude means to limit the maximum risk they may pose on the highway.

SafeSpeed’s contention that the limit is seen as a target is supported by evidence, though explicitly contradicted in the Highway Code – though many drivers profess ignorance of the contents of the Code (on which they are tested on it as part of the driving test).

Against the idea of drivers setting a "safe" speed, the following arguments are put:

• In the case of speeding, all the benefit of increased speed goes to the driver, whereas the risk applies to all road users. Few drivers are hurt in collisions which kill and injure pedestrians.
• Most drivers overestimate their own skill, a fact which has been documented by numerous studies and is not disputed by any known authority in road safety and regulation.
• Motor traffic is uniquely dangerous. For example, collisions with motor vehicles account for one in ten child injures resulting in hospital treatment, but these make up half of all injury deaths, so there is an additional burden on drivers to limit the danger they pose.
• The “ratchet effect”: a risk taken which does not result in a collision comes after a time to be discounted, and the baseline for future risks increases - Alan Lennox-Boyd stated that collisions come in the main “not from the taking of large risks, but from the taking of small risks very large numbers of times, a view which is widely accepted.

Regression to the mean is more an argument about siting and conspicuity than about enforcement. The idea of "accident blackspots" is regarded as excessively simplistic by many road safety activists, being subject both to regression to the mean and accident migration, and the Government's "three coffin rule" for siting, combined with the high visibility of fixed cameras, means that camera sites have now been placed in the same general class as accident "blackspot" treatments, with similar problems, rather than being used to reduce speeds over the road network in general. It is ironic that location constraints and conspicuity are themselves a response to motorist opposition to cameras – it could be argued that SafeSpeed has actively contributed to reducing the efficacy of speed enforcement; if cameras were concealed and their location unrestricted, contributing to an overall reduction in speeding over the wider road network, the regression to the mean effect would be decreased.

SafeSpeed do not appear to dispute the relationship between impact speed and fatality risk. The 12mph calculation is, however, invalid. It is derived form Jocksch’s empirical formula relating probability of fatality to the fourth power of impact speed, for occupants of motor vehicles in highway collisions. It is not extensible below 40mph, and does not apply to pedestrian fatalities at all.

SafeSpeed states that only 2% of pedestrians who are hit by cars die, but this rather misses the point, tacitly accepted elsewhere, that the severity of outcome is strongly dependent on impact speed, and that even in their one second “distraction” the 40mph and 30mph drivers will stop at approximately the same point: in the majority case where the driver is ‘’not’’ “distracted”, the 30mph driver will stop well before the 40mph driver, and if a collision happens it will be at a significantly lower speed.

The idea that the Government, by stating the fatality rates at various speeds, is aiming for a fatality rate greater than we currently have is apparently SafeSpeed’s alone. No Government spokesman has ever given these figures as targets, only as illustrations of the way severity increases exponentially with speed.

SafeSpeed’s page claiming that "We use figures from official sources and well respected research to show that we could reduce all UK speed limits to just 12 mph and still have the same numbers killed on the road." Is another example of a claim being proven false without being removed from the website.

First and foremost this is an example of begging the question: it assumes that speeding is not itself dangerous, an idea which is at the very least open to dispute.

Further, research into the accident records of multiply-convicted speeders has showed that those with speeding convictions were ‘’around twice as likely’’ as average to be involved in collisions, mileage adjusted: it seems that cameras do catch dangerous drivers, and, through penalty points, give them an incentive to modify their driving. Middle-aged male company car drivers, a strong constituency in the ABD, have a higher than average collision rate, mileage adjusted, though still lower than for young male drivers - but they typically have a high annual mileage.

The perception that dangerous driving is a problem of ‘’other drivers’’ rather than something to address in one’s own driving has been identified by some authorities as a key problem in achieving greater road safety.

There is at least some verifiable evidence that cameras do not catch some dangerous speeding drivers. On 22 November 2005, according to Cumbria's safety camera partnership, more than 150 vehicles were caught travelling at 70mph or more in a section of contraflow near Tebay, where the temporary limit was 50mph. However, the fog was so heavy that the cameras failed even to make out the shape of some of them.

TRL strongly dispute SafeSpeed’s interpretation of TRL 323. In particular they point out that the study was dependent on subjective judgments of primary cause, and that many of the other primary causes listed also implied excessive speed. Other TRL studies (e.g. 421 and 511) have directly examined the relationship between speed and accidents, finding a strong association. Most importantly, a study of over 300 roads and encompassing several hundred thousand observations clearly showed that the faster the average speed of traffic on a given type of road, the more accidents there are; also, injury accidents rise rapidly as average speed increases, if all else remains constant.

SafeSpeed dismiss these and other studies as “flawed” while uncritically accepting TRL 323 – another apparent example of different standards being applied to evidence perceived as supporting SafeSpeed’s agenda versus that conflicting with it.

The documented reduction in traffic police began in the early 1990s, when traffic policing was downgraded in the police performance measures. Camera partnerships started some years later and the rate of camera deployment increased most sharply between 2001 and 2003. There is no evidence of a correlation between numbers of cameras and numbers of traffic police. Nor is there any evidence that removing cameras, which are revenue-neutral, paying for their own operation through fines raised, would result in increased numbers of traffic police.

Most road safety groups would welcome an increase in traffic policing, in addition to speed enforcement. The idea that the two are necessarily exclusive is a false dilemma.

• Road safety
• Speed camera
• Speed limit

• SafeSpeed
• Paul Smith profile at The Association of British Drivers site
• "They call themselves the voice of the driver. But who do they really represent?" - a Guardian expose of The Association of British Drivers.