## Road versus rail - Casuatay Costs

Expand Messages
• Road versus rail - Facts Sheet number 4 - Casualty cost comparisons The data in the tabulation below provides the casualty costs per billion passenger-km by
Message 1 of 4 , Aug 8, 2002
• 0 Attachment

Road versus rail - Facts Sheet number 4 - Casualty cost comparisons

The data in the tabulation below provides the casualty costs per billion passenger-km by rail and non-urban bus.

The cash values are at the 2000 price base.  They have been calculated by applying the Government's values for roads of £922,874 for a fatality, £105,200 for a serious injury and £8,145 for a slight injury (at 2000 prices) to estimates of the casualty numbers and by dividing the result by the passenger-km.

As noted in my earlier e-mail rail records non-fatal casualties for passengers only when the casualty is taken directly to hospital from the scene of the accident.  Rail defines that class of casualty as "injured" although the numbers then appears in Transport Statistics Great Britain under the heading "minor" so causing confusion. Within the table below the costs attributable the "injured" class of rail casualties are under the heading "serious" and the unit cost applied to the corresponding source numbers is the "serious" road casualty cost.  On the face of it that will have underestimated the cost of rail casualties because, typically, a minor rail casualty is a more seriously damaged person than a "serious" road casualty (meaning overnight in hospital or at the discretion of the police).

"Slight" rail casualty costs in the table have been estimated by first estimating the number of slight casualties, not recorded within the source data.  To do that I assumed that the ratio of serious to slight casualties by bus (8.7) would apply to all classes of rail casualty.  The assumption may well be far from the truth for some classes of accident but, in the absence of other data, the estimates are described here as being fair.

Additionally, to allow for the differences in definition for deaths (died within 30 days for a road casualty, but within one year for rail) the rail fatalities have been reduced by 5% and the 5% has been added to the so-called minor rail casualties.

The passenger-km use to estimate the bus casualty rates and costs were obtained by multiplying the bus-km by an occupancy of 16. 16 is 20% below the value of 20 appropriate to intercity coaches but over 50% above the average of 9 for all buses and coaches.

Table 1 Casualty costs to passengers £(000)s per billion passenger km, 2000 base
(Period covered is 1996 to 2000)

Accident type                                              Itemised                        Cumulative totals
Fatal "Serious" Slight  Total     Fatal "Serious"  Slight   Total
RAIL PASSENGERS
(1) Train (in-vehicle).                     177     369       248      794       177       369         248      794
(2) Falling from trains. (in-veh)         59        9           6        74       236       378         254      868
(3) board/alight.                             15     575       388       978       251       953         642   1,846
(4) Doors                                        0     175       118       293       251    1,128         760   2,139
(5) Falling from platforms and
hit by a train.       118      22         15       155       369     1,150        775    2,294
(6) Falling from platforms and
not hit by a train.          0      87         59       146       369     1,237        834    2,440
(7) Other.                                      15    397        267      679       384     1,634     1,101    3,119
(8) Crossing the lines.                    25       4            2        31       409     1,638     1,103    3,150
Non-urban BUS PASSENGERS
(a) In-vehicle, "like for like" *            97     97         66       260          97         97         66      260
(b) In-vehicle, due to other  veh *    226   228        154       608        323       325       220       868
(d) As pedestrians                         26     18            5        49         349      343       225       917

Inspection of the last column in the table shows that rail imposes upon passengers about 3 times the cost per passenger-km imposed by buses on non-urban roads.

Other comment:

(a) The cost of fatalities in train accidents is 3.7% of total rail casualty costs.
(b) The cost of all rail fatalities is 13% of total rail casualty costs.
(c) Options for sensitivity tests include:
(i) doubling the cost of "serious" rail casualties on the basis that "taken to hospital (by ambulance)" is a more serious class of injury than the serious road injury, namely spent the night in hospital.
(ii) Halving the costs of slight rail casualties on the basis that little is known about these e.g. the ratio of slight to serious casualties for buses, used in calculations, is 6709/771 = 8.7, but the corresponding ratio for pedestrians injured by buses is 14,562/4,151 = 3.51, suggesting a wide range for the ratios applicable to rail casualties.
(iii) Increasing bus occupancy from 16 to 20.  That would increase estimated passenger-km by bus by 20% and hence reduce the costs per passenger-km by the same amount.

