Decades of propaganda have established the myth that railways cost much less, are far safer, have much more capacity, use less fuel, less manpower and are far kinder to the environment than road transport ever can be. Surprisingly the myth turns out to have no basis in fact.  Indeed the reality is so different from the myth and the propaganda so powerful, that people suddenly faced with the facts are usually unable to accept them, even though the data and the associated calculations are easy to check.

Here are some key comparisons:

(1) Government expenditure on rail divided by passenger-km or tonne-km provides unit costs that are five to six times as large as those for the strategic road network.

(2) If the national rail function were carried out by express coaches and lorries on an uncongested network, such as enjoyed by rail, then the fuel consumption would be reduced by 24%.

(3) A dedicated express bus lane has three to four times the capacity of a single railway track. An example is the contra-flow lane serving New York’s bus terminal. That carries 700 45-seat buses in the peak hour, offering over 30,000 seats. In comparison the 30,000 crushed passengers who arrive at Victoria Main Line in the peak hour require four inbound tracks, each wide enough for an express bus lane.

(4) Astonishingly, in London and in the peak hour the surface rail network is in highway terms scarcely used.

(5) The passenger or freight flows per track or lane are three times greater on the motorway and trunk road network than on the rail network.

Despite rail carrying only 6.3% of passenger-km and 8.5% of tonne-km it enjoys subsidies and capital grants running at £5 billion annually, equivalent to £200 per year in taxes for every household in the land, or to £150,000 per track-km, or to 10 pence per passenger-km or, if passengers and tonnes are added, to 7 pence per mile travelled.

In contrast the strategic road network makes a profit for the exchequer of £13 billion per year, equivalent to £250,000 per lane-mile or to 5.6 pence per passenger-km or to tax of £520 from every household in the land.

Such data underplay the poor financial performance of rail compared with road in that rail track is 6 to 7 times as expensive to build. Consequently, in terms of capital employed, the productivity of the strategic road network may outperform rail by a factor, not of 3, but of 20 (and that is before taking account of the relatively low value of rail freight, for the most part bulk minerals etc).

Moreover, nearly half the population uses rail less than once a year and those from the top quintile of household income use rail four to five times as much (the National Travel Survey) as those from either of the bottom two quintiles. Why on earth should taxpayers be subsidising the better off?

Paul Withrington

Member of the Advisory Council

Paul Withrington graduated in Civil Engineering from Bristol University in 1962. In 1966/7 he took an MSc in Transport Planning under the aegis of the Greater London Council, where he worked for two years followed by a period as a lecturer at Portsmouth.  He joined Northamptonshire County Council in1975 as Project Manager, Transport Planning. Since 1994 he has directed Transport Watch appearing at public inquires and undertaking policy analysis. In 2000/01 he appeared as the Strategic Objector at the Public Inquiry into Railtrack's West Coast Main Line Modernisation Programme.

20 thoughts on “Rail versus road”

  1. Posted 17/12/2011 at 09:05 | Permalink

    Comparing the subsidy of the rail network to the amount of tax revenue the government chooses to impose on drivers does not, as implied, compare to cost of the rail vs road. Most households are paying depreciation of £2,000 per year on one or two cars, as well as funding the road network, fuel, insurance, maintenance etc.

    It may well be useful to compare the cost of running a dedicated network of coaches to running he rail network, but implying the infrastructure has no cost because we currently raise so much tax from drivers the government has a surplus does not make sense. Imagine the cost of changing the rail infrastructure to a coach infrastructure and then the maintaining it.

  2. Posted 17/12/2011 at 12:53 | Permalink

    How does the Tube fare as a product?

  3. Posted 16/01/2012 at 19:02 | Permalink

    Reference Daniel Webb’s – I was comparing the costs to the Government. Apologies if that was not clear. The point is – roads are massively profitable to the exchequer whereas rail cost the exchequer £5 bn per year
    That is to say, road users are wiling to pay far more than their costs whereas rail passengers are heavily subsidised.
    Rail infrastructure and maintenance costs are very much higher than the same for the strategic road network. If DW cares to send me an e-mail I will send him more detail – or see the transport-watch web site. [email protected]. Otherwise there is our web site.
    I have not much on the Tube, but what is certain is that it too requires massive subsidy. If there were no subsidy to London’s infrastructure perhaps its high cost would have reduced the growth of the capital to the advantage of other cities and of the nation as a whole. Our web site does have a little on rapid Transit systems – see topic 9 – rather out of date.
    The web site is peraps best explored from the site map.

