Road Safety Authority: 2017 Road Deaths 186->158 but Cyclists 10->15

tomasrojo · 03-01-2018 6:04pm

CramCycle wrote: »

Even with the that nice graph, I think it is missing the number of road users, number of cars on the road and other such figures that would expand upon the data. I think the author says this in reply to a comment further down. Seat belts look like they do nothing to death rates, and may in fact be a hindrance to lowering death rates. This may also be explainable by an increase in the number of drivers, the much loved one of drivers felt safe therefore drove more dangerously, although I think the first option is the most plausible, or a mix of the two.

A far more interesting stat, maybe its there and I missed it, but would indicate for or against the second suggestion is the number of pedestrian and cyclist road deaths from this point on.

Ah sure, it's not a detailed analysis, but the 2000+ lives saved per year claim (or 60,000 lives over 25 years, as it's often put) isn't based on detailed analysis either. Nobody seems to know where it came from, in fact, though it's quote extensively. Anyone who listens to More or Less on BBC Radio will recognise this as a perennial characteristic of publicly quoted statistics.

He does look elsewhere at the trend in cyclist and pedestrian fatalities before and after the mandatory seat belt law.
http://www.john-adams.co.uk/2013/03/24/the-biggest-lie/

Significantly the Significance article did not make it into the Review’s list of key references on seat belts. A significant omission because the authors, all defenders of the seat belt law, acknowledge an effect of the law of important consequence to vulnerable road users. They say “the clear reduction in death and injury to car occupants is appreciably offset by extra deaths among pedestrians and cyclists.”

This will probably derail this thread if discussed in any detail though. I'm not sure whether it still applies anyway, but in the years immediately after the law, it seems plausible enough that some of the proponents of the UK seat belt law acknowledge a rise injuries and deaths among pedestrians and cyclists as having probably been an unwelcome consequence.

(Adams lets his libertarian beliefs colour his analysis in some important areas, so some caution is required with him, I feel now. But the Significance article is real enough, and he didn't write it.)

GreeBo · 03-01-2018 6:15pm

tomasrojo wrote: »

(Adams lets his libertarian beliefs colour his analysis in some important areas, so some caution is required with him, I feel now. But the Significance article is real enough, and he didn't write it.)

What was the evidence of correlation between seatbelts and increase in non-vehicular accidents?

He lost me after this stunning piece of analysis...

60,000 divided by 25 years would equal 2400 lives saved every year since 1983 by the seat belt law. In 1982, the year before the seat belt law came into effect there were only 2365 driver and front seat passenger fatalities!

tomasrojo · 03-01-2018 6:19pm

I guess the important distinction is lives saved by seat belts and lives saved by the seat belt law. In the latter case, the number of deaths of occupants of the front of cars should have dropped to near zero quickly after the law if the law itself were as effective as claimed.

In the former case, it's more complicated because there were plenty of seat belt wearers before the law.

tomasrojo · 03-01-2018 6:29pm

On what evidence there was for the rise in pedestrian and cyclist fatalities, it's something like this:

There's a clear uptick in fatalities (combined in that ratio with an overall downtick for car occupants) coinciding with the law, and the previous downward trend continues from that point, rather than the point before the law.
And in raw numbers:

tomasrojo · 03-01-2018 6:33pm

The article in Significance would be better evidence, as it's a peer-reviewed paper, but it's behind a registration wall.

GreeBo · 03-01-2018 7:08pm

tomasrojo wrote: »

On what evidence there was for the rise in pedestrian and cyclist fatalities, it's something like this:

There's a clear uptick in fatalities (combined in that ratio with an overall downtick for car occupants) coinciding with the law, and the previous downward trend continues from that point, rather than the point before the law.
And in raw numbers:

I dont doubt the raw numbers, Im just struggling to understand the correlation with seat belts.

tomasrojo · 03-01-2018 7:39pm

GreeBo wrote: »

I dont doubt the raw numbers, Im just struggling to understand the correlation with seat belts.

1982 was the year that legal penalties for not wearing a seat belt in the front of car were brought in in the UK.

