Transport Blog

July 2006
Sun	Mon	Tue	Wed	Thu	Fri	Sat
						1
2	3	4	5	6	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30	31

March 08, 2004

Comparative safety statistics

Patrick Crozier | Air Safety | Best of Transport Blog | Rail Safety | Road Safety

I found this table on the PACTS (boo, hiss) web site:

Mode	Fatalities per billion passenger kilometres
Motor cycle/moped	112
Foot	48
Pedal cycle	33
Car	3
Van	0.9
Water	0.4
Bus or coach	0.2
Rail	0.1
Air	>0.01

First of all, it is perfectly possible to cast doubt on these numbers on the basis of a) their accuracy and b) their significance but let's pretend for the sake of argument that they are accurate and that they do represent the best measure for safety possible for there are several things that really strike me about these stats.

Me: safe and sound

The first thing is the enormous spread. You are 10,000 times more likely to die on a motorbike for every mile travelled than you are in a plane.

The second thing is: who would have thought that walking was so dangerous? I would have thought that pretty much the only way you can get killed when out walking is when crossing the road. OK, the vehicle could have mounted the pavement or failed to stop at a pedestrian crossing but that is pretty rare. In other words 9 times out of 10 it's your fault. And only a third of these accidents are children under the age of 16. Personally, I would like to know what the figures are for children under the age of 7. After that age most children have a pretty good idea that cars are dangerous. In other words we are pretty reckless with our own safety. More evidence perhaps that safety is not the only thing?

The other thing that strikes me (and has struck me before) is the inverse relationship between the cost of the seat and the danger involved. An aircraft seat costs a lot of money (I would guess about £150,000 (£60m/400 seats). A train seat about £14,000 (£1m for the carriage/70 seats). A car seat (assume one passenger) about £10,000. It doesn't quite work (the big exception being motorbikes) but it's not a bad rule of thumb.

Me: about to dice with death

Why is that? My guess it's because if you have forked out £60m for something you are going to look after it. Indeed, it would appear that you are going to look after it more than you are going to look after yourself. Which says something about the real value of human life.

It is also interesting that buses and coaches are 10 times safer than cars. This seems very odd as you would have thought that buses and coaches are just as likely to crash as cars but yet almost no one on board will be wearing a seat belt.

Update

Just one final thought. If the safety campaigners are as concerned about child safety as they say they are shouldn't they be campaigning to reduce the minimum age for a driving licence to, say, 5 years old and for the removal of all forms of motoring taxes to make it easier for children to drive which, after all, would be 16 times safer than walking?

Trackbacks

Blog of the week: TransportBlog
TransportBlog looks at transport policy from a free-market perspective, often putting forward radical solutions to the problems in the world's transport systems. It is edited by Patrick Crozier, Transport Spokesman of the Libertarian Alliance. Recent p...
Adam Smith Institute Blog on March 26, 2004

Comments

Wouldn't a better measure be fatalities per hour of travel, to correct for the varying speeds of each mode?

Posted by David on March 9, 2004

Two different modes: A and B. Exactly the same journey distance. Mode A takes half the time. Exactly the same number of journeys. Exactly the same number of fatalities. Add in the correction: Mode A is twice as dangerous as Mode B.

Doesn't sound right to me.

Posted by Patrick Crozier on March 9, 2004

I'm always bemused by these statistics - particularly as I am a motorcyclist. It's what they don't tell you that is interesting. For example, a significant proportion of motorcyclists do so for leisure and therefore ride fewer miles that do motorists. This skews the figures giving a misleadingly high accident statistic for those (like me) who ride throughout the year and cover average (and above) mileages.

Of course, the likleyhood of being killed or seriously injured on a a motorcycle in the event of an accident is going to be higher as there is no metal shield to protect the rider - but I am no more likely to be involved in one when I am riding that when I am driving.

I hate statistics with a vengeance - they get in the way of the truth and consequently rational management of the risk. Which your comment adroitly points out.

Posted by Mark Ellott on March 9, 2004

I was wondering the same thing as Mark above. I did a 47mile bicycle ride yesterday, had I crashed would it have counted? I wouldn't call my ride transport because I went out the front door did a large non-stop loop around the countryside and came straight back in again. I didn't go anywhere per-se.

What about yachting and boating people? I assume more leisure sailors drown than do eople falling off the Ostende ferry? And what about fishermen?

Too many questions!!

Posted by Mark Holland on March 9, 2004

From the point of view of trying to minimise fatalities, these statistics are irrelevant. The relevant numbers should be the marginal cost of saving a life in each mode of transport.

Googling for "marginal cost of saving a life" tantalisingly gives me this hit:

Accidental fatalities in transport - J Royal Statistical Soc A... prevented (Railway Safety, 2002). This is about 100 times greater than the marginal cost of saving a life on the roads. Even more costly ... www.blackwell-synergy.com/links/doi/ 10.1111/1467-985X.00274/abs/

but following the link just gives the abstract of an academic paper which would cost $25 to read.

If I get the time, I might have a trip to the library and photocopy the paper.

Posted by Andy Wood on March 9, 2004

As for whether it should be fatalities per hour or fatalities per mile, I think it depends on the circumstances. I want to go for a day out on a Sunday from my hypothetical home near Waterloo. I can go for a lengthy walk along the bank of the Thames, have lunch in a pub somewhere, and walk back. Or I can go for a drive to Brighton, have some fish and chips, and drive back. Or I can buy a day return to Paris on the Eurostar, have a really fine French meal, and return on the train that evening. In this instance, the number of miles I travel is heavily influenced by the time it takes to get there, and as a consequence deaths per hour is probably the appropriate safety metric.

On the other hand, I have to go to Paris for a business meeting. The number of miles I have to travel is fixed, and therefore deaths per mile is the appropriate safety metric, regardless of whether I go by air, train, coach/ferry or car/ferry.

The question is to what extent we are willing to substitue distance and time for one another when we travel. Which metric is appropriate varies in response to this quesiton.

Posted by Michael Jennings on March 9, 2004

I've never come across any stats, dodgy or otherwise, as to what kind of interactions were involved.
How many of the people who died on motorcycles, bicycles and their own two feet died under someone else's car?

Posted by neil on March 9, 2004

Neil, there are stats that suggest that the majority of motorcycle accidents involve another road user usurping the rider's road space. However, as to reliability and where to access them.....(RoSPA? TRL?)

I try not to worry too much about stats - I just get on with defensive riding ;-)

Posted by Mark Ellott on March 9, 2004

When reading safety stats one needs to bear in mind whether they are actually anwering the most pertinent question; "which modes *cause* the most danger?". Asking simply "who is in the most danger" seems pretty dumb.

The question as to whether modes should be assessed by time, distance or journey is interesting. Cars are relatively safe places to be (even if they make everywhere near them rather more dangerous places to be) only if KSI per kilometre is the measure.

Posted by Not Responding on March 13, 2004

Permalink

IN BRIEF

November 23, 2004

'Captain commuter' wins Sydney a free day on the trains ...link