The above is of interest when comparing casualty costs to passengers.   However if we wish to compare on a system wide basis then staff and trespassers by rail and all pedestrians etc. by road need to be included.

Currently my calculations show that system-wide the casualty cost per billion passenger-km are:

National Rail  ……………………………………………..........……  £ 7.3 million
All roads including pedestrian, cyclist and motor bike casualties  £14.2 million
All roads excluding pedestrians etc                                            £  8.3 million

Pedestrian, cyclist and motor bikers are substantially absent from railways and most roads have non-engineered alignments whereas rail alignment are of course engineered.  Hence if the rail function were carried out by rubber tyred vehicles occupying the rights of way used by trains the casualty costs would probably fall far below that provided by the steel tyred option.

Paul F Withrington August 2002
• I think it is critical as transport proponents to try and assess these kind of statistics with as little bias as possible, so I d like to commend Paul for
Message 2 of 4 , Aug 12, 2002
• 0 Attachment
I think it is critical as transport proponents to try and assess these kind
of statistics with as little bias as possible, so I'd like to commend Paul
for trying. I think this exercise has been useful for demonstrating why a
radical review is needed to try and get some consistency between road and
rail casualty definitions and figures. This will always be difficult, owing
to variations in the actual environment of road and rail travel, but more
consistency should be possible.

Unfortunately, statistical comparisons like Paul's built on a large number
of assumptions and comparing radically different figures and situations are
pointless. It's not enough to say "These are the comparisons given x, y and
z, and therefore these assumptions have to be bourne in mind while
interpreting the data". Fundamentally you are comparing two completely
different situations using a set of untenable assumptions.

In his first set of figures Paul has compared ALL rail figures versus
NON-URBAN bus figures. Is there an explanation? Secondly, Paul compares
bus/coach with rail. Buses make up a small percentage of vehicles on the
road, but this is compared with ALL rail deaths, which as far as I'm aware
extends to freight. Many of the statistical assumptions are unjustified in
their magnitude.

From what is then, at best, incomparable situations, Paul has then drawn a
total non-sequiteur:

"Pedestrian, cyclist and motor bikers are substantially absent from
railways and most roads have non-engineered alignments whereas rail
alignment are of course engineered. Hence if the rail function were
carried out by rubber tyred vehicles occupying the rights of way used by
trains the casualty costs would probably fall far below that provided by
the steel tyred option."

That is not an argument against rail- it's an argument for giving trains
rubber wheels. It's not correct to say that pedestrians are absent from
railways- they are just concentrated in a completely different manner-
hence the large number of "falling in front of a train" fatalities.

The problems currently facing railways in Britain are more to do with the
particular circumstances of rail in Britain than inherent in rail, as can
be easily experienced if you travel on rail in other European states.
Specifically, the problems with rail originate with systematic
under-funding of the railways (partially owing to its strong
labour/unionist history) and over-funding of roads, as part of the classic
"rail is an expensive subsidy, roads are public investment" concept.

Improving the rail system would reduce the amount of traffic on roads,
ironically reducing the total amount of road casualties, and make safer and
more pleasant urban environments for people. The reason you have pedestrian
enough to allow pedestrians and cyclists to share their space. It is
rail, and therefore have a much greater area over which to injure pedestrians.

Even were you to have accurate data from which to work, and you treated
that data correctly, simply comparing injuries per billion passenger
kilometers does not give you a full picture of the complexity of the issue
which is system wide, not per kilometer. Rail works by concentrating travel
down narrow corridors, and is fed by pedestrians, bicycles, bus and car.
The nearest analogy is the motorway, but that is fed nearly exclusively by
cars, causing a number of safety issues on a network level.

