Crossing roads can be (gasp) dangerous!

timlennon
Posts: 41
Joined: 8 Feb 2011, 1:24pm

Crossing roads can be (gasp) dangerous!

Post by timlennon »

So I read through the links and some of the papers in John Franklin's research page: http://www.cyclecraft.co.uk/digest/research.html

And it seems to me that the overwhelming body of the research demonstrates - in what seems at the moment to be one of the most unsurprising outcomes of the 20th C. science, that when you take a cycle path and have it cross a road (typically when the road maintains priority) then that makes the cycle path 'more dangerous' than cycling on the road.

Which seems something of a lame justification for not endorsing wider use of segregation. I've seen a number of articles /posts which draw this link, and it seems poor reasoning to say the least. Given that whenever a cyclist is involved in an accident it is overwhlemingly likely to be with a motor vehicle, one can only ask 'how come it's unsafe to separate cyclists and motor vehicles?' And the answer appears to be that if you give cyclists a segregated facility, the average cyclist will prefer it, and then use it with a lower level of attentiveness than is needed on a road. Thus when you arrive somewhere like here (http://is.gd/Hr40tg) then as a cyclist it's hard not to keep going. (Note, the image is a bit old: a marked cycle lane now runs up to the two bollards, with a give way sign to allow cars to turn off the main road.)

I could probably construct this argument better, but it's late on Saturday night: I'll attempt an edit tomorrow if it isn't clear. But, in summary, Franklin's research page looks like a bogus, misleading pile of rubbish ...
snibgo
Posts: 4604
Joined: 29 Jun 2010, 4:45am

Re: Crossing roads can be (gasp) dangerous!

Post by snibgo »

timlennon wrote:... a bogus, misleading pile of rubbish ...

How come? What is bogus or misleading?

The papers seem to show that cyclists are generally in reduced danger between junctions (because they are segregated there), but not at junctions (where they need to interact with traffic at what is now a more complex junction). And that's what you seem to say in your first substantive paragraph.

And that leads me to suppose that if we want to reduce danger to cyclists, we might build a network of paths that don't have junctions with roads. Do you agree?
irc
Posts: 5192
Joined: 3 Dec 2008, 2:22pm
Location: glasgow

Re: Crossing roads can be (gasp) dangerous!

Post by irc »

Cycle paths men the user riding along a main road exchanges having priority over sidestreets and being in a position drivers expect to see other vehicles to having to yield priority at every side street and being less visible to drivers.

So of course they are more dangerous and yes the danger is at juctions. So what is bogus?

Seperated paths with long distances between junctions, like canal towpaths for example can be safer than roads depending on the number ofd other users and the surfaces etc. But routes like these are rare and other than the historical accidents of old towpaths and railway routes they are not going to happen in modern cities.
No one believes more firmly than Comrade Napoleon that all animals are equal. He would be only too happy to let you make your decisions for yourselves. But sometimes you might make the wrong decisions, comrades, and then where should we be?
User avatar
Cunobelin
Posts: 10801
Joined: 6 Feb 2007, 7:22pm

Re: Crossing roads can be (gasp) dangerous!

Post by Cunobelin »

There are many reasons why this is the case.

There is a problem with 'priority' as quite often this is unclear. There are many driveways or entrances where technically the cyclist has priority, but vehicles will not give way, despite the press insistence otherwise, a cyclist approaching the junction on a "pavement" is unexpected and not necessarily looked for and hence not seen. Even when they are seen, drivers will drive out anyway as they are nor restricted by the signage.

The field of view is less for both cyclist and road user as the cyclist is closer to obstructions such as walls and infrastructure so the field of view is limited

Edited... and of course the BIG one, where cyclists are "left hooked" because motorists refuse to recognise the cycle facility and turn across the cyclists path illegally

A classic example - not as bad as some, but dangerous for both the cyclist and pedestrian at this junction.
thirdcrank
Posts: 36776
Joined: 9 Jan 2007, 2:44pm

Re: Crossing roads can be (gasp) dangerous!

Post by thirdcrank »

timlennon wrote:....I could probably construct this argument better, but it's late on Saturday night: I'll attempt an edit tomorrow if it isn't clear. But, in summary, Franklin's research page looks like a bogus, misleading pile of rubbish ...


Please do. Whatever you may feel about John Franklin's stuff, I think he is entitled to more than your final unfinished sentence.
timlennon
Posts: 41
Joined: 8 Feb 2011, 1:24pm

Re: Crossing roads can be (gasp) dangerous!

Post by timlennon »

Please do. Whatever you may feel about John Franklin's stuff, I think he is entitled to more than your final unfinished sentence.