  4. Posted 17/01/2013 at 16:08 | Permalink

    “If the national rail function were carried out by express coaches and lorries on an uncongested network, such as enjoyed by rail,”
    It seems to me that this is the big if. If there were as many coaches as is suggested here, it would not be uncongested; it would be very congested! There just isnt room to get all those coaches in and out again!
    You cannot just convert railways into roads; roads need to be a lot wider, they take up much more real estate.
    A principal reason why the railways are not working at prime efficiency is the restriction on stations, etc being expanded to allow for longer trains…

    The main issue is safety; railways are much safer – many fewer casualties per mile. if all those people switched to roads it would push up insuranance premiums due to the increase in accidents, increasing costs. There would be more deaths, so fewer to share the remaining burden

    Every few years someone new comes along suggesting this idea, and every time it gets dismissed for very good reasons. A very good antidote to this theory is E A Gibbins; Railway Conversion: The Impractical Dream, which has just be republished as a Kindle edition (I am not connected to the author or publisher in any way)

    Tanya Jackson

  5. Posted 17/01/2013 at 17:15 | Permalink

    (1) Government expenditure on rail divided by passenger-km or tonne-km provides unit costs that are five to six times as large as those for the strategic road network.

    this is another interesting one. Where does this data come from? Expenditure on roads is distributed amongst many authorities. the Highways Agency only pays for motorways and some other major roads. the rest of the expenditure comes from many different councils and it has been notoriously difficult to pin down how much in total is spent in any one year – there is much research on this going way back when.

    Rail equipment is more expensive, but usually lasts much longer and its amortisation profile is much longer. It can recoup its expenditure quite quickly, though.

    4) Astonishingly, in London and in the peak hour the surface rail network is in highway terms scarcely used.

    it may seem that way to an observer who does not understand railway logistics, but paths for trains are generally crammed at peak hour. There have to be long stopping distances between trains. the payoff is they carry many more passengers.

    5) The passenger or freight flows per track or lane are three times greater on the motorway and trunk road network than on the rail network.

    I am not sure what this means, or proves?

    Tanya J

  6. Posted 17/01/2013 at 18:49 | Permalink

    Tanya J has not done the arithmetic, nor does not know the width a railway or a road. For widths see Fact Sheet 3 here
    The data comes from national sources – go divide costs by passenger-km.
    In Central London and in the morning peak hour the surface rail network is, in highway (or vehicle flow) terms substantially disused see topic 15 here
    For system wide death rates see Facts Sheet 2. Small numbers and the variability make it dishonest to compare or quote death rates for passengers by rail or express coach. However, system-wide, and quite contrary to the myth the death rates are in favour of road.
    Mr Gibbins and Co often compare the rail network with the entire road. That these people make such stupid comparisons is sufficient for them to be dismissed. – The rail network is 16,000 km long and contains circa 32,000 km of track. The road network is 400,000 km long. Most of its length is comprised of urban back street and country lanes. That network carries traffic a well as giving direct access to every house, retail outlet, factory and railway station in the land. Instead we compare rail with the strategic road network. It has a lane length of between 50,000 km and 55,000 km. It is used 2.5 to 3 times as intensively per lane as is rail per track despite rail having the advantage of serving the hearts of our towns and cities. For comment on Mr Gibbins’ book see Transport-watch topic 21.
    Those interest in the origins of the Railway Conversion should see the archive at topic 7. Readers may also like our evidence to the Transport Committee, Topic 18.

    The data comes from national sources. Ed Gibbins likes to ignore them. Go divide costs by passenger-km or see Facts Sheet 4 or topic 18.

  7. Posted 18/01/2013 at 02:23 | Permalink

    “At Waterloo 50,000 crushed passengers alight in the morning peak hour. They could all find seats in 1,000 50-seat motor coaches. Those coaches would occupy no more than one lane of a motor road. At Waterloo there is room for 3 or 4 lanes in each direction. The waste is lamentable.”

    I think the above quote really highlights the problem here. This ignores two crucial elements; lateral movement and kinetic continuity.

    On most roads there is the ability to remove broken down vehicles off to the side. On the expressways converted from two track railways there would be no space to enable this. Once a vehicle had broken down, that would be it, until the broken down vehicle was repaired you would into a situation where the coaches would have to pull into the opposite lane to pass it.
    Railways overcome this problem through kinetic continuity. Railway vehicles can buffer up against each other; one can come to the aid of another and push it out of trouble. Also, of course, well maintained trains are simply less likely to break down than road vehicles – you get extra reliabilty for those extra bucks.
    The casualty figures in
    are very interesting and are worth more study; of course the focus is so much on the passenger casualties on railways other, so called “movement accidents tend to be ignored. The sad fact is that a lot of people (particularly children) seem drawn to railways and their dangers in a way that they are not drawn to motorways. Once you extract those groups from the figures I am sure that they will reflect more what common sense tells us should be the case.

