Road Safety News

Cameras may have ‘deterrence effect’ even when switched off

Friday 11th March 2016

A report looking at casualty stats at speed camera sites in Northamptonshire after the cameras were switched off found “no significant change in collision rates at the fixed camera sites post-switch-off compared to the collision trends seen elsewhere in the county”.  

The report’s author, Richard Owen from Road Safety Analysis, says this is “significant as many people would have expected a rise in collisions”.

The report looks at the four year period prior to the cameras being switched off (April 2007 to March 2011) and the four years after (April 2011 to March 2015).

In the period after the cameras were switched off, the findings highlight a 45% reduction in KSI at camera sites (29 to 16), compared to a 27% reduction across the rest of the county’s road network (1,628 to 1,193). In the same period, casualties of all types at camera sites were down 21% (from 90 to 71) while across Northamptonshire’s other roads there was a 29% fall (from 7,293 to 5,189).

The report concludes that “the changes between the two periods for Northamptonshire’s roads and the speed camera sites are not statistically significant”.

Richard Owen said: “The statistical analysis of the results show that there has been no significant change in collision rates at the fixed camera sites post-switch-off compared to the collision trends seen elsewhere in the county.  

“This is significant as many people would have expected a rise in collisions if people started to ignore the presence of the cameras and increased speeds.

“The report does not look at speed survey data as this is not publicly available so it is impossible to say whether speed have changed at the sites, although anecdotal evidence suggests that people do indeed slow down for the cameras.

“The findings suggest that fixed cameras have a deterrence effect, regardless of whether they are operated, and this could be important information for other areas.”


Comment on this story
Report a reader comment

What's your view - comment on this story:

I confirm that I have read and accept the moderation policy and house rules relating to comments posted on this website.
Your comment:
Your name and location:
Your email:
Captcha [What is this?]

Actually, Rod, the opposite is true.

When you have a "simple set of circumstances", a simple before/after test might be accurate but, if the system is more complex, there may be many other factors that influence the result. The effect of the intervention therefore has to be separated out from all the other factors and this can be difficult to determine, inaccurate and open to interpretation.

The whole point of a scientific trial is that all other factors are automatically and accurately compensated for, leaving the intervention as the reason for the differences found. The scientific trial is a simple, cheap and accurate test specifically relevant for use in complex systems.
Dave Finney, Slough

Agree (0) | Disagree (0)


Of course, your simplest, cheapest and most accurate test will only work when you have a simple set of circumstances. With road casualties being a combination of so many social, attitude, technical, regulatory, expertise and random factors then "randomised control trials" are simply not viable. And that is the reason why they are not being used.
Rod King

Agree (3) | Disagree (3)

I note Richard has said no data is available on vehicle speeds in which case, the exercise couldn't hope to be conclusive. All fixed camera sites, to my knowledge, have loop detectors downstream (typically), so surely before and after speeds would have been available as a good indication of their influence if, as he says, 'local' motorists (whatever that means) were aware (how?).
Hugh Jones, Cheshire

Agree (6) | Disagree (1)

I agree with Richard. All his data shows is that there was a 45% reduction in KSI where speed cameras were switched off, but we don't know why that reduction occurred. The possibilities are:

1. random variation
2, speed cameras contribute to more KSI
3. other factors

Richard and I are both trying to determine what is causing which effects to occur, but how can this be achieved in a way that everyone can agree on? What we need is a test that is completely neutral and unbiased.

Let's all join together and agree to use the simplest, cheapest and most accurate test known to mankind, the standard scientific trial known as a "Randomized Controlled Trial".
Dave Finney, Slough

Agree (7) | Disagree (4)

Have I got this right.. the cameras were deliberately 'switched off' to see if crashes would increase as a result of people ignoring the cameras? Inducing crashes in the name of research? Or - and what I'd like to think - were the cameras going to be switched-off anyway for an unrelated reason, in which case, why on earth tempt fate by advertising the fact? It'd be like advertising the fact that the police had run out of roadside breathalysers on New Years Eve to see if crashes increased.
Hugh Jones, Cheshire

Agree (7) | Disagree (3)

To address a few points:

The study was carried out because the local motorists were told that the cameras were switched off. The thought at the time was that crashes would increase in frequency as people ignored the cameras. There is no evidence to show that collisions have increased and no data is available on vehicle speeds.

