Memorandum by the London Transport Users
Committee (LTUC) (OPT 11)
OVERCROWDING ON PUBLIC TRANSPORT
Thank you for inviting us to submit a memorandum
in connection with the Transport Committee's forthcoming inquiry
into overcrowding. This is a topic which has long been of concern
to the London Transport Users Committee (LTUC) and its predecessors,
and we welcome the recognition of its seriousness which is implicit
in its selection by your Members as an issue deserving of their
scrutiny. In the astonishingly short interval allowed for receipt
of submissions, it has not been possible for us to compile as
comprehensive a review of the evidence as we would ideally have
wished. But we hope that the following points will be of relevance
and value.
1. IS OVERCROWDING
A SOURCE
OF CONCERN
TO USERS?
1.1 "Overcrowding" is an elusive
concept, though everyone knows it when they encounter it. What
passengers regard as intolerable overcrowding may be seen by commercial
transport operators simply as efficient asset utilisation. On
the railways, it has no formal definition, because there is no
legal limit to the number of people who can be carried on a train
(or be present in a station), if they choose to force their way
on board. And short of requiring all seats to be reserved, which
is hardly a practicable option on busy commuter routes, there
is no means of imposing one. Buses are subject to restrictions
on their carrying capacity, which are set by law and (because
they have fewer entrances and exits than trains) these are enforceable
by their crews. But there is no systematic surveillance of the
degree of compliance with these, and/or of the extent to which
their capacity falls short of demand. On both modes, the operators'
"conditions of carriage" specifically exclude any entitlement
for ticket holders to be provided with a seat, except when one
has been reserved, or to be accommodated on any particular journey.
1.2 Complaints about overcrowding per
se do not dominate the caseload of representations which the
LTUC receives from members of the travelling public in and around
the capital. Regular users have long since come to regard it as
endemic on the busiest parts of the network at the busiest times
of the day and week. Their failure to complain is doubtless conditioned
by an awareness that it is not a problem susceptible to an instant
solution which an intervention on the part of LTUC with the relevant
operator is likely to secure. But when it occurs most acutely,
it is often a symptom of other deficiencies in the planning and
delivery of the service, such as delays, cancellations and (in
the case of trains) short formations. So it is commonly mentioned
in that context, where the primary purpose of the complaint is
to secure redress for a specific journey failure.
1.3 The Strategic Rail Authority commissions
a twice-yearly poll of users' satisfaction with the quality of
service offered on the National Rail network. Respondents are
asked to rate their satisfaction with the overall journey experience,
and with eleven individual service attributes, on a five-point
scale from "very poor" to "very good". LTUC
regularly monitors the findings, and reports these in terms of
the net satisfaction rate, ie the excess of the proportion awarding
positive ratings over the proportion awarding negative ratings
(so the maximum possible range of values is +100% to -100%).
1.4 The most recent results available at
the time of drafting this letter were those for the seventh "wave"
of the survey, conducted in the spring of 2002. The overall net
satisfaction rating given by rail users in the London area was
+55%, compared with +60% nationwide. Londoners were also more
dissatisfied with all but one of the eleven individual facets
of the service which are covered by the survey. In the case of
"amount of seats/standing space" (the nearest approximation
to a specific question on overcrowding), their average net satisfaction
rating was +27%, compared with +38% nationally. Only two other
service attributes were awarded even lower scores : "value
for money" (-3% and "upkeep and repair of trains"
(+15%).
1.5 Not surprisingly, users of longer distance
services were generally more content with this facet of performance
than users of local trains in the London and south east area,
because seats on longer distance journeys can usually be reserved
in advance (and for some types of ticket, this is automatic).
"Amount of seats/standing space" received a net satisfaction
rating of +61% from passengers on the longer distance trains to
and from London (ie the former InterCity routes), but a rating
of only +32% from the users of the London and south east network.
And within the latter group, there were wide variations between
individual train companies. While Gatwick Express and Chiltern
Railways achieved scores of +86% and +51% respectively, the equivalent
ratings for Connex South Eastern and Thameslink were only +21%
and +23%.
1.6 Passengers declaring themselves to be
"regular users" were also asked a more specific question
about their experience of seat availability. Of these, 31% stated
that there are always seats available, and 41% that this is usually
so. Of the 16% who declared that they usually stand, 7% said that
their trains were crowded and 5% that they were very crowded.
The residual 10% reported that their experience varied. But these
data are not disaggregated by operator, route or time of travel,
and therefore give no direct insight into the location or duration
of the problem if it occurs.
1.7 London Underground also conducts regular
satisfaction surveys amongst what it is wont to describe as its
"customers". The methodology differs from that used
for the Strategic Rail Authority's polling. For example, respondents
are interviewed face-to-face on completion of their journeys rather
than being asked to fill in questionnaires during the course of
them, the content and wording of the questions varies, the range
of issues covered is not identical, and satisfaction is recorded
on a scale from 0 to 10 (with average scores scaled up to a range
from 0 to 100). So direct comparisons cannot be made between the
reported satisfaction levels of the users of the two networks,
although many will be the same passengers.