    Tanya J

  8. Posted 18/01/2013 at 15:17 | Permalink

    In answer to Tanya’s above – there is a contra flow express coach lane 11 feet wide in New York, which serves the main bus terminal.. That lane is 4 miles long including 1.5 miles in tunnel. It carries nearly 700 45-seat coaches in the peak hour offering over 30,000 seats. In comparison 30,000 crushed rail commuters arrive at Victoria Main Line in the peak hour in trains requiring four inbound tracks…..
    We do not have data on the relative breakdown rates of trains and express coaches, but what is certain that trains are uniquely fragile to minor problems such as leaves on the line and the wrong sort of snow. A simple prank, such as a rope thrown over the catenary, may disrupt a line for hours and hours. If the branches of trees encroach, so that a pigeon may touch the branch and the catenary when taking off, then the pigeon will be fried and the catenary brought down.
    The idea that it is easier to rescue a broken down train than a coach is difficult to defend. A train cannot be easily pushed out of the way and will block a track for hours and hours. A coach may be quickly pushed aside be towed away. Meanwhile other vehicles can usually drive round it, or do so easily enough on ordinary roads.
    The advantage of Tanya’s “kinetic continuity” rests with roads in that the gaps between vehicles may be a few tens of metres whereas that between trains is miles and miles. Trains cannot steer and have stopping distances several times that of road vehicles. For that reason signalling on the national rail system costs tens of billions of pounds. In contrast there are scarcely any signals on the more heavily used strategic road network.
    The railway lobby focuses its sasfety propaganda on passengers killed in train accidents. These account for less than 5% of those killed on the railway – if you fall of a platform you are not a passenger in a train. The myth of near perfect rail safety is nurtured (a) ignoring usage, and there are 17 times as many passenger miles by road as there are by rail, and (b) comparing passengers killed in train accidents with all those, system-wide, who are killed on an open access road network………..
    It would be dishonest to ignore any group of people who are killed, as canvassed by Tanya. It is up to the industry to create a safe environment. For rail that is difficult because of the long stopping distances, the inability of trains to take avoiding action and the low usage, which lulls the trespasser into a sense of false security.
    Common sense has nothing to do with it. Instead Tanya and the public at large are the victims of decades of shameless railway propaganda, which have turned the railways into a kind of religion. The truth is that the gap between the railway myth and reality is so big that it beggars belief. Sensible discussion, based on a careful examination of the facts, is thereby sabotaged.

  9. Posted 18/01/2013 at 19:14 | Permalink

    I really don’t think there is pro-rail propaganda at work here
    I think that what there is is a weary acceptance that if we were starting all over, we wouldnt start from here. The reason why railways are inadequate for our transport needs are often down to the mistakes of the 19thC constructors, not down to any inherent flaws in railways. I dont doubt that in some applications dedicated busways could offer a more flexible solution, if planned and constructed as such from scratch. Converting existing rail track beds to that purpose is really a second rate alternative with little to be gained and a lot to be lost. the fact that it hasn’t been done and has never even appeared on the agenda of even the most pro-road government speaks volumes for its practicality. I don’t believe that that is all down to pro-rail propaganda – the politicians don’t listen to anything else the rail lobby says, why should they listen to that?


  10. Posted 18/01/2013 at 19:25 | Permalink

    Having thought about it I realise that the safety figures presented in Transport Watch’s Facts Sheet 2 Casualty Rates, Rail And Road Compared are misleading. The reason why railways are subject to so many “movement” accidents, in this case trespass deaths, is because they are more likely to be found in densely populated areas, near to habitation, where the temptation to use the railway as a short cut is sadly inevitable. On the other hand, motorways are usually out of city centres or else have more robust defences that just a fence. Plus, also, with more frequent flow of vehicles they actively discourage people from trespassing on them. I am sure that if you looked into it you would find that railways suffer very much the same attrition rate as any type of industrial site in a city centre – and that includes areas where road vehicles move about off road. It is just that there is so much more railway – and it does obstruct peoples movement across the landscape. But then so do motorways. So would a rail to road converted expressway. How would the latter differ from a railway in a city centre?
    Railways also suffer the blight of level crossings, including pedestrian crossings, dating from an era before people knew better. The modern motorway network has been built with the hindsight gained from the experience of building the railways – there are bridges, & underpasses for pedestrians and cars.

