Wednesday, 29 April 2009

People Are Going to Die at 50mph

I have been mulling this one one for a while. This is quite a bold claim, I realise, but if we press ahead with the reduction in the National Speed Limit (NSL) to 50mph then people are going to die as a result.

In case anyone needs clarification, the NSL is a default speed limit. Unless some other rule applies, then the limit is the NSL - currently 60mph. A different limit of 50, 40, 30 or 20 can be signposted, or a 30 limit can be assumed from the presence of lamp-posts, or a 70 limit can be assumed when on a motorway. In all other locations, the NSL applies.

So the discussion is clearly limited to non-urban roads; urban streets will default to 30 unless signed otherwise. Non-urban roads will typically be very variable, but generally break down into three types. The most common type is fairly straight, not so straight that you see to safely overtake, but not so bendy so as to cause concern. These sections are typically interspersed with shorter sections that either require a much slower speed to negotiate, or which offer an excellent view and allow an overtake if the vehicle in front is not making progress.

First, let's deal with the slow bits - the danger points. At the moment, these are marked. Some will have 30 or 40 limits posted, some will have other signposts. Many, however, have signs indicating a 50 limit. All of these signs will, over time, be removed. This means, in effect, that less warning will be given to drivers of the forthcoming hazard. Drivers will receive information from officialdom that the forthcoming stretch is no more dangerous than the one they have been negotiating for the last few minutes; this information will be false. There will be accidents as a result.

If you have a hazard ahead, it is not difficult to realise that reducing the palette of information for drivers is not going to help.

Then there is the question of the quick bits - the overtaking sections. Human nature is such that drivers who have been negotiating these same roads for years (decades, even) at 60 will not be pleased to be stuck behind someone firmly sticking to an indicated 50mph - often equal to an actual 45mph*. So they will be keener to overtake. When an overtaking section arrives, there will be more pressure to go for it. I have also observed that there is a class of driver who always drives at "limit minus x". Their speed on NSL roads seems to drop from 45-50 to 35-40 when the limit drops from 60-40. This is frankly ridiculous; a road section that is objectively safe at 60 then has a queue sitting at 35 - this is often the source of a perfectly reasonable desire to overtake. I have, in fact, noticed more overtaking manoeuvres now that many formerly NSL roads have a 50 limit. There will also be some drivers that will accept more marginal overtaking locations; create a slow-moving roadblock and you create a queue of people who want to get past. Not all of them will have time to do so at the safe spots.

Overtaking can be safe, and can be wise. It can also be very dangerous. It is not difficult to realise that placing more pressure on some drivers to get past is not going to help.

I'm not making this up, either. A long open road near to me used to be NSL, but about three years ago - upset at the rate of deaths on the road - they put up a big yellow sign telling us to be careful because 60 people died or were serously injured over the 5 mile stretch over the previous 3 years. They also dropped the limit to 50. The sign has just been updated again. 93 died or were seriously injured** over the last 3 years. That is, roughly speaking, an additional death (or serious injury) every month or so.

I'll just repeat that, because it frightened me. An additional death (or serious injury) every month or so. Somehow, I don't think it was the big yellow sign that caused this.

*Speedometers are allowed to be up to 10% optimistic, but must not under-read. Engineering-led manufacturers therefore set them to over-read by 5%. Marketing-led manufacturers set them to over-read by 10%.

**The original version of this post simply said "died"; Steve Jones commented that this would be very surprising and, on that point, I have to agree with him. I'll go back and check; the sign may well have said deaths or serious injuries, rather than just deaths. In my defence, I was driving past (!) and therefore my attention was (and needed to be) elsewhere. Nevertheless, I should have spotted that. Until I can check it, I think it would be best if the post referred to the more likely of the two possibilities.


  1. The government is not interested in facts, the government is interested in taking decision-making responsibility out of the hands of the drivers and putting into the hands of the ministers.


    70mph is the speed limit for any road with more than one carriageway, not just motorways. Except in the case of lorries, who have a limit of 60mph on dual carriageways and 70mph on motorways - although many lorries are physically limited to around 60mph anyway.


    I think your logic on the danger of reducing the speed limit is excellent. The chances that the decision will be based on logic, facts, research or statistics, though, is very low.

    As with most things in life.

  3. By the way, I take it you have read J.J.Leeming's research on the subject of speed limits, the results of which seem to have been largely ignored.

  4. I think you all need to go on a Speed Awareness Course who test the knowledge of speed limits .....I am not sure Stu would pass :-)

    I think the government's hidden agenda is to discourage the private ownership of cars (and, oh yes, fundraising) but this cannot be said in public because of the vested interests of car manufacturers et al. Given the statistics, fatalities from road accidents are given surprisingly high prominence especially as some accidents will always be inevitable. There must be an argument about diminishing returns.