1.8 The Underground satisfaction survey
covers 19 separate attributes of the service. In the most recent
period for which data have been published, July to September 2002,
the overall rating awarded by users was 75. The ratings for train
crowding and platform crowding were, respectively, 71 and 73,
ranking 15th and 11th out of the 19 attributes (the lowest score,
64, being given to train cleanliness). By this measure, therefore,
London Underground is clearly regarded by its users as serving
them less adequately in relation to crowding than in relation
to the majority of the elements of its service.
1.9 A similar survey is conducted on behalf
of London Buses, although again there are differences in methodology
which make direct numerical comparisons invalid. In the most recent
period for which results are available, July to September 2002,
bus users' overall satisfaction rating was 79 (on a scale from
0 to 100). The score for "level of crowding on bus"
was 77, placing it 11th out of the 17 service characteristics
covered, the lowest score being given to reliability (66).
1.10 Users of London's bus stations are
also polled. In the same period, the overall rating was 67, and
the rating for the level of crowding was 70, placing it 10th out
of 22 elements covered. The lowest rating in this survey was for
the condition of toilets (44).
1.11 Taken in isolation, these figures may
have little apparent meaning. But the overall message they convey
is that, for any of the modes of travel cited, users are more
likely to be dissatisfied with the level of crowding they encounter
(which they may therefore regard as overcrowding) than they are
with the quality of service experienced overall, and with the
majority of the other individual attributes about which they are
asked.
1.12 What these surveys do not reveal, however,
is the relative importance attached by users to the various attributes
of the services they encounter, and therefore the relative weighting
they give to the need for improvement in each aspect. An insight
into the importance of understanding this was offered by a recent
survey of Public attitudes to transport in England 2001,
sponsored by the Commission for Integrated Transport. Respondents
were asked to rank a number of possible areas of improvement according
to the level of priority they believed each should receive. Bus
users ranked overcrowding sixth out of nine service elements listed,
while rail users ranked it third and Underground users ranked
it first. This indicates that the issue which the Select Committee
is addressing is one with high salience in the eyes of those who
are affected by it.
1.13 It is also worth bearing in mind the
comparative ratings awarded by the Government's People's Panel
to the service quality delivered by a sample of retail services
(on a scale from 0 to 100). Energy and communications utilities
received net satisfaction ratings of over 80, and high street
banks or building societies were rated at 78. The equivalent figures
in the transport sector were 44 for London Underground, 37 for
local buses and 28 for train companies.
1.14 The clear impression which emerges
is that public transport in general is regarded as offering a
much lower quality of product than other commercial services,
and that amongst people who do use it, the level of crowding ranks
high on the list of service attributes they dislike. If the Government's
ambition that trains and buses should recapture a larger share
of the total transport market is to be fulfilled (and LTUC fully
endorses this as an objective), the problem of overcrowding will
have to be addressed, because it is a major deterrent to the use
of public transport on the part of those travellers who have a
choice. The irony of the position is that if the Government's
policy is successful, then at least in the short term the problem
is likely to grow worse, because overcrowding is a symptom of
undercapacity at the times and in the places of peak demand, and
a modal shift away from other forms of transport will simply exacerbate
the situation. But enlarging capacity is expensive, particularly
on the railways, and offers a low rate of return on the investment
required if the additional space provided is fully used for only
a short period each day.
2. WHERE DOES
OVERCROWDING OCCUR?
2.1 In the absence of any objective definition
of overcrowding, it is impossible to describe its incidence precisely.
But on some parts of the public transport network, loading levels
are recorded, and the resulting data can be used to obtain a partial
insight into the extent of the problem.
2.2 The operators of all franchised National
Rail services (which constitute the vast majority) on the national
rail network are required under the terms of their contracts with
the Strategic Rail Authority to "use all reasonable endeavours
to ensure that sufficient capacity is provided on each train .
. . to carry without excessive overcrowding all passengers intending
to travel on such a train . . ." But the term excessive overcrowding
is not defined. The contracts merely set an "initial number
of vehicles" as a benchmark, and operators face financial
penalties if the number of vehicles included in their train plans
(timetables) persistently falls below this.
2.3 There is no systematic tracking of passenger
numbers relative to train capacity on the longer distance services.
The relevant train companies are perfectly willing to carry standing
passengers for long journeys if they have not reserved seats or
if all seats have been taken, although their rolling stock appears
to be designed on the assumption that such a situation will seldom
if ever arise, because no specific provision is made for the presence
of standees. If standing passengers travel in the aisles of the
saloons, they will obstruct movement along the train, and will
find only vestigial handholds to support them. If they cluster
in the vestibules, they will find no grab rails, perches or other
means of support at all (and will probably find conditions extremely
draughty). Anecdotally, we are aware that travel in such circumstances
is a common event, and has become more so where older types of
train (typically loco-hauled stock) have been replaced by newer
ones (typically multiple units, with fewer carriages). But in
the absence of any centralised monitoring and reporting of train
loadings, we have no means of quantifying the precise distribution,
frequency or scale of the problem.
2.4 Most local rail services in and around
London and the south east are subject to more exacting requirements
in the form of "load factor specifications", related
to the notional carrying capacity of each type of train. This
varies according to the configuration of each type of rolling
stock, depending on the number of seats and on the available standing
space, but in broad terms it equates to about 110% of the seating
capacity of slam-door trains and to about 135% of the seating
capacity of sliding door trains. First class accommodation is
disregarded. In addition, it is assumed that no passengers should
have to stand for involuntarily for more than 20 minutes, so on
trains where scheduled intervals between stations exceed this,
the notional capacity is equal to the number of seats.