The report was not paid for by either the Northamptonshire Police or County Council.

Dave – You need to read the section on statistical tests. The small numerical changes in the samples were not statistically significant and could therefore have arisen by chance. That means you are incorrect in your conclusion about changes post-switch off. Feel free to apply the same methodology to your previous projects.
Richard Owen, Banbury

Agree (6) | Disagree (8)

The data in the report doesn't seem to support it's main conclusion. What the report does show is that, after the speed cameras were switched off, there was a 45% reduction in KSI at the speed camera sites (compared to a 27% reduction elsewhere).

It is quite possible that the speed cameras were contributing to greater numbers of KSI and, now that they are switched off, citizens are able to concentrate more fully on avoiding other road users. This could help to explain why the sites are now much safer with the speed cameras switched off.

I argued to retain half of the speed camera sites in a "Randomized Controlled Trial". Sites should have been selected into pairs, and then one of each pair chosen at random to be removed or covered. The remaining sites enforced as normal (1/2 of the total). This would have provided several benefits;

1) the world's first ever scientific trial of a road safety intervention.
2) the most accurate evaluation of speed cameras that it is possible to achieve
3) cheaper than performing estimates of what effect might have occurred

Let's start running scientific trials and produce the proof of effectiveness in road safety that we have come to expect in other fields of safety engineering.
Dave Finney, Slough

Agree (7) | Disagree (10)

Negative effect number 41.
It’s probably not well known, but there was secret government research many years ago into the suspected effects of the electro-magnetic wavelength of shade of yellow used by speed camera housings when exposed to motorists travelling at exactly the 85th%ile speed on a Tuesday on a rainy day and whilst listening to any piece of music with a 12/8 time signature and - most importantly - in a vehicle with a VIN ending in 7 and in a vehicle whose colour's wavelength was a perfect fraction of the aforesaid camera housing yellow.
In certain circumstances this can apparently lead to the observer developing paranoia about speed cameras.
Hugh Jones, Cheshire

Agree (10) | Disagree (7)

Some years ago, Paul Smith identified about 40 negative effects of speed cameras:
They also all apply whether or not cameras are switched on. Any safety assessment needs to look at negative as well as positive effects.
Eric Bridgstock, St Albans

Agree (8) | Disagree (6)

I think that we have to remember that by cameras being so conspicuous, they have an effect of raising speeds on the rest of the network by their clear non-existence. Their conspicuity promotes the idea that only "here" is speed limit compliance required.

We would all be served better by less conspicuous and more numerous cameras which endorsed the idea that speed compliance is required everywhere rather than just where history shows a greater prevalence of casualties and trajedy.
Rod King, 20's Plenty for Us

Agree (6) | Disagree (13)

Is someone trying to justify their position? Why waste time and money undertaking a study and preparing a report that will probably be read by only those that knew this anyway.

Agree (12) | Disagree (0)

"The presence of fixed cameras have a deterrent effect, regardless of whether they are operated or not". Well, yes, of course they do. Why would anyone think they didn't? The driver's "not knowing" if a camera is live is certainly a deterrent. Reporting on the blindingly obvious I would say.
Pat, Wales

Agree (10) | Disagree (0)

Shock news, Drivers who don't know if cameras are on or not value their licence and don't want to be fined and the subsequently have insurance price increased so slow down.
Steve Armstrong, Halifax UK.

Agree (8) | Disagree (0)

I'm not sure why anyone would think there would be a difference in speeds and driver behaviour in the vicinity of a camera just because it is not-operational. If the outward appearance remained the same, how would the drivers know whether they were operational or not? Does 'switched-off' mean they still 'dummy flashed' as a deterrent, or were they strictly non-active?
Hugh Jones, Cheshire

Agree (8) | Disagree (0)

Makes sense because who is going to risk speeding through a camera in the hope that it isn't active? Here in Staffordshire I made an FOI request a few years ago - only around 34 of the fixed cameras out of 263 were ever live - they move them from camera housing to housing so drivers don't know which ones are active.
Paul Biggs, Staffordshire

Agree (11) | Disagree (0)