    Still, at least, it lets you all have a jolly good discussion about the effectiveness of road signs. Nice piece, Patently.

  5. Oops, thank you for the (almost correct) correction Stu. I was thinking only of cars (without trailers) and mentally merged motorways and dual carriageways. Tsk.

    You, BE & Measured are, I think, absolutely right on the motives and methods of this decision.

    YMB - yes, I have heard of JJ Leeming. He made a serious mistake, though; he based his work on evidence. From that point, it was doomed to be forever ignored.

    A final thought, inspired by Measured's comment. How many know that more die in the UK from hospital-acquired infections than in road traffic accidents?

  6. So, if you are speeding to get to the hospital, you are probably as good as dead already.

  7. Measured is right. It's an anti-car & and anti-individual thing.

    It's about maufacturing congestion (and indeed death) as part of the statist hate for the indivdualist mode of transport. Contemplate the contribution to safety and the flow of traffic of no-car lanes as you drive or the easy flow of traffic your local council's one way system is producing.

    Oh and as a bonus we get a nationwide survellancie and tracking system too!

    Still, there is hope, some people have woken up: Destroyed Gatsos - look at each one and contemplate the scale of this.

  8. There is no road in the UK where there are as many as 93 (or even 50) deaths in a 5 mile section. The AA Motor Trust analysed the figures annually (now taken over by the IAM), and the worst in recent years is on a 24km (15 mile) section of the A682 with 27 fatalities over the three years from 2001 to 2003.

    Given the one bit of evidence that you quote (on an unidentified road) is in direct conflict with the AA analysis, then I think it certain that you've confused "casualty" with "fatality". If we take the official government statistics on what proportion of casualties are fatalities (1.2% in 2007), then that figure of 93 deaths would be reality be 1 death in the 3 year period. Without a breakdown for the section of road in question, it's impossible to say what the fatality numbers were in reality (the higher speeds on trunk road could well mean it is a notably higher than the national average of 1.2%). An increase in the casualty rate of 1 per month is still something to be concerned about of course, but I'd have to run this through a statistical significance test to see if it is within the bounds of random fluctuation (and it's more important to know if the definition of casualty has changed).

    As for the rest of the analysis, it follows a very weak chain of speculation and logic.

    Personally I doubt very much that a reduction in the NSL to 50mph will increase fatilities, if only for the reason that, if complied with, there is a 30% reduction in the amount of kinetic energy to be dissipated in event of a collision. It would also increase the available thinking time. It may be that behavioral changes, like more risky overtaking, would counteract it, but that is very much speculation.

    One area that the government could look at is the rate of motorcycle fatalities. In contrast with the overall fatality rates (down steadily since 2003), those for motorcycle riders increased in 2007 to 599. That's about 20% of total road fatalities, and is vastly out of proportion to the total passenger miles. The fatalities are disproportionately concentrated on the riders of more powerful "sports" motorcycles. Discouraging the riding of powerful motorcycles, which are inherently more vulnerable, less visible and less stable than cars would undoubtedly save lives, even with some displacement of riders to cars. However, I think the biggest cause (unacknowledged by motorcycle pressure groups) is that such riders are less risk averse than most car drivers as such machines are inherently more dangerous by any objective measure.

  9. I've made a related mistake, but it only emphasises the point. The figures I quoted for the 15 mile section of the A682 (claimed to be the most dangerous in the UK) are for "serious and fatal" accidents. According to the BBC report, the total number of fatalities and serious injuries on that section of the A682 was just under 100, so the average number of deaths per year would have been below ten. The BBC don't provide a breakdown between serious injuries and fatalities (it is incorrectly reported as almost 100 fatalities by some media outlets)

    The following from the IAM has a better breakdown of accidents on major roads in the UK (but still combining fatal & serious into one category)

    It is notable that on the A682 (according to this report) 47% of fatal and serious accidents on that stretch involve motorcycles, a simply staggering proportion.

    What this does show is that the claim of 93 deaths in 3 years on 5 mile section of a road in the South East simply has to be wrong. If you can identify the road and the notice showing 93 fatalities in 3 years on the section of road you refer to, then I will stand corrected. But now it looks even less likely.

  10. I'll go and check the sign; you may well be right. 93 actual deaths does sound unlikely, now you point it out. I was driving past at the time, so my concentration was elsewhere! If so, I'll correct the original post.

    As you acknowledge, though, the point remains valid as whatever I am measuring, I am comparing like with like as the sign originally said 60 (whatevers).

    I think my point of disagreement with you is at the point:

    "if complied with"As I have argued before, the use/abuse (depending on your opinion) of speed limits and speed limit enforcement has reduced the general level of respect for road traffic laws and reduced the level of compliance. This means that on roads where the limit is genuinely justified (of which there are many), the level of driving is in fact more dangerous.