2.5 All with-flow peak period trains entering
or leaving London are monitored once a year, usually in the autumn
(on dates selected to avoid known fluctuations in demand, eg school
holiday periods). The peak period trains are defined as those
arriving between 0700 and 0959, and those departing between 1600
and 1859. The count is carried out by observers located on platforms,
because the level of crowding is often such as to make it impracticable
to conduct counts on board, and only a small number of trains
are yet fitted with automatic counting equipment (which detects
variations in the load borne by the suspension system). There
is thus an element of estimation in the results, but their accuracy
is regarded as sufficient to identify the routes and trains on
which crowding is most acute. The monitoring sites are selected
to reflect the point of maximum loading on each route, which is
normally on the approach to the London terminus, though it may
be further afield if there is substantial level of transfer to/from
the Underground at an intermediate station (such as Finsbury Park).
Some sections of line are omitted where there is a high level
of interchange between services for short-distance feeder trips,
eg between Charing Cross and London Bridge, where many users simply
board the first train to depart regardless of its ultimate destination.
2.6 These counts are used to calculate the
number of "passengers carried in excess of capacity"
(PIXC). The total number of such passengers on any trains where
the notional capacity is exceeded is expressed as a percentage
of the total number of all passengers on the relevant route in
the relevant direction. "Undercrowding", ie unoccupied
spaces on trains with fewer passengers than their notional capacity,
is disregarded because it is assumed that such spaces are not
pertinent to the needs of "excess" passengers on other
trains.
2.7 It should be noted that PIXC counts
take no account of whether or not trains are actually running
in their planned formations (or at all). So carrying 800 passengers
on a train with a notional capacity of 1,000 would not result
in a breach of the PIXC target even if in reality half the carriages
are not provided and only 500 spaces are actually available. Similarly,
no breach would occur if one of two trains each with 500 places
and 400 passengers is cancelled and all 800 are carried on the
other. Operators are penalised financially by means of other incentive
regimes, if they fail to provide the planned number of spaces
(and trains), but for PIXC monitoring purposes the phantom spaces
on missing carriages are still deemed to be present. This means
that although these data can be used as a theoretical indicator
of the incidence of crowding, they necessarily understate its
true prevalence.
2.8 The published PIXC totals are averages
based on all trains to or from London on a particular operator's
network, taken over the whole of the peak period. Loadings on
individual trains and individual routes may be very much higher
or lower. But because operators cannot directly control the number
of passengers who may choose to travel on any particular train,
they are not subject to penalties if individual services are excessively
crowded. Instead, they are required to make "reasonable endeavours"
to match their resources to passenger demand, if and when the
recorded PIXC rates exceed the SRA's "threshold" levels
for their network as a whole. These thresholds have been set at
4.5% in any one peak, or 3.0% in both peaks combined. The more
exacting threshold value for the combined peaks reflects the fact
that the evening peak is more extended than that in the morning,
with passenger numbers during the peak three hours about one fifth
lower, and pressure on available capacity is consequently less
acute (partly because, for example, schools open at around the
same time as shops and offices in the morning, but close earlier
in the evening).
2.9 The most recent PIXC census for which
results are available is that conducted in the autumn of 2001.
The published results (shown in the accompanying table) are expressed
purely in percentages, to facilitate inter-operator comparisons,
and the SRA does not disclose the actual number of passengers
affected. In the case of the operators with larger or more complex
networks, sub-totals for individual "inner" and "outer"
route groups are provided, although the thresholds still apply
only to the average for the entire company. Results in excess
of the applicable threshold value are shown on a shaded background.
2.10 The table reveals that three of the
10 London area operating companies to which load factor specification
applies exceeded their PIXC thresholds in 2001 in respect of the
morning peak period (Silverlink, South Central and South West
Trains). The same operators plus a fourth, Thameslink, were in
breach of the combined peaks threshold. These thresholds were
also exceeded by the combined load factor performance of the London
and south east network taken as a whole. And First Great Eastern
would also have been in breach, in respect of its inner area services,
if the thresholds applied to route groups separately.