    I also think that the fuel use statistics given in Facts Sheet 5: Fuel And Emissions: Trains Compared With Replacement Express Coaches And Lorries are extremely misleading. These fail to extrapolate the extra value gained through the usage of the fuel railways use. For instance, inter-city trains usually travel at speeds far in excess of road traffic – they use more fuel to do this, inevitably, but also more value is gained through the use of that fuel because people get to their destinations more quickly.
    The same for urban transport – many stops and starts uses more fuel – but then we gain the advantage of a service which serves many more stations.
    Comparing either of these usages to a road coach travelling along a motorway on a simple journey from a to b, maybe to c, is to miss the point. It does not provide the same service. We get less value for it.

    How would a route way on which x number of road coaches all driving in a line cope with the frequent station stops that a railway demands? All the coaches would have to stop in a line, tailing back, queuing for each station in a huge tailback. Even if they pulled off of the roadway (demolish existing platforms to construct loading bays?) there would still have to be allowance for them to get back into the traffic stream. Not all the coaches would serve all of the stations. Some would be rushing past whilst others loaded. You would need traffic lights to stop the passing stream and let the coaches out of the station bay. Congestion!
    No. The only practical – but far from perfect – way is to get as many passengers as possible into a single vehicle, – a train – and then multiply those and put sufficient headways between them to enable them to stop and start without inconveniencing other traffic. This is the logistical solution railways offer that I hinted at above – lines of road coaches would not enable this.

    Statistics and numbers can be used to prove many things. The rail to road case is a classic example of how they can be used to produce results contrary to common sense and simply counter to the practicalities of the real world

  11. Posted 19/01/2013 at 21:59 | Permalink

    Tanya – if bitumen macadam and pneumatic tyres had bee available at the start of the railway age perhaps they would have built system of reserved moor roads. Nowadays rail has become a kind of religion impervious to rational argument. Instead we have shameless propaganda. For example, in evidence to the Transport Committee’s inquiry into the Future of the Railway, 2003-04, Bombardier told the committee that, “to carry 50,000 people per hour in one direction we would need a 35 metre wide road used by buses or a 9 metre track bed for a metro or commuter railway”. The reality is that 1,000 express coaches per hour may offer 75,000 seats. If those coaches were travelling at 100 kph in one lane of a motor road the headways would be 100 metres.
    Similarly, Ralph Smyth, of the CPRE, when speaking at the Westminster Forum’s seminar “Getting UK rail on track, 6th Dec 2012, said, “the SNCF say that a 2 track high speed railway has the same capacity as a 10 lane motorway”. Again the gap between the claim and reality is stunning. If HS2 achieves 18 1000-seat trains per hour in one direction there will be 18,000 seats. In contrast one lane of a motor road used by express coaches could, as we have seen, offer 75,000 seats. Incidentally we heard at the same seminar that half of all rail commutes are less than 15 miles long……..
    I ask – is there is any possibility of jail time for those who deliberately mislead Commons Committees or for those who give out wildly inaccurate data at seminars chaired by MPs and members of the House of Lords?
    You could have some fun by reading the quotes in iten six in our archive here Item 7 is also fun.

  12. Posted 19/01/2013 at 22:01 | Permalink


    You make some good points with regard to accidents, but overlook that rail has no frontage access and is, in urban areas, generally well and truly walled off. In contrast the trunk roads are open to motorbikes, cyclists, horses and pedestrians whilst offering access to the property that they pass.

  13. Posted 19/01/2013 at 22:20 | Permalink


    Your comment on fuel use overlooks the fact half of all railway journeys are less than 20 miles long and that 90% of are less than 80 miles long. For most of those the express coach would beat the train for journey time – after taking account of a service frequency perhaps ten times that of the train.
    Also, our calculations do take account of the fuel used, in that we calculated the emissions and found that if the railway function were discharged by express coaches and lorries using rail’s rights of way, managed to a void congestion, the emissions would be reduced. Better still, tens of thousands of lorries and other vehicles would divert from the unsuitable rural roads and city streets that they now clog, cutting their emissions no end
    You may like to know that it costs the Taxpayer circa 7 times as much to move a passenger or tonne of freight by rail as it does by the strategic road network
    You are also misled with regard to the stopping of coaches. They would pull off the main line and use the vast areas now devoted to platforms. Those areas would be converted to slip roads. There would be no more of a problem entering and leaving than there is on motorways. As previously, there are no or virtually no signals on the strategic road network and none, or very few, would be needed on the converted rail system.
    If you want to appreciate the disaster that railways are imagine the strategic road network paved with railway lines. The place would be at a near standstill as are the railways in highway terms.