    You also acknowledge that:

    "It may be that behavioral changes, like more risky overtaking, would counteract it, but that is very much speculation. So my core point, you dismiss as speculation. Except that is isn't, is it? It is based on my personal observation over 20+ years of driving, and it is based on the actual figures from when this was done to the formerly NSL road near to me.

    It is the reduction of the NSL to 50 is in fact based on pure speculation and assumption, not my concerns

  11. I can bring my personal experience to bear to. I've been driving (including motorcycles) since 1971. I wouldn't agree with your general observation over compliance. Since 1971 there has been a huge drop in virtually all types of road casualties. In 1979 there were 6,352 deaths on the road whilst in 2007 there were 2,946 and rates have declined markedly in the last four years (for a few years before that the figures plateaued). That's against the background of a massive increase in traffic, although there have also been lost of car improvements, better crash protection et al so the picture is complex.

    As for whether you are comparing like-with-like, then that depends if the highway authorities have changed the criteria for "casualties". We live in an age of public propaganda, so I wouldn't be too sure that it is like for like. However, even then one statistic on one road without any statistical significance check does not make a convincing case, especially when it runs directly opposite to wider, national trends.

    Yes - and I do consider what you came up with as highly speculative. You are the one asserting that lowering the NSL will lead to an increase in deaths. You call it dismissing, but if you make bold claims then you need to make a convincing case in order to overcome the obvious points which come straight from Newtonian physics.

    I think a more reasonable claim might be that it won't result in the expected saving of lives, and that it might not be justified. One thing I can say, is that the proliferation of speed limits has made driving a good deal more stressful in terms of dealing with all the roadside signs. That doesn't mean it has become more dangerous, but the possibility of accidentally transgressing has definitely been increased.

    Lost of stats available here :-

  12. Newton's laws do indeed state that of two crashes, the one that happens at the higher speed will cause more harm. But to use this as your only guide is to fall into the "Speed Kills" argument, and to mistake the intuitive for the correct.

    The fact remains that there will be less harm as a result of a crash that an attentive driver avoids than as a result of a crash that happens to an inattentive driver travelling below the speed limit. Therefore, if we are really interested in reducing injuries and deaths, we will work primarily on improving the attentiveness of the driver(s) involved, improving their level of awareness as to what is going on around them. That itself can lead to a reduction in speed, in places, as drivers start to notice the hazards and adjust their driving accordingly.

    We are not doing that, though. We are fiddling with every last speed limit in order to create a “drive by numbers” world in which people are given a “safe” speed at which to drive and encouraged to stop thinking. So often, I am overtaken or bullied simply because I slow for an evident hazard, only to either leave the same driver way behind or become stuck behind them after the hazard has been passed. Clearly, they are just sticking to one predetermined speed rather than looking outside and driving according to the conditions.

    You do note: “One thing I can say, is that the proliferation of speed limits has made driving a good deal more stressful in terms of dealing with all the roadside signs”. Now, a skilled and experienced driver or rider such as yourself may be able to cope with this, but others will find it harder. Either way, both are being distracted from their primary task of controlling the car, which is not a good idea. Yes, their accident might be less harmful, which is a good thing, but if it becomes more likely then we are much worse off overall.

    My driving experience, which regularly includes several formerly NSL roads that now have 50 limits, has been that since those limits were introduced I have seen more near misses, I have seen more police boards asking for witnesses, and (on the particular stretch I mentioned) the official KSI tally has gone up not down. As a result, when I hear that this is to be applied blindly to every road in the land, I fear for our safety.

  13. Interesting discussion, however, I think Patently and Steve are at crossed purposes. It is perfectly consistent with Patently’s point that some roads may have fewer fatalities with a lower limit. His target seems to be a one size fits all approach.

    In order to defeat Patently Steve needs to prove that there is no road anywhere which would have more fatalities with the lower limit. I wouldn’t wish to defend such an absolute proposition until it had been tested: Patently only needs one example to falsify it.

    Patently has given good reasons to suppose there is at least one such example. Granted, the evidence is rather limited, but then this is a blog not a policy paper, and he only needs one road to prove his point. My guess is that he is right: some people on some roads will simply continue to drive at 60, only because the lower limit has caused them to over-take, they will be on the wrong side of the road when they do so.

    Steve says “I think a more reasonable claim might be that it won't result in the expected saving of lives, and that it might not be justified.” This is interesting, if there are the same number of deaths, but people are less likely to die when they have an accident, it would seem to follow that there must be more accidents. So that means with a lower limit, my journey takes longer, I’m more likely to be involved in an accident, and still I’m no less likely to be killed. I wouldn’t describe that as safer!