| Operator/route group |
AM peak | PM peak
| Both peaks |
| c2c | 0.3% |
0.6% | 0.5%
|
| Chiltern | 2.4%
| 0.6% | 1.6%
|
| Connex | 3.2%
| 1.3% | 2.3%
|
| Inner | 2.8%
| 0.3% | 1.7%
|
| Outer | 4.4%
| 3.6% | 4.0%
|
| First Great Eastern | 3.7%
| 1.6% | 2.7%
|
| Inner | 6.0%
| 2.4% | 4.3%
|
| Outer | 1.5%
| 0.8% | 1.2%
|
| Silverlink | 8.1%
| 3.4% | 5.9%
|
| Inner (orbital) | 15.9%
| 8.9% | 12.7%
|
| Inner (radial) | 2.2%
| 0.0% | 1.1%
|
| Outer | 3.5%
| 0.0% | 1.9%
|
| South Central | 11.2%
| 1.0% | 8.9%
|
| Inner | 10.5%
| 0.7% | 6.5%
|
| Outer | 12.5%
| 1.7% | 8.0%
|
| South West Trains | 6.3%
| 2.5% | 4.6%
|
| Inner | 6.5%
| 2.1% | 4.6%
|
| Outer | 5.9%
| 3.4% | 4.7%
|
| Thames | 3.3%
| 1.6% | 2.5%
|
| Inner | 2.8%
| 2.9% | 2.9%
|
| Outer | 3.6%
| 0.5% | 2.2%
|
| Thameslink | 4.3%
| 3.9% | 4.1%
|
| Inner | 9.7%
| 8.6% | 8.2%
|
| Outer | 3.0%
| 3.2% | 3.1%
|
| West Anglia Great Northern | 2.3%
| 1.6% | 2.0%
|
| Inner | 1.8%
| 1.7% | 1.8%
|
| Outer | 2.7%
| 1.4% | 2.1%
|
| London and south east: total | 5.0%
| 1.7% | 3.6%
|
2.11 The SRA's bulletin On Track, in which these
results are reported, records that "there are now more than
50 additional trains running into London every morning compared
with the figure in 1996, when 809 services ran." But the
report does not indicate whether or not total train capacity has
grown by the same proportion, or how this growth compares with
the change in passenger numbers. In fact, at the aggregate London
and south east level, since the final year of BR operation the
PIXC result has risen by 2.4% for the morning peak and 0.5% for
the evening peak. Nor does the bulletin mention any particular
steps that the operators which are in breach of the planning thresholds
are being required to take to mitigate the situation.
2.12 There are no published data on the incidence of
crowding at stations on the National Rail network. Operators are
not required to carry out any census of station usage.
2.13 London Underground has adopted a statement of its
Customer Service Delivery Standards, in which it aims to
"identify the constituents of a customer-facing service and
define the quality of delivery required in each service area to
meet customers' needs and priorities". One of the underlying
themes is stated to be "the aim to minimise potential stress
by offering a pleasant, safe and calm travel environment."
The standards are very much more comprehensive and prescriptive
than any requirement placed on the franchised train companies,
but are purely aspirational in character as they have no mandatory
force.
2.14 The standards lay down a minimum average service
interval of five minutes over the central part of the network,
and go on to state that "service frequencies and patterns,
when new or refurbished rolling stock is being introduced, must
be sufficient to obviate the need for any customer to stand for
more than 15 minutes. Seating for at least 1/6 of the maximum
capacity of each car must be provided. Any major upgrade should
take account of the fact that the existing standard for train
loadings is often unavoidably exceeded on some sections, and should
have congestion management as a key objective." An internal
reference manual, containing a commentary on these standards,
notes that "Crowding is a complex issue, influenced by service
volumes, train capacity, reliability, and demand. Significant
improvements in the on-train crowding penalty will only be possible
through major upgrade projects." The requirement for seating
to be provided for only 1/6 of passengers does not include perch
or tip-up seats, and relates to the theoretical maximum crush-
loading capacity of the trains, rather than anything which would
be tolerable in normal operating conditions. It is noteworthy
that significant congestion relief is seen as attainable only
in the context of "major upgrade projects", rather than
through any change in day-to-day operating practices.
2.15 London Underground conducts regular monitoring of
train loading levels during weekday peak periods, the results
of which are expressed as a statistic showing the percentage chance
of being on a train with a given ratio of passengers to seats.
Prior to the creation of Transport for London, these data were
routinely published. This appears no longer to be the case, although
they are available on application. This table shows the results
for the period July to September 2002.
|
| % chance of being on a train with:
|
| Line | all seats full
| at least
1 person standing
for each 1 sitting
| at least
2 people standing
for each 1 sitting
|
|
| Bakerloo | 18
| 4 | 0
|
| Central | 69
| 15 | 1
|
| Circle & Hammersmith | 84
| 45 | 4
|
| District | 23
| 12 | 2
|
| East London | n/a
| n/a | n/a
|
| Jubilee | 68
| 9 | 1
|
| Metropolitan | 16
| 8 | 1
|
| Northern | 72
| 7 | 0
|
| Piccadilly | 50
| 4 | 1
|
| Victoria | 54
| 19 | 1
|
| Waterloo & City | 95
| 83 | 4
|
| All lines | 52
| 16 | 1
|
|
2.16 The table shows that at peak times the majority
of Underground passengers are required to travel standing. But
the figures are averages which do not necessarily reflect the
situation on the most heavily used sections on each route, and
this can fluctuate markedly from minute to minute depending on
the evenness of the intervals between trains. Assuming that passengers
enter platforms at a constant rate, any significant delay will
cause excessive loading on the next train, and this in turn will
result in under-loading of the train following that. The distribution
of platform entry and exit points can also cause uneven loadings
between individual carriages on the same train.
2.17 There are wide variations between lines, but it
should be noted that average journey lengths differ markedly depending
on the geography of the route. What may be tolerable on the Waterloo
& City line, for a maximum journey time of four minutes, is
unlikely to be acceptable on the District or Metropolitan lines
where journeys can last more than an hour. The design of the rolling
stock varies between the lines, so that on the Circle and Hammersmith
& City lines (where loadings are high but most journeys are
short) there is a much higher proportion of standing space to
seats, and for any given passenger loading, the percentage chance
of all seats being occupied is automatically higher. These data
also take no account of the turnover of spaces within trains.