This, of course, completely ignores the issue of air pollution. Air
pollution from cars is estimated in several European countries and
Australian cities to cause more fatalities than road accidents. I suspect
the discrepancy is far higher in many majority world countries (what other
people choose to call less-developed nations) such as India and Mexico
where air pollution is more serious. If we included the deaths from air
pollution causes in Paul's analysis we would see deaths per kilometer for

In conclusion, even if Paul's figures weren't fundamentally flawed due to
what he was comparing and hadn't taken unjustifiable assumptions he has
left out critical aspects of the problem which would render his conclusion
actually reversed if he had incorporated them.
Thanks

Robert Murray-Leach

Robert Murray-Leach
Green Transport Officer
The Conservation Council of South Australia
120 Wakefield St
Tel. (08) 8223 5155
Fax. (08) 8232 4782
E-mail. smogs@...
• In reply to Robert Murray-Leach - I collect such E-mails so as to be able to republish as examples of the comment that railway people tend to make whenever
Message 3 of 4 , Aug 14, 2002
• 0 Attachment
In reply to Robert Murray-Leach -

I collect such E-mails so as to be able to republish as examples of the comment that railway people tend to make whenever rail is criticised.

- Mr Leach says I have incorrect data without saying which and where and that I have made untenable assumptions again without saying which or where so:
1    My source data is the national statistics
2    We assume the laws of arithmetic have not been suspended
3    I divide casualty costs buy passenger-km to find casualty costs per passenger-km
4    Within the calculations I assumed that the average occupancy of a bus/coach on non-urban roads was 16.  That compares with the national average for all buses of 9 and with the occupancy of 20 routinely achieved by intercity coaches.  It is easy for the reader to factor my results accordingly but, because the data is so devastatingly in favour of the bus, the conclusions will not change.
5    My other most important assumptions are:
(a)    an "injured" passenger by rail (= taken to hospital (by ambulance) directly from the scene of the accident) has the same cost as a "serious" road casualty (= spent a night in hospital).
(b)    A rough estimate of slight rail casualties can be made by reference to the ratio of slight to serious bus casualties.

The layout of the results in Facts Sheet 4 will enable the reader to carry out sensitivity tests on different assumptions but I believe my assumptions are, if anything, in favour of rail e.g. probably taken to hospital by ambulance is a more costly injury that spent the night in hospital.

Most of my data relates to buses on non-urban roads since, if the railways were dug up and replaced by roads, the buses on the railway alignments would approximate to buses on non-urban roads except that they would be safer and faster due to the better alignments, and widely spaced junctions.

Then Mr Murray-Leach says that rail is "equitable " compared with road overlooking the fact that generally only the rich can afford to travel by rail and that, when they do, they are heavily subsidized.  In the real world the poor scarcely ever use rail.  Instead they use buses and cars.

For amusement I annotate Mr Murray-Leach's below:

Paul Withrington

PS If any person wants the original text underlying the last Facts Sheet I can provide either as an AmiPro attachment or with some considerable difficulty as a Word one or by slow mail.

Paul W
----- Original Message -----
From: Robert Murray-Leach <smogs@...>
Sent: Monday, August 12, 2002 8:59 AM
Subject: Re: [World-Focus] Road versus rail - Casualty Costs

> I think it is critical as transport proponents to try and assess these kind of statistics with as little bias as possible, so I'd like to commend Paul for trying. I think this exercise has been useful for demonstrating why a radical review is needed to try and get some consistency between road and rail casualty definitions and figures. This will always be difficult, owing to variations in the actual environment of road and rail travel, but more consistency should be possible.
>
> Unfortunately, statistical comparisons like Paul's built on a large
number of assumptions and comparing radically different figures and situations are pointless. It's not enough to say "These are the comparisons given x, y and z, and therefore these assumptions have to be borne in mind while interpreting the data". Fundamentally you are comparing two completely different situations using a set of untenable assumptions. There are few assumptions and all of them are, if anything biased in favour of rail.  Additionally the source provides or suggests sensitivity tests (Available as an AmiPro attachment or by slow mail or as a Word attachment but with all the paragraph numbers lost).

> In his first set of figures Paul has
compared ALL rail figures versus NON-URBAN bus figures. Is there an explanation? Yes  - I am interested in Railway Conversion.  If the railways were converted to roads the rail passengers would be sitting in express coaches at one quarter the cost and with 4 times the capacity.  A proxy for the casualty cost they would then suffer is the cost suffered by passengers in buses and coaches on non-urban roads - one third that suffered by their compatriots sitting in trains or waiting on platforms.

i.e. , Secondly, Paul compares bus/coach with rail. Buses make up a small percentage of vehicles on the road, but this is compared with ALL rail deaths, which as far as I'm aware  extends to freight. In 1994 - the last year for which data is available - freight train-km amounted to just 10% of all train-km Many of the statistical assumptions are unjustified in their magnitude. Which ones please? I really do object to generalised and unsubstantiated comments such as these.

