Latest reported road casualty figures covering 2015-2020

[Arjimlad]

https://www.gov.uk/government/stati...s-in-great-britain-pedal-cycle-factsheet-2020

Key findings
Between 2004 and 2020:

• fatalities increased from 134 to 141 (5%)
• serious injuries (adjusted) rose by 26%
• pedal cycle traffic grew by 96%

Averaged over the period 2015 to 2020:

• an average of 2 pedal cyclists died and 83 were seriously injured (adjusted) per week in reported road casualties
• a majority of pedal cycle fatalities (59%) do not occur at or within 20m of a junction compared to 32% of all seriously injured (adjusted) casualties
• almost half (46%) of pedal cycle fatalities in 2 vehicle accidents involved a car
• 56% of pedal cycle fatalities occurred on rural roads compared to 29% of traffic
• 83% of pedal cycle killed or seriously injured (KSI) casualties were male

The most common contributory factor allocated to pedal cyclists in fatal or serious accidents (FSA) with another vehicle was ‘Driver or rider failed to look properly’. It was also the most common factor allocated to the other vehicles involved

 

[mjr]

Thanks for spotting that release and posting it here.

I feel the "failed to look properly" wasn't worth the added bold. It mainly signifies that investigator believes the collision could have been avoided if either or both had noticed the other. That's pretty common and I think could be expected.

The bit about most fatalities not being at junctions but serious injuries are, that seems unsurprising too because average heavy vehicle speeds are hopefully lower at junctions, so less likely to kill. It's interesting in "Table 4" that T/Y/staggers account for the next highest percentage, then crossroads and only then roundabouts, despite how much cyclists worry about them.

Also, about 50% more drivers than cyclists involved in KSI (Killed or Seriously Injured) collisions were thought to misjudge another persons's path or speed, over 50% more were "careless, reckless or in a hurry" and over 100% more did a "poor turn or manoeuvre".

The "rural roads" being overly lethal looks worrying. It would be interesting to break that down by road class once someone gets their teeth into the full dataset, to see if that's country lanes (C/U) or primary (A) roads. I expect it's probably mostly 60+mph limit roads.

The 83% KSIs being male looks bad, but last data I saw said 66% of cyclists were male, so it's not quite as bad as it looks. Still curious.

And staying on curiousities, the graph in "Chart 1" with the traffic line says "Source: STATS19" but STATS19 is casualty data and did not contain traffic data when I last saw it, so I suspect they omitted a source there.

While they point out that cars are involved in more cycling fatalities, I note further down it says "the highest proportion of casualties that are fatal occur in 2-vehicle accidents involving an HGV (6.1%)" suggesting HGVs are still disproportionately dangerous and more are killed in car collisions maybe because cars are much more numerous.

The "time of day of collisions" is not very informative because it is not scaled by any traffic volume (either cycling or all).

It's also good to remember that this is data collected by humans who sometimes misclassify things, which may explain the 63 non-cyclists who were found to have contributed to collsions by being "cyclist entering road from pavement".

Fascinating. Let's see what other people notice in it...

 

[presta]

Spot the risk compensation. People drive more carefully on urban roads because they look more dangerous.

 

[mjr]

Spot the risk compensation. People drive more carefully on urban roads because they look more dangerous.
View attachment 611658

I don't understand. If that's so, why does 71% of traffic on urban roads produce 81% of casualties despite presumably lower impact speeds?

 

[Dolorous Edd]

• a majority of pedal cycle fatalities (59%) do not occur at or within 20m of a junction

This would be more meaningful if we understood what proportion of miles (or hours) cycled were at or near 20m of a junction.

For example, if the figure is 20%, that would make those segments, accounting for 41% of cycle fatalities, twice as dangerous as non-junction segments.

 

[HMS_Dave]

I don't understand. If that's so, why does 71% of traffic on urban roads produce 81% of casualties despite presumably lower impact speeds?

Because lower impact speeds have little bearing on the casualty rate, only the severity of it but significantly increased traffic likely does. I would also agree with Presta's assessment.

 

[mjr]

Because lower impact speeds have little bearing on the casualty rate, only the severity of it but significantly increased traffic likely does. I would also agree with Presta's assessment.

I suggest lower speeds makes it easier to drive carefully.

But even without that, 71% of traffic is producing 81% of casualties, whereas wouldn't we expect more careful driving to result in less than 71% of casualties?

 

[mjr]

This would be more meaningful if we understood what proportion of miles (or hours) cycled were at or near 20m of a junction.

For example, if the figure is 20%, that would make those segments, accounting for 41% of cycle fatalities, twice as dangerous as non-junction segments.

To put the claim in a bit of context, it used to be that 75% of cyclist casualties were within 20m of a junction. That was a big argument against the sort of shoot cycleways that councils used to build that vanished at junctions, doing only the uncontroversial easy straightforward bits where drivers were happy to see 20mph cyclists out of the way of their 60mph motorised tanks.

 

[Arjimlad]

Thanks for the analysis of these figures.

Am I correct in reading Table 3, that of the 106,370 total casualties, only 4885 had no other vehicle involved, and that motor vehicles were involved in (106,370 less 5,800) 100,570 of the pedal cycle casualty cases - that's 94.5%

 

[mjr]

Thanks for the analysis of these figures.

Am I correct in reading Table 3, that of the 106,370 total casualties, only 4885 had no other vehicle involved, and that motor vehicles were involved in (106,370 less 5,800) 100,570 of the pedal cycle casualty cases - that's 94.5%

Yes, I think so at a glance, but remember that only motorists were legally required to report injury collisions to this data system, so an unknown number of cyclist-only, cyclist-cyclist and cyclist-walker collisions are not included here.