Memorandum submitted by Transport 2000
INTRODUCTION
Transport 2000 is the independent national body
concerned with sustainable transport. We look for real solutions
to transport problems and aim to reduce the environmental and
social impact of transport by encouraging less use of cars and
more use of public transport, walking and cycling.
Transport 2000 believes it is imperative that
rail fares are structured and priced so as to:
— encourage a modal shift and attract
people out of their cars and on to trains;
— offer rail passengers off-peak discount
travel so as to maximise capacity on off-peak services, but not
penalise passengers for travelling at peak times; and
— thereby, offering a socially inclusive
means of travel in the UK and still generating revenue for the
rail industry.
We have identified as the main and inherent
problems with existing rail fares that they are expensive, complicated
and inflexible.
HIGH-COST
FARES AND
THE EFFECT
OF FARE
CHANGES ON
THE DEMAND
FOR RAIL
Whilst the costs of motoring have fallen, rail
fares have increased substantially and above inflation. Since
1975 the cost of owning and running a car (taking into account
purchase, maintenance, petrol, tax and insurance) has fallen by
11% in real terms while rail fares have climbed 70% (Department
for Transport 2005).
Table 1 shows the changes in rail ticket prices
since 1999 as reported by the SRA and clearly shows that because
London and South East operators have a higher proportion of regulated
fares than other operators that their overall change was (unlike
other sectors) below the rate of inflation.
Table 1
CHANGES IN TICKET PRICES ADJUSTED FOR INFLATION
| Long Distance
| London and SE | Regional
| All Operators |
January 1999 | 100
| 100 | 100 | 100
|
| January 2000 | 105 | 99
| 100 | 101 |
| January 2001 | 105 | 97
| 100 | 101 |
| January 2002 | 109 | 97
| 101 | 102 |
| January 2003 | 110 | 97
| 100 | 102 |
| January 2004 | 112 | 98
| 108 | 103 |
| |
| | |
Source: calculations from data published on SRA website.
However, what is of real concern is the actual price paid
for travelling by train (earnings per passenger km—some
times referred to as yield).
Table 2
EARNINGS ADJUSTED FOR INFLATION PER PASSENGER KM IN REAL
TERMS BY SECTOR
| Long Distance
| London and SE | Regional
| All |
1998-99 | 100 |
100 | 100 | 100
|
| 1999-2000 | 103 | 100
| 99 | 101 |
| 2000-01 | 105 | 94
| 109 | 100 |
| 2001-02 | 107 | 96
| 97 | 100 |
| 2002-03 | 109 | 93
| 98 | 100 |
| 2003-04 | 111 | 94
| 96 | 100 |
| |
| | |
Source: calculations from data published on SRA website.
Table 2 shows that the actual price paid per km travelled
in real terms varies more widely between sectors than the change
in ticket prices. It is also clear particularly in the case of
Regional operators that passengers react to fare increases by
changing their travel patterns. It would be tempting but premature
to conclude at this stage that, since there are more realistic
alternatives to the typically shorter train journeys in regional
and south east sectors, the longer distance operators have exploited
their position. However it should be noted that short distance
earnings per passenger kilometre fell for both long distance and
shorter operators, whilst ticket prices only fell in real term
for London and South East operators. Clearly more detailed analysis
is required.
Using data published by the SRA and from a series of reports
published in the journal Transit over the years, which provide
summary of the annual published accounts of selected train operators,
the results shown in table 3 were calculated.
Table 3
CHANGES BETWEEN THE FINANCIAL YEARS 1998-99 and 2003-04
Operator |
Sector | real terms % change in
earnings per passenger km
| % change
passenger km |
Thameslink | London and SE
| ¸0.87 | 20.50 |
| Scotrail | Regional | ¸4.92
| 15.34 |
| C2C | London and SE | ¸5.80
| 14.71 |
| GNER | Long distance | ¸5.20
| 13.49 |
| First Great Western | Long distance
| 4.74 | 12.84 |
| Central Trains | Regional |
0.10 | 12.04 |
| RRNE/Northern Spirit/ATN | Regional
| 11.22 | 3.47 |
| North West Trains | Regional
| 34.09 | ¸0.66 |
| Virgin Trains | Long distance
| 10.98 | ¸1.52 |
Source: Calculations from Annual Accounts as published in Transit/SRA web site.
| | | |
It can be seen from table 3 that in general, the higher the
real cost of travel per kilometre the lower the growth in passenger
kilometres. Such data can be used to determine the relationship
between the amount paid for travel and the demand for travel as
well as the long run elasticity for rail fares. This is shown
in figure 1.