From what is then, at best, incomparable situations, Paul has then drawn a total non-sequiteur:

"Pedestrian, cyclist and motor bikers are substantially absent from railways and most roads have non-engineered alignments whereas rail alignment are of course engineered.  Hence if the rail function were carried out by rubber tyred vehicles occupying the rights of way used by  trains the casualty costs would probably fall far below that provided by the steel tyred option."  There is no non sequitor here - the point is  - if the rail function were carried out by rubber tyred vehicles using the rights of way taken by railways then pedestrians, cyclists and motorbikes would be absent from those rights of way and the rights of way would have better alignments than existing roads.  If the rights of way were restricted to freight and buses/coaches then flows would be ever so light in highway terms so eliminating accidents caused by cars and improving the safety comparison with rail.>

> That is not an argument against rail -
it's an argument for giving trains rubber wheels. Trains with rubber tyres = road vehicles It's not correct to say that pedestrians are absent from  railways- they are just concentrated in a completely different manner - hence the large number of "falling in front of a train" fatalities.  Pedestrians are, of course, almost entirely absent from the rights of way used by rail and would remain absent if the rights of way were converted to roads.
>
> The problems currently
facing railways in Britain are more to do with the  particular circumstances of rail in Britain than inherent in rail, as can be easily experienced if you travel on rail in other European states. Specifically, the problems with rail originate with systematic under-funding of the railways (partially owing to its strong labour/unionist history) and over-funding of roads, as part of the classic "rail is an expensive subsidy, roads are public investment" concept.  I reject all this absolutely - as will be seen from my later e-mails rail is at least 4 times as expensive and has one quarter the capacity of the rubber tyred alternative.  (E.g. (a) the cost of the West Coast Main Line Modernisation Programme is to cost £13 billion concentrated on just 1000 km on track.  In comparison the M1, with 1800 km of track, has a replacement cost including land of £2.1 billion: (b) At Waterloo 50,000 crushed passengers alight in the peak hour.  They could all sit down in 1000 50-seat coaches requiring one lane of a motor road.  At Waterloo there is room for 3 to 4 lanes in each direction.  The waste is lamentable)

Improving the rail system would reduce the amount of traffic on roads, This is a silly myth - Rail generally serves city centres where cars can go with difficulty.  In any case the Government's hope is to increase rail usage by 50% i.e. from 6% to 9% of present passenger-km.  If half the increase is to come from people otherwise travelling by car then car use in the next 10 years will rise to 118% of today's value instead of to 120% - The rail modernisation is to cost £73 billion.  The effect will be quite trivial.
ironically reducing the total amount of road casualties, and make safer and  more pleasant urban environments for people. The reason you have pedestrian fatalities on roads are not because, unlike rail, roads are egalitarian Only the rich can afford to travel by rail - the poor go by road.  So, to suggest rail is egalitarian compared with road is ......enough to make me despair of this commentator allow pedestrians and cyclists to share their space. It is because roads have invaded pedestrian space much more significantly than rail, and therefore have a much greater area over which to injure pedestrians.  So get the traffic off the historic roads onto rail alignments by taking the railway lines away and giving them to China.

Even This word "Even" is misplaced - my data is the national data were you to have accurate data from which to work, and you treated  that data correctly, simply comparing injuries per billion passenger kilometres does not give you a full picture of the complexity of the issue which is system wide, not per kilometre. Rail works by concentrating travel down narrow corridors, and is fed by pedestrians, bicycles, bus and car. I protest - the way to compare systems is to compare deaths or casualty costs per passenger-km.  If any one wants to see the detail I will send them copy

The nearest analogy is the motorway, but that is fed nearly exclusively by cars, causing a number of safety issues on a network level.

This, of course, completely ignores the issue of air pollution. Air pollution from cars is estimated in several European countries and Australian cities to cause more fatalities than road accidents. I suspect the discrepancy is far higher in many majority world countries (what other people choose to call less-developed nations) such as India and Mexico where air pollution is more serious. If we included the deaths from air pollution causes in Paul's analysis we would see deaths per kilometre for total road at least double.  Calculations of mine yet to be posted show that the entire rail function, if carried out by buses and lorries on the alignments occupied by rail, would be discharged using 35% to 50% less energy than does rail.  A passenger train is no more fuel efficient that a diesel powered car containing two people.  Express coaches would use half to one third the fuel required by trains.