As I said, there were plenty of people wearing seat belts before that date, and promotion had begun years before. Adams' point, I think, is mostly what happens when the final tranche of the population, who wouldn't voluntarily wear a seat belt, are forced to do so. And it seems, though I'm not sure how robust the statistical analysis is, that the increase in safety (real or perceived or both) is offset by more risk-taking, with cyclists and pedestrians feeling the consequences.

tomasrojo · 03-01-2018 7:57pm

The other point, which I think is a fair point, is that there's no basis for saying that the seat belt law saved 60,000 lives over 25 years, or maybe indeed no basis for saying that seat belts themselves saved that many lives.

The first claim is close to impossible, given that there were only 2000 or so lives that could be saved that way in 1982, when the law came into force.

The second claim is definitely possible (seat belts were pretty widely in use before the law, so it's plausible they were saving thousands of lives per year before 1982 and those saved lives were already bringing the total down to 2000 or so), but nobody can explain how they did this calculation -- or even claims that they tried.

GreeBo · 03-01-2018 8:07pm

tomasrojo wrote: »

1982 was the year that legal penalties for not wearing a seat belt in the front of car were brought in in the UK.

As I said, there were plenty of people wearing seat belts before that date, and promotion had begun years before. Adams' point, I think, is mostly what happens when the final tranche of the population, who wouldn't voluntarily wear a seat belt, are forced to do so. And it seems, though I'm not sure how robust the statistical analysis is, that the increase in safety (real or perceived or both) is offset by more risk-taking, with cyclists and pedestrians feeling the consequences.

Was there an equal increase in the stats for car crashes then?

Also I think there is some confusion about seat belts preventing deaths versus seatbelts preventing crashes.

I think its pretty easy to demonstrate the difference between crashing with and without wearing a seat belt.

I have a really hard time accepting the "increase in risk taking", for any of the arguments its ascribed to (seat belts, helmets, lights, etc)

tomasrojo · 03-01-2018 8:27pm

Yes, Adams accepts that you're more likely to survive a crash wearing a seat belt.

I'm not totally on board with Adams' outlook, as I said. I actually do think risk compensation is a real phenomenon. But it's not a slam-dunk argument against the use of precautions, or even the imposition of mandatory use of precautions (though the precautions have to be very good, and the attendant risk has to be pretty big).

GreeBo · 03-01-2018 8:31pm

tomasrojo wrote: »

Yes, Adams accepts that you're more likely to survive a crash wearing a seat belt.

I'm not totally on board with Adams' outlook, as I said. I actually do think risk compensation is a real phenomenon. But it's not a slam-dunk argument against the use of precautions, or even the mandation of the use of precautions (though the precautions have to be very good, and the attendant risk has to be pretty big).

I think it is a real phenomenon but not for run of the mill everyday activities.

Would I be more likely to drive aggressively if I was in a rally car complete with roll cage and 5 point harness, most definitely.

Do I cycle differently on days with my helmet versus without? I'd have to say no, it wouldnt even be a though in my head.

tomasrojo · 03-01-2018 8:40pm

It is supposed to be a largely unconscious process of compensation though.

They have done some studies (this is all off the top of my head, so may not be accurate) that showed people who usually wear helmets cycling more slowly when they didn't wear them, and some studies of taxi drivers driving faster when they were told they had ABS brakes compared with when they were told they didn't. They did a study of children running round an obstacle course, and the children ran more quickly when wearing protective equipment.

The effect isn't enormous, but I think there's reasonable evidence for it.

ted1 · 03-01-2018 9:22pm

Seaswimmer wrote: »

tomasrojo wrote: »

As I've said before, saying there's a 50% rise is true, but misleading, because it makes it sound worse than it is. Basically, the yearly total has gone up, but it's gone 12->10->15. Which is not the direction anyone would want, but it's all in the same ballpark.

Where it differs is from the 2007-2012 period, which was consistently low, even getting down to 5.

It's worth noting, and I do think the total may be on an upward trend, due to both higher numbers of cyclists, and more HGVs back on the road, but analysing small numbers is hard to come up with vey meaningful stuff, especially on small timescales.

Strangely the cycle counter numbers on the Rock Road were down this year.

2016 275,000
2017 262,000

I expected an increase in line with the perceived notion that numbers of cyclists are up generally so not sure why there was such a substantial drop this year.