Figure 1
The line of best fit was obtained using the Microsoft Excel
package. An "R2", also known as the coefficient of determination
of 0.65, implies a coefficient of correlation of 0.8, which suggests
that there is a link between the two variables, and in this case
the gradient of the graph, -0.4863 is the long run average price
elasticity.
This figure suggests that the demand for rail travel is much
more elastic than often assumed—the elasticity derived here
is closer to ¸1 than the often quoted one of ¸0.3. This
implies that there is considerable scope for reducing fares to
fill otherwise empty seats. Indeed this approach is at the heart
of both the new GNER franchise and Midland Mainline's forthcoming
initiative to attract new custom through a sophisticated computer
system designed to maximise the passenger numbers on each train,
rather than the approach of Virgin which has been to increase
the yield per passenger kilometre. Conventional wisdom suggests
that faced with an elasticity of around ¸0.5 the response
of any business is to increase the price to achieve more revenue
from a declining number of customers in order to improve profitability
of the company. However, the two rail operators in the sample
(Virgin and North West Trains) that pursued these policies have
required massive extra amounts of government money to keep them
solvent, whilst some of the operators who have pursued a policy
of increasing train loads, such as GNER and Thameslink, are paying
ever increasing premiums into the Treasury. In summary, then,
headline fares increases hide significant variations. Different
rail operators have pursued different pricing strategies, and
the evidence suggests that strategies that aim to maximise passenger
numbers (yield per train) do better for the company, the taxpayer
and the wider community than the strategies that aim to get more
revenue from declining numbers of passengers (yield per seat).
VARIATIONS IN
LOCAL FARES
Much rail comment in the media and from the government has
focused on long distance journeys, with the implication is that
rail is best at that and should focus on that at the expense of
local journeys. Indeed, many of the recent Route Utilisation Strategies
and franchise specifications explicitly propose reductions or
closures of local services/stations to make way for longer distance
trains. But this ignores the fact that rail is used overwhelmingly
for shorter distance journeys. Table 4 shows the proportion of
journeys undertaken in each rail sector and the average length
of those journeys
Table 4
| Long Distance
| London and South East | Regional
| All operators |
Percentage of all journeys undertaken by sector
| 8% | 69% | 24%
| 100% |
| Average journey length (miles) for sector |
100 | 18 | 19 |
24 |
| | |
| |
Source: Calculations from SRA website.
It is clear that around 92% of all rail journeys have an
average length of less than 20 miles. It is tempting to say that
these are all journeys made by London commuters. However in 2001,
before the massive expansion of their London—Birmingham
services, Chiltern Railways web site revealed that 1/3 of their
passengers arrived in London before 10:00, 1/3 arrived in London
after 10:00 and 1/3 did not travel to London at all. One only
has to visit stations at regional centres such as Southampton
and Leeds in the morning peak to see the importance of rail commuting
outside London.
Table 5 shows the average journey length for a range of operators—it
is clear that there is a lot of very short rail journeys being
made which are typically less than 30 minutes duration.
Table 5
| Sector
| average journey length (miles) |
Southern | London and SE
| 14.6 |
| North West Trains | Regional
| 16.9 |
| Silverlink | London and SE |
17.4 |
| C2C | London and SE | 17.5
|
| South West Trains | London and SE
| 18.7 |
| Great Eastern | London and SE
| 19.2 |
| Arriva Northern Trains (RRNE) | Regional
| 19.7 |
| Thameslink | London and SE |
20.1 |
| Scotrail | Regional | 20.9
|
| Central Trains | Regional |
22.1 |
| Great Western | Long Distance
| 77.3 |
| Midland Mainline | Long Distance
| 79.2 |
| Virgin Trains | Long Distance
| 99.7 |
| GNER | Long Distance | 155.8
|
| |
|
INCONSISTENT AND
COMPLICATED FARES
For this vast majority of local rail journeys, there is a
huge array of different fares available with very little consistency.