If the number of passengers alighting at a succession of stations
is equal to the number boarding, the aggregate loading level will
be unchanged, but any individual passengers obliged to stand involuntarily
at the start of their journeys will not necessarily remain standing
for the whole of the trip. Although London Underground has set
a target maximum of 15 minutes, the actual length of standing
times experienced is not reported.
2.18 In relation to stations, London Underground stipulates
that "Minimum dimensions and station capacity requirements
are described in station planning standards. These are to be observed
in respect of ticket halls, routeways (including stairs) and platforms.
Capacity limitations at stations are kept under review and plans
developed to address changes in demand." These standards
have been applied at the very limited number of new stations which
have been constructed in recent years, notably on the Jubilee
line extension from Green Park to Stratford, whose spaciousness
is in striking contrast to most of the rest of the network. But
London Underground still has many stations built to the standards
in use a century ago, and it is the lack of access and egress
capacity, as much as a lack of train capacity, which is the primary
cause of congestion and delay for a large proportion of its passengers.
2.19 Regular users of stations such as Victoria and Oxford
Circus have long since become accustomed to the need for "station
control" to be imposed at busy times, when access to the
station is denied until congestion at platform level eases. In
the straitened financial circumstances in which the organisation
has been compelled to operate in recent years, little or no funding
has been available to address this. While staff are enjoined "to
encourage passengers to move to less crowded sections of station
platforms" by means of notices and announcements, at the
busiest stations at the busiest times there is little or no underoccupied
platform space available. Published details of the outputs which
will be required of the "infraco" consortia which will
become London Underground's partners under the Government's PPP
plans for the network are scarce, but it appears that the anticipated
capacity enhancements relate primarily to the train service rather
than to stations. If the PPP is successful in raising service
quality, and in generating additional demand as a result, the
under-capacity of the busier stations will become ever more acute.
2.20 At one time, London Buses regularly monitored and
published statistics relating to "the percentage chance of
boarding the first bus to arrive." This survey was abandoned
several years ago, and there is now no systematic tracking of
the incidence of crowding on buses. Loadings are checked as part
of the "key points survey" conducted on all major routes,
but this is only undertaken at 18-month intervals. At other times,
the organisation appears to rely on ad-hoc feedback from users
and/or operators, plus occasional "mystery traveller"
reports, to alert it to the extent of under-capacity both on vehicles
and at bus stations.
3. DOES OVERCROWDING
MATTER?
3.1 Self-evidently, uncomfortable travel conditions are
a deterrent to the use of public transport. Those who have the
option of travelling by other means will do so, unless rail or
bus offer substantial advantages in terms of time and/or cost.
The persistence of overcrowding on parts of the public transport
network indicates that at busy times it is catering for a largely
captive market, at least in the short term, and that those who
endure these conditions do so because the perceived value of the
journey exceeds the perceived cost, including the element of discomfort.
The fact that well over 80% of people entering central London
every day do so by rail or bus (a higher proportion than in virtually
any other major city in an OECD country) does not imply that they
regard the experience as satisfactory, let alone enjoyable, but
simply that there is no practicable alternative mode available.
Even if central London's roads could accommodate more people travelling
by car, there are no legal or affordable parking spaces available
for their use. Providing extra capacity (in the form of tracks,
vehicles and staff) is costly if it is only required for a small
proportion of the time—many trains on the London and south
east network already make only one fully laden journey each way
per day. So there is little or no commercial incentive for operators
to take mitigating action.
3.2 When passengers complain about overcrowding, they
commonly assert that they regard it as a threat to their safety,
particularly on the railways. Trains are built to carry as many
people as can be physically accommodated on them, and their braking
and suspension systems are designed to function correctly under
maximum crush-load conditions. There are no instances known to
LTUC of crowding having led directly to an accident. So such concerns
generally relate to the number of people potentially at risk in
the event of an accident (such as a collision, derailment or fire),
particularly if they are standing, and to ease of evacuation or
escape.
3.3 This concern has found expression from time to time
at inquiries arising from major railway accidents. For example,
the "Hidden Inquiry" into the Clapham Junction collision
in 1988 took evidence to the effect that "the severity of
injury, or the risk of fatality, was no greater for standing than
seated passengers." The report concluded, however, that "the
fact is inescapable that the higher the number of passengers on
a train, the higher the number of casualties is likely to be in
absolute terms." It therefore contained a recommendation
that "BR shall ensure that overall train loading criteria
are achieved. The Department of Transport and BR shall keep the
criteria under review." In practice, the criteria then in
use were the same as the threshold values subsequently incorporated
into the franchise specifications set for privatised operators,
as set out in paragraph 2.8 above. On substantial parts of the
network, they were never achieved by BR and they have not been
achieved by its successors either.