Big increase over 2013 which was last full year count (approx 210,00) but I expected trend to be continually upwards.

It was out of order for a while

tomasrojo · 03-01-2018 11:38pm

Back on topic:
http://irishcycle.com/2018/01/03/cycling-deaths-2017/

Level-headed analysis of this subject.

1bryan · 04-01-2018 9:30am

GreeBo wrote: »

I suspect its an issue of "distinguish-ability" rather than visibility, but I take the point.

what an absolute nonsense. Pretty much every fatality I'm aware of that involved a motorist (which is pretty much every fatality there is), the cyclist was exactly where they were meant to be - ie: on the left of the road.

I think your comment about sums up the problem though. Motorists react, they don't anticipate. If they see a cyclist they will avoid the cyclist. But what they should be doing is respecting the part of the road where cyclists are, AT ALL TIMES.

1bryan · 04-01-2018 9:31am

This is interesting. It's Florida, rather than Dublin, but the findings ring true for here too.

https://usa.streetsblog.org/2018/01/03/study-cyclists-dont-break-traffic-laws-any-more-than-drivers-do/

Ferris · 04-01-2018 11:43am

1bryan wrote: »

This is interesting. It's Florida, rather than Dublin, but the findings ring true for here too.

https://usa.streetsblog.org/2018/01/03/study-cyclists-dont-break-traffic-laws-any-more-than-drivers-do/

I read the data analysis section of that study and they haven't for bias on the cyclists behaviour due to being monitored? Maybe cynical but I would be a good boy too if I was being monitored by sensors, cameras, and a GPS. Driver behaviour would improve too under similar scrutiny.

Not to mention that they performed all analysis based on a 90% confidence interval which I think is insufficient when you have good quality data, in my industry 95% would be the norm.

Shame that these aspects have not been considered as I believe studies like these hold the key to making good value infrastructure changes rather than looking at the high level cyclist mortality rate.

GreeBo · 04-01-2018 12:16pm

1bryan wrote: »

what an absolute nonsense. Pretty much every fatality I'm aware of that involved a motorist (which is pretty much every fatality there is), the cyclist was exactly where they were meant to be - ie: on the left of the road.

I think your comment about sums up the problem though. Motorists react, they don't anticipate. If they see a cyclist they will avoid the cyclist. But what they should be doing is respecting the part of the road where cyclists are, AT ALL TIMES.

I dont think I said that cyclist was anywhere other than where they should be?

My point is that you can be perfectly visible in a lab setting, but not distinguishable from pedestrians, buggies, displays on shops, parked cars, etc etc that you find in the real world.

I didnt make any attempt to blame anyone, yet you jump straight into the usual, angry blame motorists response.

BTW, if it was that easy then we wouldnt have car crashes either and not crashing into a car is to both drivers benefits and yet it still happens every day.

GreeBo · 04-01-2018 12:20pm

tomasrojo wrote: »

Back on topic:
http://irishcycle.com/2018/01/03/cycling-deaths-2017/

Level-headed analysis of this subject.

I wouldnt agree its level headed at all, its (obviously biased towards cyclists) its title gives that away immediately.

But also points such as

Significantly, all of the 15 collisions in 2017 included the involvement of motorists — that compares to the 2016 and 2015

when we almost all agree that there isnt enough data to draw any significant conclusions.

I might as well say thats its significant that when tossing a coin I got HEADS 15 times compared to last year where it was only 10.

1bryan · 04-01-2018 12:49pm

GreeBo wrote: »

My point is that you can be perfectly visible in a lab setting, but not distinguishable from pedestrians, buggies, displays on shops, parked cars, etc etc that you find in the real world.

my point is that your point is irrelevant. Someone offering the excuse, 'I didn't see them', when they hit a cyclist is completely unacceptable because, by-in-large cyclists occupy the part of the road that they're supposed to. If that part of the road was respected at all times, and not just when motorists 'see' a cyclist, it wouldn't even be an issue.

1bryan · 04-01-2018 12:49pm

GreeBo wrote: »

I didnt make any attempt to blame anyone, yet you jump straight into the usual, angry blame motorists response.

yes, motorists who are responsible for pretty much all cyclist fatalities.