And that's true. On some reflection, I think my issue is less with the research per se, than with the way it tends to be used or quoted. I can't imagine anyone thinks it is a surprise that, when a segregated path meets the road, there are more accidents. So much discussion around this research, and around segregation seems to then use this to suggest that segregation is 'more dangerous than sharing the road'. Perhaps it's one for the people from Freakonomics, but it would seem clear that, if a given cycle lane is measurably more dangerous than a parallel length of road, there is something missing in the design of the lane, rather than it being ab initio more dangerous. (For example, why is there the presumption that drivers have the priority at every junction?

Yes, such changes involve things like driver education, and a dozen other things, but that doesn't make cycle lanes / segregation an inherently bad option, does it?

(And in case you're wondering I've purchased and read Cycle Craft, I just don't imagine it's something my girls will be ploughing through when they're old enough to go to primary school.)
snibgo
Posts: 4604
Joined: 29 Jun 2010, 4:45am

Re: Crossing roads can be (gasp) dangerous!

Post by snibgo »

I'd say Cyclecraft [EDIT: isn't] aimed at primary school children. But it does have a chapter addressed to parents.

A long thread here (viewtopic.php?f=6&t=47121) discussed a paper that found six cycle tracks in Montreal were safer for cyclists than corresponding reference streets. Such papers are rare; most find such facilities are more dangerous.

I agree that the increased danger isn't at the actual segregation, but junctions with roads. It seems obvious to me that when a junction becomes more complex, it also becomes more dangerous, unless other measures are taken.
Last edited by snibgo on 20 Feb 2011, 2:39pm, edited 1 time in total.
thirdcrank
Posts: 36776
Joined: 9 Jan 2007, 2:44pm

Re: Crossing roads can be (gasp) dangerous!

Post by thirdcrank »

Perhaps this is a case where theory has to be set in a real life context.

Are cyclists safer in the absence of motor traffic? Almost certainly. (That's the theoretical bit.)

Motor traffic is not going to disappear, so is separate provision a solution? Possibly: it seems* that in the Netherlands at least, extensive separate provision has been created which treats cycling as a viable transport mode and where the two modes have to share space, different attitudes prevail - notably that cyclists are respected.

Might the experience of the Netherlands be replicated here? (That's the real life context bit.) As the genie said "Let's go back to your first wish. How wide did you say you wanted that staircase to the moon?" :wink:

* I have only reports to rely on, including one from my son who has lived and worked in the Netherlands for the greater part of the last decade.
Steady rider
Posts: 2749
Joined: 4 Jan 2009, 4:31pm

Re: Crossing roads can be (gasp) dangerous!

Post by Steady rider »

In the Netherlands they have a cyclist fatality rate of about 11 per billion km and the UK about 24 per billion.

The Netherlands is full of cycle paths and road crossing, therefore if the junction risk was high their fatality rate would not be much lower than the UK.

By requiring and designing for motorists to give way to cyclists the risks are reduced plus drivers have become used to this procedure in NL.

If you look closely at the data for NL you will see many older cyclists dump up their fatality rate, medium value for most age groups is about 5 to 6 per billion km.

As a side issue pedestrian data shows their fatality rate to be quite low per billion km walked, in part probably due to being able to use cycle paths.

For the UK, poor design, loss of priority, short sections, all add to inconvenience and little value. On top, is the issues of motorists and others wanting cyclists to use the low quality facilities sometimes provided. The police have occasionally stopped cyclists and requested them to use the cycle path adding to opposition.

When cycle paths are provided you are more likely to get less experienced cyclists using them plus if they are not really well designed, then risks increase, falls at entry on raised edges are one risk.

UK data about 55% of cyclists accidents at junctions, NL may be higher 73%, figures would need checking.
NL say 11 per billion km, at junctions 73% = 8 per billion km.
UK, say 24 per billion km, at junctions 55% = 13.2 per billion km

Away from junctions
NL 11 per bill km, 27% say, 3 per billion km
UK 24 per bill km, 45% say 10.8 per billion km

To start with a full assessment of the data would be required.

Apart from the accident side there is the health side, cycling provides health benefits many times higher than the injury dis-benefits, 20 to 1 Hillman etc and cycle paths encourage more cycling. Roughly about twice the rate can be achieved.

UK now, injury rate X, health benefits 20X,
Good cycle paths or similar, injury rate probably about 1.8X, health benefits 40X, gain 18X.

To move forward higher standards for roads and cycling facilities,
cycling facilities – markings - lanes – off road paths – junctions etc.

Currently the UK designs are not up to the level required. The CTC should provide their own basic standards to highlight where changes are required. Cycling is too low a priority for progress to occur easily.
snibgo
Posts: 4604
Joined: 29 Jun 2010, 4:45am

Re: Crossing roads can be (gasp) dangerous!

Post by snibgo »

Your methodology confuses accident rates with fatality rates; it should use one or the other.

However, the conclusion is plausible: NL is somewhat safer than the UK for cyclists at junctions, and very much safer between junctions.
timlennon
Posts: 41
Joined: 8 Feb 2011, 1:24pm

Re: Crossing roads can be (gasp) dangerous!