3.4 The inquiry which followed a buffer stop collision
at Cannon Street in 1991 commissioned some computer modelling
work on the relative levels of risks faced by seated and standing
passengers. Its broad conclusion was that in the case of collisions
at speeds exceeding 15 mph, both groups would be likely to be
thrown around uncontrollably, and would be equally exposed to
a number of potentially injurious impacts. At lower speeds, standing
passengers might be at greater risk because they were more likely
to lose their balance and fall in an uncontrolled way. But much
would depend on the rate of deceleration and whether the movement
of the train gave any forewarning of the impending collision.
The inquiry concluded that although there were no safety grounds
for prohibiting passengers from standing on a train, "care
needs to be exercised in the way standing passengers are conveyed".
It noted in particular that "with a free choice many passengers
will choose to travel in the front coaches even if this means
having to stand" in order to be able to exit more quickly
on arrival at the terminal station. The report therefore contained
a recommendation that BR should examine what measures could be
introduced to distribute passengers more evenly along the train,
such as "better advice to intending passengers of the available
space . . ., arranging access points to platforms away from the
front of trains, and making arrangements at terminal stations
. . . to remove the perceived advantage to those at the front
of the train." Whatever examination BR may have made in the
light of this recommendation does not appear to have led to any
subsequent action.
3.5 The "Cullen Report" into the Ladbroke Grove
collision in 2000 also touched briefly on these issues. It noted
approvingly that the train company First Great Western was "considering
the possibility of dissuading passengers from moving forward in
the train shortly before arrival at the terminus" but made
no specific recommendation on the matter. Like the Clapham Junction
and Cannon Street inquiries before it, the Ladbroke Grove inquiry
heard suggestions that seat belts might be fitted to trains, but
like them it declined to endorse such a proposal. The safety benefits
were open to question, and even if they were real, it was felt
to be impracticable to make the use of belts obligatory on trains
used for very short journeys. To the extent that any real difference
existed, seated passengers were already at less risk than those
obliged to stand (or who chose to do so).
3.6 The available evidence from accidents prior to that
date was reviewed in a 1999 report for the Health & Safety
Executive on Implications of overcrowding on railways.
This confirmed that "in higher-speed collisions, whether
a passenger is seated or standing makes little difference to the
overall severity of injuries sustained". Although seated
passengers may be less likely to sustain serious injuries in lower
speed collisions, "in overcrowded conditions . . . the situation
may be somewhat more complex". This was because while a large
number of standing passengers may have a cushioning effect which
would protect them against some forms of injury (such as those
sustained by hitting carriage fittings or other passengers), it
was possible that other injuries such as head-to-head collisions
would be more prevalent. Overall, the report concluded that "there
is no evidence to suggest that overcrowding per se is a
safety issue." But it should be noted that this was purely
a desk study, and no original research was undertaken in the course
of it.
3.7 The HSE report endorsed the view that any policy
requiring all passengers to have seats in order to travel, and
to remain in them while the train is moving, would be impossible
to implement and police. It also noted that if "excess"
passengers were removed and put on an extra train, there might
be "an increased risk to the total population of rail travellers
simply due to the addition of the extra trains required to deal
with the excess passengers." This comment is based on the
apparent assumption that there may be a positive relationship
between the intensity of traffic on a railway and the level of
risk faced by its users. This is a highly debatable issue, giving
rise to complex questions about (eg) the performance of train
control systems and/or the rate of asset deterioration and maintenance,
which are outside the parameters of this letter. What can be said
with some certainty, though, is that if the effect of overcrowding
on railways is to displace some travellers onto other modes, then—except
in the case of those who might go by air—the net effect
will almost certainly be to increase the overall level of casualties,
because all other modes entail significantly higher levels of
risk.
3.8 Perhaps surprisingly, none of these reports addresses
the extent to which ease and speed of escape or evacuation in
the event of an accident is likely to be affected by the number
of passengers on board—possibly because none of the actual
accidents reviewed involved trains that were particularly heavily
loaded. The only recent incident of which LTUC is aware in which
this issue has arisen in practice occurred on the Underground's
Victoria line in July 2001, when some 4,000 passengers were trapped
in three trains held in a small-bore tunnel for up to 90 minutes
as a result of malicious interference with train equipment. The
only ventilation on such trains is provided by the movement of
the trains themselves, so rapid build up of body heat can occur
in heavily-loaded carriages held stationary for long periods.
Because this incident took place on one of the hottest days of
the year, ambient temperatures were high, and 17 people were taken
to hospital suffering from the effects of heat exhaustion. But
trains do not have to be heavily crowded, or in tunnels, for similar
problems to arise. Much of the newer rolling stock on the main
line railways has air conditioning, and no openable windows in
the passenger saloons. When electric trains are held stationary
because of power failures, the air conditioning is rendered inoperative,
as the demands it would place on emergency battery supplies are
excessive. The result, on warm days, is that conditions on board
can rapidly become extremely uncomfortable. The industry appears
to have no solution to this problem.
3.9 The HSE report makes some mention of research which
has been conducted into other effects which overcrowding may have,
of a psychological and physiological nature, which manifest themselves
through their impact on the health of those affected, rather than
their physical safety—though they may have indirect safety
consequences if, for example, they lead to an increased level
of violent conduct. Although these effects have been studied in
the behaviour of crowds in other contexts, very little research
appears to have been undertaken into the health effects of the
various sources of stress to which public transport passengers
are routinely exposed. The current, very restricted state of knowledge
in this field is demonstrated by a recent (and as yet unpublished)
study of Rail passenger stress and health, commissioned
by the Rail Passengers Council from the Institute of Work, Health
and Organisations at the University of Nottingham.