GreeBo · 04-01-2018 1:03pm

1bryan wrote: »

my point is that your point is irrelevant. Someone offering the excuse, 'I didn't see them', when they hit a cyclist is completely unacceptable because, by-in-large cyclists occupy the part of the road that they're supposed to. If that part of the road was respected at all times, and not just when motorists 'see' a cyclist, it wouldn't even be an issue.

Any my point is that if it was that simple,you wouldnt have head on collisions or cars side-swiping each other as they would be respecting other lanes.

You need to live in reality and accept that these things happen, rightly or wrongly.

I also didnt say it was an acceptable excuse, I provided it as a reason, since, most sane people dont drive into other humans on purpose, even if they are only cyclists.

1bryan · 04-01-2018 1:16pm

GreeBo wrote: »

even if they are only cyclists.

I figured you were trolling all along. That comment confirms it.

beauf · 04-01-2018 1:19pm

GreeBo wrote: »

I suspect its an issue of "distinguish-ability" rather than visibility, but I take the point.

I don't think its either.

I'd suggest our driving habits are not to look for cyclists. But is probably also true of car vs car accident. We are reckless because there is weak enforcement of dangerous driving. You see it daily on the M50 (but all over to be honest) just really dangerously bad driving. People would be more careful driving if they thought they would be more likely to be caught and punished.

You can be lit up like a xmas tree. But if someone is looking at their phone, or just not looking at all, they won't see you.

CramCycle · 04-01-2018 1:29pm

1bryan wrote: »

I figured you were trolling all along. That comment confirms it.

MOD VOICE: If you think that someone is trolling or breaching other site rules, use the report post button. DO NOT CALL THEM OUT IN THREAD. Any questions via PM

tomasrojo · 04-01-2018 1:32pm

GreeBo wrote: »

I wouldnt agree its level headed at all, its (obviously biased towards cyclists) its title gives that away immediately.

I was thinking most specifically about him putting the 50% number in context, and looking at longer-term trends. Most of the cycling campaigners (the ones I'm aware of anyway) are making more of the 50% than it warrants, which is a pretty standard thing to do when you're essentially involved in politics, but I admire people who don't do it.

GreeBo wrote: »

But also points such as
when we almost all agree that there isnt enough data to draw any significant conclusions.

I don't think he was using "significantly" in its statistical sense, to be fair. But the majority (and sometimes, as this year, the totality) of cyclist fatalities are caused by collision with motorised vehicles every single year. It is beyond any uncertainty the leading cause of cyclist deaths in Ireland.

GreeBo wrote: »

I might as well say thats its significant that when tossing a coin I got HEADS 15 times compared to last year where it was only 10.

If every year you toss a coin you get 95-100% heads, then you know you've got a biased coin. We know that simple falls are not a large cause of cyclist deaths. We know beyond any doubt that the area we have to work on is reducing the likelihood of collisions with motorised vehicles.

1bryan · 04-01-2018 1:36pm

CramCycle wrote: »

MOD VOICE: If you think that someone is trolling or breaching other site rules, use the report post button. DO NOT CALL THEM OUT IN THREAD. Any questions via PM

done. Report sent.

Roadtoad · 05-01-2018 4:50pm

Of the 2017 fetalities, half the vehicles were 'commercial' (4x4, tractors, campervan, trucks), which seems out of kilter with the numbers of such vehicles on the roads.

Is there a driver gender count and age count? I'd prefer this to anecdotal opinion gleamed from reports at the times of the events.

CramCycle · 05-01-2018 5:44pm

It should be noted that this year, one of the fatalities involving a commercial vehicle and cyclist was the cyclists fault (the tourist), but none of the others appear to be. This was the only one where I know of the incident being the fault of the cyclist. The others (that I know of) where all the fault of the motorist involved.

GreeBo · 05-01-2018 6:13pm

CramCycle wrote: »

It should be noted that this year, one of the fatalities involving a commercial vehicle and cyclist was the cyclists fault (the tourist), but none of the others appear to be. This was the only one where I know of the incident being the fault of the cyclist. The others (that I know of) where all the fault of the motorist involved.

Were the details of the road sweeper incident in Rathfarnham released yet?

Road Safety Authority: 2017 Road Deaths 186->158 but Cyclists 10->15

Comments