  14. Posted 12/02/2014 at 16:24 | Permalink

    what is the average annual maintenance cost of rail road , pls .

  15. Posted 14/02/2014 at 20:48 | Permalink

    The costs of rail and road are both buried in commercial confidentiality as well as the added complication of local roads being funded by local authorities. Finding out that sort of data for a comparison is impossible.


  16. Posted 15/02/2014 at 14:19 | Permalink

    Since I last engaged with this discussion I have cast around for more information about the Lincoln Tunnel XBL and have found that it is very far from the accident free, always smooth running system its proponents claim. It is subject to very regular issues.
    To pick just two from a lengthy assortment; This was a serious crash involving one bus rear-ending another –

    The incident in the following link also highlights the primary dangers of the system

    it may have been caused by a rogue motorcyclist with no business to be there but the follow on incident again highlights the dangers of the system. Two buses collided with the bus involved in the initial collision. The supposed professionalism and training did not [prevent a serious incident which blocked up the system for a full ninety minutes at rush hour.

    All of the stuff I have found highlights what one intuitively feels must be the dangers of the system. Poorly paid drivers doing boring repetitive work become mind-numbed and when they need to put the brakes on, they don’t.

    There will come a time when we will have to re-evaluate the relationship between road and rail. The possibilities of dual system vehicles could be coupled with the possibilities of signalling systems such as ERTMS. This may enable road vehicles to safely join and leave rail systems for long journeys – thus taking advantage of the massive reduction of rolling resistance that rail offers.
    As technology speeds up the family saloon with new automated safety and driving features its occupants will find themselves increasingly in conflict with HGVs on the roads. These will be confined to low speeds for reasons of efficiency. At that point the majority of road users will likely demand moving freight onto rail.


  17. Posted 15/02/2014 at 18:03 | Permalink

    Rail is vastly expensive and low capacity compared with the express coach on an uncongested motor road. The longer term prospect is driverless cars/buses/coaches, vastly increasing capacities..
    The express coach lane in New York is low teck. E.g. they insert plastic poles in the carriageway to separate the lane from oncoming traffic when operative……..

  18. Posted 05/12/2014 at 07:31 | Permalink

    I’m baffled by the assertion that the road network makes a profit. No charges are made for the use of roads, save the BNRR and a few bridges and tunnels. Hence it is purely a cost.

  19. Posted 23/01/2016 at 05:59 | Permalink

    Aside from the 50,000 figure being a gross underestimate (try 100,000) those thousand buses would be arriving at Waterloo 3.6 seconds apart. Any idea where they’re all going to park whist the passengers alight or are they supposed to jump off whilst the bus is still moving? With vehicles running so close to each other the allowable speed will not be great and anybody who’s ever experienced the ripple effect on a Motorway will know that one driver has only got to blow his nose to bring everything to a standstill a mile behind him.

    Aside from finding drivers for all these buses London’s air pollution already breaks EU limits and kills tens of thousands every year. All the rail lines leading into Waterloo are electrified and replacing these electric trains with diesel buses spewing NOX and particulates into already polluted air doesn’t strike me as a very sound idea.

    Heavy goods vehicles, including buses, cause the majority of damage to Britain’s roads but this is not reflected in the taxes they pay. In this respect the road haulage industry receives a £2-3 billion subsidy from the rest of us every year.

  20. Posted 25/01/2016 at 17:04 | Permalink

    In reply to Deipnosophista at second above – has he (or is it she) not noticed that taxes from road vehicles far outweigh the expenditure, providing a profit to the exchequer?

    In reply to Mr Kempe: from the White Paper “Delivering a Sustainable railway” July 2007 Cm7176 we, for 20008/9, 36,900 high peak hour arrivals at Waterloo with 41,700 forecast for 2013/14. If we have as many as 50,000 then they would all find seats in 1,000 50-seat coaches or 670 75-seaters. In comparison many of the peak hour train passengers are standing in crush conditions. 1,000 vehicles an hour at 100 kph would have headways of 100 metres, more than commonly seen on motorways. Mr Kemp goes on to bleat about where the coaches would park. Well, what on earth do they do with the trains, a far bigger problem? In the case of coaches, some would drive on to serve London, others would go back out as do the trains but with a far shorter dwell time, so cutting the terminal space required. As to pollution, firstly the NOX and particulate risk is much overdone, see Transport Watch topic 34. Secondly there would be a huge environmental gain from opening these trivially used railway rights of way to lorries and other vehicles which currently clog unsuitable city streets and rural roads, provided only that the system should be managed to avoid congestion.

Comments are closed.