3.10 LTUC is not aware of any published work relating
specifically to the risks associated with congestion at stations.
Under paragraph 71 of the Reporting of Injuries, Diseases and
Dangerous Occurrences Regulations 1995, station operators
are required to notify the HSE of incidents of serious congestion,
defined as "Any case where planned procedures or arrangements
have been activated in order to control risks arising from an
incident of undue passenger congestion at a station unless that
congestion has been relieved within a period of time allowed for
by those procedures or arrangements." But the term "undue
passenger congestion" is not defined, and the annual report
of the HSE's Chief Inspector of Railways makes no mention of the
frequency with which such cases are reported, or of the circumstances
which give rise to them.
3.11 The Public Service Vehicles (Carrying Capacity)
Regulations 1984 require maximum seating and standing capacities
to be set for buses and coaches, and it is an offence to allow
these to be exceeded. LTUC understands that (except in the case
of buses operated with conductors) the primary purpose of these
regulations is to ensure the stability of moving vehicles. A report
by the London Accident Analysis Unit on Public service vehicle
accidents in Greater London showed that in 1991 standing passengers
accounted for 42% of all reported casualties to bus and coach
users. Since it is likely that on most buses the proportion of
standees is less than this, it is reasonable to conclude that
the risk of injury is higher for those who are standing than for
those who are seated. It is our impression that the frequency
of claims made by passengers against operators for minor injuries
sustained in the course of journeys, eg as a result of sudden
braking, has been increasing. But it is not clear whether this
is the result of a higher incidence of such events, or a greater
propensity on the part of passengers to seek compensation as a
result of them. If fault cannot be shown to lie with the driver,
or with another road user who is required to carry insurance—eg
where the bus is forced to brake to avoid colliding with a pedestrian,
cyclist or animal—the risk is likely to be uninsured and
any victim to remain uncompensated.
4. HOW CAN
THE PROBLEM
OF OVERCROWDING
BE ADDRESSED?
4.1 There are no simple or cheap solutions, particularly
on the railways. Trains cannot be made wider, or higher (to allow
double deck operation), without immensely costly alterations being
made to railway infrastructure—platform edges, tunnels,
bridges and overhead wiring. They can be made longer, but this
often necessitates the lengthening of platforms (and sometimes
sidings), which is also expensive and sometimes physically impossible.
Their maximum speed is set by such factors as the curvature of
the line, the capacity of their traction units, the number of
powered axles, and (in the case of electrified lines) the available
power supply. On routes with closely-spaced stations, such as
the Underground and much of National Rail's London and south east
network, the effective overall speed is set by the spacing of
the stops and the maximum tolerable rates of acceleration and
deceleration.
4.2 On-board capacity is controlled by the number and
distribution of seats provided, and the ease of circulation within
the train. Metros and suburban railways in other countries typically
have only a limited number of seats, arranged longitudinally,
and it is assumed that most passengers will travel standing. The
removal of seats is not popular where passengers have been accustomed
to having them, but a recent consultation exercise by Connex suggested
that many of its passengers would accept this if it could be shown
that, in the short term, it was the only means of allowing them
to continue to board the train of their choice. Wider aisles and
a more generous provision of handrails encourage passengers not
to cluster in vestibules, thus making better use of the space
available. Wide inter-carriage connections (similar to those found
on articulated trams and buses) would encourage passengers to
move between carriages, distributing themselves more evenly along
the train, and would have the incidental benefit of mitigating
concerns about personal security at times of low usage. But the
rail industry in Britain has been reluctant to embrace this feature,
although it is commonplace on metros overseas. This appears to
be due to concerns about cost, crashworthiness, the spread of
smoke in the event of fire, and the difficulty of draught-proofing
cabs when passages are created through them to allow units to
be coupled.
4.3 In the past, all trains were provided with guards'
compartments which were also used to carry passengers' luggage
and cycles. Some also carried buffets on board. Capacity has been
increased by removing these, albeit at some loss of passenger
amenity. On the National Rail network, first class accommodation
is provided at the discretion of the operators, and excluded from
PIXC calculations. It has lower seating density, and (typically)
a lower occupancy rate. One operator, Chiltern, has recently decided
to discontinue it. Underground trains do not have toilets on board,
but generally these have been provided on National Rail services,
except on some trains used only on shorter-distance inner-suburban
routes. Connex has canvassed opinion on the option of eliminating
them from some new rolling stock, to enhance capacity, with the
proviso that the quantity and quality of toilet facilities on
stations would be increased instead.
4.4 On high-capacity routes with frequent stops, "dwell
time" at platforms represents a significant element of journey
time, and thus of route capacity. The time required for boarding
and alighting is a function both of the total number of passengers
carried, and of the length of their journeys (ie the numbers using
each station). As trains become more crowded, movement within
them becomes more restricted, and passengers take longer to enter
and leave, thus prolonging each stop—especially if some
are obliged to alight involuntarily simply to allow others behind
them to exit. The rate of boarding and alighting can be made faster
by providing more (and wider) doors and larger vestibules, but
this is necessarily achieved at the expense of seating space.