Though the cost of rail travel per mile is roughly the same throughout
the former Network SouthEast area there are wide discrepancies
between comparable journeys elsewhere in the country. For example
a Radcliffe (5 miles east of Nottingham) to Nottingham day
return costs £3.10 (£3.20 in peak), whereas Attenborough
(4¼ miles west of Nottingham) to Nottingham day return
costs £1.70 (any train). This means that it costs over 50%
more per mile to travel on the same train. The cheapest day return
fare from Radcliffe to Sheffield (46 miles) is £16.20 yet
Nottingham to Sheffield (41 miles) cheapest day return is £9.90,
a difference of £6.30 for five extra miles. Differences such
as this cannot be justified and perhaps explains why over twice
many passengers" use Attenborough station compared with Radcliffe.
For less than the extra charged for the 5 miles to Radcliffe,
the fare for the 16 miles from Nottingham to Derby is £3.90
off peak return. These inconsistencies cannot be justified and
the higher fares can act as a disincentive to rail travel.
In East Anglia cheap day return fares are quoted only to
and from larger stations, such as Lowestoft, Great Yarmouth, Norwich
and Ipswich. This means that on the East Suffolk line, for example,
it is cheaper to buy a cheap day return from Beccles to Lowestoft
(£2.40, 8Ö miles) than a day return from Beccles to
Oulton Broad South (£3.60, 6Ômiles). There are countless
other anomalies around the country of unexplainably high fares
for short journeys which often result in increased car use and
congestion, or isolation for those who cannot afford to travel.
INFLEXIBLE FARES
If a group of people want to travel together by train, the
cost can become a real barrier. Some companies offer a "carload"
fare where four or five people can travel together at a discounted
group price. This should be standardised and available on all
operators' trains. To make a round trip to one town, or to another
and back again, three single tickets would currently be required,
again making the trip overly expensive. A simple ticket should
be available to allow passengers to make round trips to return
from a different station to their original destination.
THE EFFECT
OF ALLOCATION
FORMULAS ON
BRANCH LINE
EARNINGS
Revenue for through journeys from branch lines onto the rest
of national network are calculated on the basis of mileage. Though
this appears to be fair, in reality the branch line operator actually
receives much less than the passenger thinks they have paid for
the extra journey on the branch. Table 6 illustrates this point
with an example of fare and allocation to branch operator.
Table 6
ALLOCATION OF FARE TO BRANCH OPERATOR (ALL RETURNING SAME
DAY USING SAME BRANCH LINE TRAIN)
| Exeter
| Bristol | London
|
Distance (miles) | 11
| 87 | 184 |
| Fare for day return travel | £3.50
| £22.20 | £49.00 |
| Allocation to operator of branch line | £3.50
| £2.81 | £2.93 |
| |
| |
Source: NRES website.
Where the journey on the branch is typically very short—for
example Island Line or St Ives—then the branch allocation
for long journeys can be as low as one fifth of what the passenger
thinks they have paid extra for that part of the journey. Clearly
if the branch operator increases their fare to make up for the
losses it will only discourage further use of the branch line.
And of course there is every incentive for the branch operator
to issue two tickets rather than one in order to get the full
allocation.
MITIGATING THE
IMPACT OF
ENGINEERING WORKS
ON TICKET
REVENUES
Engineering works not only cause considerable disruption
to passengers but also affect the revenue of operators not only
at the time of the works but some time afterwards. The aftermath
of the engineering works following Hatfield on the passenger numbers
clearly demonstrates this. Network Rail has followed the practice
of "big bang" engineering possessions in order to minimise
its own costs, rather than a series of short overnight possessions.