Post by timlennon »

I agree that the increased danger isn't at the actual segregation, but junctions with roads. It seems obvious to me that when a junction becomes more complex, it also becomes more dangerous, unless other measures are taken.


Although I saw research a couple of years ago which suggested that, counter-intuitively, the more complex a junction - at least for cars - the safer it was. The reasoning seemed to be that when one approaches a complex junction, one automatically slows down / pays more attention / thinks a bit harder about what's about to happen, whereas with simpler junctions, there's a lot of auto-piloting. I'm not sure how this would work with mixed cycling / walking / driving junctions, mind.
snibgo
Posts: 4604
Joined: 29 Jun 2010, 4:45am

Re: Crossing roads can be (gasp) dangerous!

Post by snibgo »

I think the question is whether a motorist perceives a junction to be complex, and therefore more demanding of attention.

In Stevenage, at roundabouts with dual-carriageway approaches, they once had the bright idea of planting hedges in central reservations. The theory was that drivers couldn't see what traffic was already on the roundabout so would have to slow right down.

It had the opposite effect. Drivers couldn't see traffic already on the roundabout, so slammed into it at 40 mph.
thirdcrank
Posts: 36776
Joined: 9 Jan 2007, 2:44pm

Re: Crossing roads can be (gasp) dangerous!

Post by thirdcrank »

I think the issues at complicated junctions are influenced by familiarity. Not only do the locals become used to the layout, they also adopt local practices, influenced by things like traffic light sequences.

The notorious local junction J27 on the M62 was 'improved quite recently with multiple traffic lights. It quickly became apparent that by following legal but unsigned routes through the junction, time could be saved. I was recently driving through doing just that when a driver behind began hooting really aggressively. When he was eventually able to overtake he was still hooting and wildly gesticulating at me, before he shot off, weaving from lane to lane to leapfrog the traffic. It was only after a bit of thought that I realised he had assumed that I would turn out of his way, rather than continue on the same route as his, but observing the 30mph limit.

I think this means that there's a mix of drivers who know where they are going and others who are bemused. Impatience, sometimes growing to aggression are overlaid, especially if some 'idiot' betrays signs of being lost.

Apart from that, since accident stas only record collisions involving injury (and the trend is towards only considering so-called KSI) the countless knock- for- knock shunts and scrapes go unrecorded and are increasingly regarded as inevitable. Any of these might kill or seriously injure a cyclist or pedestrian so these modes are then seen as being somehow dangerous.
irc
Posts: 5192
Joined: 3 Dec 2008, 2:22pm
Location: glasgow

Re: Crossing roads can be (gasp) dangerous!

Post by irc »

timlennon wrote:if a given cycle lane is measurably more dangerous than a parallel length of road, there is something missing in the design of the lane, rather than it being ab initio more dangerous. (For example, why is there the presumption that drivers have the priority at every junction?


You can't design out driver behaviour. A main road has priority so drivers entering it are mostly cautious. Drivers don't expect to have to give way to cyclists on cycle paths. This may be one reason the Dutch figures are lower. Cycling is so common that drivers look for them.

The other thing about road cycling is that it puts cyclists where drivers perceive their danger is coming from. Drivers only see what they look for and that is traffic on the road. Drivers won't spend a great deal of time looking for bikes or pedestrians when it is only other motor vehicles that pose a risk to them. If an rhino is charging at you from one direction and a mouse from the other where will you be looking?
No one believes more firmly than Comrade Napoleon that all animals are equal. He would be only too happy to let you make your decisions for yourselves. But sometimes you might make the wrong decisions, comrades, and then where should we be?
Steady rider
Posts: 2749
Joined: 4 Jan 2009, 4:31pm

Re: Crossing roads can be (gasp) dangerous!

Post by Steady rider »

snibgo » Sun Feb 20, 2011 4:27 pm
Your methodology confuses accident rates with fatality rates; it should use one or the other.


trying to provide a bit more info.

http://policy.rutgers.edu/faculty/puche ... stible.pdf
fig 10 provides the fatality and injury rates

UK study PPR 445
UK –“A high proportion of collisions occurred at junctions; almost two-thirds of cyclists reported killed or seriously injured at or near junctions”

NL fig 4, fatality rates per age group, http://www.swov.nl/rapport/Factsheets/U ... clists.pdf
Near to fig 4,
“resulted from a crash in an urban area; 67% of these occurred at intersections and 33% on road sections”.

http://www.swov.nl/rapport/Factsheets/U ... lities.pdf
Table 1 deaths
30 road section, 61 intersection.

Table 2, injuries
333 road section, 919 intersection, total 1252, intersection 73%

It appears the accidents at intersections in the UK and NL may be of similar percentages but NL a level of safety higher, safety in numbers perhaps and their driving practices reflecting the higher cycling levels.

Excluding the 60+ age group the NL fatality rate may be about third of the UK's per km cycled.
With improvements the UK could be below 100 cyclists per year, instead on 100+.
Post Reply