4.5 Even if the planned capacity of a service is sufficient
to meet demand, any shortfall or irregularity in its delivery
can result in a rapid build-up of overcrowded conditions. Poor
reliability plays a major part in exacerbating the problem. In
the period from April to June 2002, 0.9% of planned longer distance
trains to or from London did not run, nor did 1.2% of those on
the London and south east network. But, in addition, 20.9% of
longer distance services reached their terminus more than 10 minutes
late, and 15.6% of those on the London and south east network
terminated more than five minutes late. On the Underground, the
overall incidence of delay is measured by "excess waiting
time", ie the average amount by which passengers' actual
waits exceed those they would experience if the service ran exactly
as planned. In the year 2000-01, the average waiting time experienced
across the network was 3.2 minutes, but the excess element of
this was 0.8 minutes. In other words, passengers were typically
obliged to wait one third longer than they should, if the full
level of service was operated, and delivered on time. The causes
of this unreliability are outside the parameters of this discussion,
but they are a significant contributory factor in the overall
incidence of crowding.
4.6 Even if additional rolling stock is available, or
can be provided, it is not necessarily possible to increase the
overall frequency of service, since this is governed by the capacity
of the route—which is in turn governed by such factors as
the spacing of signals, the number of stations and platforms,
and the number and layout of junctions and crossovers. In general,
the more complex and variegated the service pattern, the lower
the number of trains that can be run. The greatest utilisation
of capacity is usually obtained where all trains run between the
same destinations and observe the same stops—a situation
which is rare in London, even on the Underground (because higher
frequencies are operated over the sections with greater levels
of passenger demand). Requiring fast and slow trains to operate
on the same tracks (as occurs on parts of the National Rail system)
invariably reduces the frequency of both. And timetabling services
at or close to the theoretical limit of capacity is unwise, because
it eliminates the necessary "recovery" margin which
is essential if any delay to one train is not to result immediately
in a cascade effect which repercusses on the timekeeping of all
following trains. Indeed, experience on parts of the Underground
has shown that overall reliability—and thus evenness of
passenger loadings between trains—can be improved by removing
trains from the timetable at the busiest times, rather than adding
them, when the planned schedule has proved itself to be over ambitious.
4.7 If demand is not evenly distributed between trains
on a route, passengers can be reallocated from overcrowded to
undercrowded trains by varying the stopping pattern, so that not
all trains serve the same stations. This is very effective in
redistributing loads (provided that passengers do not change their
choice of stations in response). But it makes local journeys between
intermediate points more difficult if they are no longer served
by the same trains, and is a technique which is therefore best
used sparingly and selectively.
4.8 Finally, fares can be adjusted to "choke off"
excess demand, or to redistribute it to times and/or routes when
and/or where pressure on capacity is less. This technique has
long been employed by the railways, eg by offering discount fares
for off-peak travel. In the era of BR, prices were deliberately
raised as a device for curbing the unwanted growth of peak period
traffic. But fears that privatisation might lead to excessive
profiteering by train operators catering for a captive market
caused a price capping regime to be introduced for certain key
fare categories, including the standard class season tickets which
are most likely to be used by commuters travelling at the busiest
times. Pegging these fares to a level below the prevailing rate
of price inflation, and imposing additional fares reductions as
a penalty for poor reliability, has made the cost of rail travel
cheaper than might otherwise have been the case. This is a welcome
development, from the users' perspective, where the system has
sufficient spare capacity to cater for the additional demand thus
attracted. In principle, we favour any policies designed to make
the use of the railways more attractive. But they have a perverse
effect where overcrowding is already acute. In the London area,
the differential pricing of bus and rail travel has had some limited
success in diverting trips from rail to bus, but the two modes
are not serious alternatives for any but fairly short journeys,
and the zonal pricing basis of the Travelcard (which has largely
replaced point-to-point season tickets) means that it cannot be
used to deflect demand from one route to another.
4.9 On the buses, there are far fewer constraints to
increasing capacity in line with demand, either by enlarging the
size of vehicles or increasing their frequency. Both the volume
and usage of bus services in London are growing steadily at present.
But reliability remains problematic: on the higher-frequency routes,
actual average waiting times are currently about 40% longer than
planned. The principal obstacle to increased reliability is the
traffic environment in which the bus service is required to operate.
The Mayor's London Bus Initiative is designed to achieve much
more extensive bus priority measures on major corridors, coupled
with more systematic policing and enforcement. But there is a
looming shortage of depot space to handle the growing size of
the bus fleet, and in a buoyant labour market the cost of recruiting
and retaining sufficient staff of the required calibre (both in
operating and engineering roles) is rising, particularly as housing
costs escalate. As routes are re-tendered, operators' bids are
rising steeply, and with fares frozen in money terms the subsidy
requirement is escalating at a rate which the Greater London Authority
may find difficult to fund. Unless the planned increase in capacity
can be sustained, however, particularly in the central area, the
success of the Mayor's congestion charging strategy may be compromised—because
this acknowledges that there is no realistic prospect of being
able to handle many additional peak period travellers on London's
railways, and if the strategy is successful in inducing some diversion
of travel away from cars, much of this demand will have to be
met by the bus network.
December 2002
|