In conclusion, even if Paul's figures weren't fundamentally flawed I object to this patronising tone - there is no "even" about it nor are my figures "fundamentally flawed" unless the government's data is itself failing or the rules of arithmetic have changed due to what he was comparing and hadn't taken unjustifiable assumptions Please say which and in what way my assumptions are "unjustifiable" he has left out critical aspects of the problem (Which ones please) which would render his conclusion actually reversed if he had incorporated them.

Please address my numbers one by one otherwise the comment can have no value.

Paul W

> Thanks
>
>
> Robert
Murray-Leach
>
>
> Robert Murray-Leach
> Green
Transport Officer
> The Conservation Council of South Australia
>
120 Wakefield St
> Tel.  (08) 8223
5155
> Fax. (08) 8232 4782
> E-mail.
href="mailto:smogs@...">smogs@...
>
>
>
>
href="http://us.click.yahoo.com/pt6YBB/NXiEAA/RN.GAA/KlSolB/TM">http://us.click.yahoo.com/pt6YBB/NXiEAA/RN.GAA/KlSolB/TM
>
---------------------------------------------------------------------~->
>
> Post message:
href="mailto:WorldTransport-Focus@yahoogroups.com">WorldTransport-Focus@yahoogroups.com
> Subscribe:
href="mailto:WorldTransport-Focus-subscribe@yahoogroups.com">WorldTransport-Focus-subscribe@yahoogroups.com
> Unsubscribe email to:
href="mailto:WorldTransport-Focus-unsubscribe@yahoogroups.com">WorldTransport-Focus-unsubscribe@yahoogroups.com
>
Questions to: WorldTransport-Focus-owner@yahoogroups.com
>
>
>
>
> Your use of Yahoo! Groups is subject
to http://docs.yahoo.com/info/terms/
>
>
>
• In reply to Paul Withrington, Hi Paul, Just to clarify my position, which was misunderstood. My problem is not with the original statistics (the national
Message 4 of 4 , Aug 15, 2002
• 0 Attachment

Hi Paul,

Just to clarify my position, which was misunderstood. My problem is not
with the original statistics (the national statistics) but the conclusions
that:
a) have been drawn from them
b) more importantly, would be possible to draw from them
I'll focus specifically on casualties as an example.

If you compare all casualties from the existing rail system with only those
from the extra-urban bus system, then that is what you are comparing. You
are not comparing casualties from trains on rail to casualties from buses
on roads set aside exclusively for buses. You're familiar with the national
statistics, so you know that a significant proportion of rail casualties
arise from intra-urban travel. As you point out, casualties during rail
travel are outnumbered significantly by casualties getting on and off
trains and while at the platform. Therefore short train journeys (mainly
intra-urban) are likely to have a larger number of casualties per kilometre
travelled than long journeys (mainly extra-urban).

When you compare total rail (including long and short distance journeys) to
long distance journeys only bus you end up biasing the data. There is
therefore a hidden assumption in your analysis- that all rail and bus
journeys, irrespective of length, will have an equal risk of casualties per
km travelled.

This kind of hidden assumption is very difficult to guard against in
statistics, and make it extremely difficult to accurately compare aspects
of train and bus travel. In this case, I would be reluctant to draw a
conclusion either way from the comparison of these two figures.

This is not to say that I buses might not be good fulfilling a similar
function to rail- there is an "TransMilenio" bus system running in Bogota
very similar to the system you're proposing Paul (although you're probably
aware of it). Neither am I a diehard rail advocate- I was just pointing out
some problems specifically with your analysis.

I definately think public transport is more equitable than private
vehicles, but I have yet to be convinced that either rail or bus on
dedicated lanes are universally better than the other. I'm afraid that I
didn't say that rail was more equitable than bus. What I said was:
"The reason you have pedestrian fatalities on roads are not because, unlike
rail, roads are egalitarian enough allow pedestrians and cyclists to share
their space." Sorry if that wasn't clear.

Cheers

Robert Murray-Leach

Robert Murray-Leach
Green Transport Officer
The Conservation Council of South Australia
120 Wakefield St