However, it may well be that the total industry cost is higher
when lost revenues are taken into account. Island Line with its
unique track access arrangement insist that nearly all works are
carried out at night in a series of short possessions in order
to minimise disruption to passengers, the revenue flow and the
cost of alternate travel. Chiltern Railways are looking to adopt
a similar approach. Chiltern's chairman Adrian Shooter was quoted
in Transit (20/5/05) posing the question "Motorways
are busy on Sundays—so why do we close the railway?"
TRACK ACCESS
AND LEASING
CHARGES
Two of the major costs faced by train operators are track
access charges and train leasing charges, which are typically
much more than operator's labour costs. Table 7 shows a wide variation
in track access costs.
Table 7
TRACK ACCESS AS A % OF ALL COSTS SELECTED OPERATORS
| SilverLink | 48.70% |
| Virgin Trains | 39.00% |
| Scotrail | 38.50% |
| Central Trains | 36.40% |
| GNER | 26.10% |
| C2C | 26.00% |
| South West Trains | 24.00% |
| |
Source: company accounts published in Transit.
Further detail analysis of selected company accounts presented
in table 8 show again a wide variation in the amounts paid by
operators. Given that both Scotrail and Central operate trains
that are a lot shorter and a lot slower than GNER one must ask
why they are paying more per mile to run their services, with
all the pressures this puts on operators to increase fares. Much
concern has been made in recent years of escalating costs and
the implications for the level of fares charged and amount of
public subsidy required in certain cases. Little or no attempt
has been made to understand why rail based projects and many times
the cost of similar non-rail projects. An independent railway
constructed a passing loop for around £150,000 using a Network
Rail contractor yet Network Rail have quoted a franchised train
operator £500,000 for a similar loop operating at similar
line speeds. It is often the case on many rural lines that additional
passing loops could allow a more attractive timetable which in
turn would increase passenger numbers and reduce pressure on the
operator to increase fares, but if even the simplest of works
are prohibitively expensive then it will not be possible to attract
new passenger journeys.
Table 8
| Operator | Access charges per train mile
|
| Scotrail | £6.76 |
| Central Trains | £5.36
|
| GNER | £5.32 |
| Island Line | £5.20 |
| |
Source: company accounts published in Transit.
Rolling Stock Leases are generally the second biggest item
of expenditure for train operators. Table 9 shows the proportion
of each £1 collected in fares that is spent on rolling stock
leases. Once again there is a wide variation in the figures.
Table 9
PERCENTAGE OF FARE INCOME SPENT ON ROLLING STOCK LEASES
| North Western Trains | 46.48%
|
| Central Trains | 39.89% |
| Arriva Trains Northern (RRNE) | 37.24%
|
| Scotrail | 29.76% |
| C2C | 28.60% |
| South West Trains | 25.54% |
| Midland Mainline | 21.32% |
| SilverLink | 18.43% |
| GNER | 16.25% |
| |
Source: company accounts published in Transit.
With such high ratios for regional operators like North West
Trains, there can be little incentive to grow the market by providing
more trains, as any new revenue will be almost swallowed by additional
leasing costs.
CONCLUSIONS
The Government says it wants more people to leave the car
at home and take the train, but this requires train travel to
be a simple and affordable alternative. High and inconsistent
unit fares for short journeys in parts of the country do little
to encourage train travel and the evidence here suggests that
fares strategies that get more people on to trains are better
for both operators and taxpayers than strategies that try to wring
more money out of fewer people. There seems to be no consistent
basis for the levels of many fares that passengers pay or for
the two biggest items of expenditure paid by train operators,
namely track access charges and rolling stock leases, and this
needs to be addressed.
Transport 2000 would like to see a fares system that is affordable,
simple and flexible
— Simpler fares: one set of uniform restrictions
should apply as a base right across the network. Advance fares
and other offers should have the same names and comparable conditions
anywhere in the UK so passengers know what they can expect and
what they are buying.
— Flexible fares: tickets for groups of people travelling
together and people wanting to make a round trip of more than
one destination.
— A national rail card: there is scope to grow the
market for rail travel by easily understood discount schemes and
serious consideration should be given to a National Railcard along
the lines of the Network Card before restrictions were imposed.
Such a card should be available to everyone, and should incentivise
people to use the train and reward them for travelling regularly.
| 
