Chapter 5
Transport costs
Learning Outcomes:
In the course of this chapter, you will learn about:
᭿ The theory of production
᭿ Deﬁnitions of technical, cost and allocative efﬁciency
᭿ How time is divided in economics
᭿ How transport costs behave in the short and long runs
᭿ Returns to scale and economies of scale in the transport industries
᭿ The relevance of production costs to the supply of transport services.
INTRODUCTION
Whilst elasticity of demand impacts upon the demand side of the market, a major factor aﬀecting
the quantity supplied is the cost of production. Transport costs however fall into a variety of
diﬀerent classes. There are those costs that impact on the individual user of a particular mode
of transport who directly beneﬁts from undertaking a journey. These are known as private costs
and would include both the ﬁnancial costs involved, such as the fare in the case of public transport,
as well as non ﬁnancial costs, such as the time involved in undertaking the journey. Taken together
these are known as the ‘generalised cost’. There are then the costs of transport that fall on non
users of the transport service who do not beneﬁt from that transport service. This includes what
could almost be termed the unwanted by-products from the undertaking of transport activities,
such as polluted air, the congested road, noise and visual intrusions. These are generally referred to
as public costs, and as these are a signiﬁcant factor in the provision of transport services these will
be examined in some depth in various chapters later in the text. Finally there are production costs
that fall on the operators of a transport service or in the case of private transport the ﬁnancial costs
incurred when undertaking the activity. In many ways these are essentially private costs, as the
individual that incurs the cost (the operator/road user) is the one that beneﬁts from the provision
of that service (i.e. proﬁt/beneﬁt from the journey). This chapter is speciﬁcally concerned with
costs that fall into this last group, particularly in the production of public transport services. Road
user issues are examined later in Chapter 8 under pricing.
95
Public transport costs have received a large amount of attention over the years, as these services
are a vital component of the economy and society and have been to a greater or lesser extent
subsidised by national and local governments. In simple terms, high cost levels restrict transport
authorities’ options and constrain the level of public transport services that can be provided. Even
where such services are not subsidised and are provided by proﬁt-making private companies, the
constraints placed on supply arising from costs still apply: you should remember from Chapter 3
that the cost of production is one of the main determinants of supply. Following that logic, if costs
can be reduced more transport services can be provided, i.e. supply can be increased. In subsidised
markets, of course, lower costs reduce the need for subsidy.
Attempting to reduce and maintain downward pressure on public transport costs has therefore
been a main concern of government policy. This has been one of the main reasons for on-going
moves by authorities to shift transport operators more towards ‘the market’ rather than operating
as state owned and controlled public enterprises. This move has resulted in a substantial shift from
publicly owned operators towards privately owned companies and the introduction of market
principles, most visibly, through competition in one form or another in the supply of transport
services.
The actual size of costs and authorities’ attempts to keep them as low as possible however are not
the only issues surrounding costs. How costs are incurred is also important. For example, in many
instances there may be a large initial overhead incurred by the ﬁrm but the actual cost of providing
one extra service or carrying one extra passenger may be very small. For example, the extra cost to
a bus or train operator of transporting an additional passenger, if they are operating at less than full
capacity, is negligible. Alternatively, there are other cases where there is a very low initial overhead
but most costs are incurred as a direct result of providing the transport service. Even at a more
basic level, are large companies more cost eﬀective than small companies or does an ‘optimum’
size exist, i.e. not too big and not too small? There may for example be certain advantages to being a
large-scale operation that leads to certain cost savings; however, in other cases the sheer size of the
ﬁrm may cause certain ineﬃciencies to arise. All of these issues will heavily inﬂuence the structure
of the transport sector and determine how transport services are provided to the market.
THE EFFICIENT PRODUCTION OF TRANSPORT SERVICES
The costs of transport operations are primarily dependent upon a combination of the production
processes used and the eﬃciency of the management of that production process. Not only that,
however, but the physical characteristics of the operating environment in which transport services
are provided will also impact upon costs. The start/stop nature of urban bus operations due to high
densities of population, for example, tend to make such an environment more costly to provide
public transport services per vehicle kilometre than in rural areas – DfT statistics (DfT, 2007) for
example show the cost per bus kilometre to be some two and a half times higher in London than
for the rest of England outside of the metropolitan areas. The reason for this is the considerably
longer running distances between stops and overall faster running speeds in rural areas. There is
little that transport management can do however about these aspects of operation. In terms of
controllable costs, therefore, it is only the production process that is under the direct control
of management, and consequently is the main determinant of costs to the operator.
96
TRANSPORT COSTS
It may be diﬃcult to reconcile transport operators with the idea of the traditional ﬁrm, where
inputs are fed in at one end of the factory and ﬁnished goods emerge from the other. The main
diﬀerence of course is that transport is a service, and thus production is concerned with producing
‘service units’. In order to measure the output of the transport ﬁrm, therefore, such ‘service units’
need to be quantitatively deﬁned. This ﬁrst raises the question of what exactly do transport
operators produce, which is not as easy a question to address as it may ﬁrst seem. What needs to be
considered is what are transport operators actually attempting to achieve. The answer may appear
to be obvious; a public transport operator, for example, simply moves people from A to B. Output
would therefore be measured in terms of the journeys undertaken, either directly by the number
of journeys or by the total number of passenger kilometres to also account for diﬀerences in
distances travelled. Strange as it may appear, however, in many cases moving people from one
location to another is not the aim of a public transport operator. Many instances exist where
operators are contracted by transport authorities to provide transport services within a given area
or location, irrespective of the number of people moved. Rather than journeys, therefore, the
output would be measured in terms of the number and level of services produced, and hence could
be expressed in terms of vehicle kilometres. How, therefore, should the output of the transport
ﬁrm be measured, by journeys or by vehicle kilometres produced?
These two diﬀerent outputs can be equated by the idea that what operators attempt to do is
move people from A to B, but how this is achieved is through the production of vehicle kilometres.
The idea of producing vehicle kilometres is far more consistent with the traditional view of
production, but this whole topic needs to be related to the aims of the producer, which you may
remember is one of the determinants of supply. These can vary from proﬁt maximisation in a
completely free transport market where revenue is only related to the carriage of passengers, to
sales maximisation in a tightly regulated market where the transport company is directly paid
to operate the service and hence proﬁt maximises by attempting to ﬁll the available capacity.
Passenger journeys as the output would be more consistent with the former as revenue is only
connected to the carriage of passengers, whilst vehicle kilometres speciﬁed as the output would be
more consistent with the latter, as the ﬁrm is paid to provide services and hence proﬁts are
maximised through ﬁlling the available capacity, i.e. sales maximise. What is important in this
context is that there is some measurable output at the end of the production process, whether that
be the number of journeys or the number of vehicle kilometres produced, and this will largely be
dependent upon the type of market the ﬁrm is operating in.
You should already be aware of the concept of production; however, in this context it speciﬁcally
relates to the process whereby economic inputs or resources in the form of the factors of
production (recall from Chapter 1 that this relates to inputs in the form of land, labour and capital)
are brought together by entrepreneurs (or ﬁrms) in order to produce goods and services. This
basic process is outlined in Figure 5.1.
As illustrated by the ﬁgure, the production process consists of converting the inputs of Land/
Raw Materials, Labour and Capital, by way of a combination process, into the outputs of Transport
Services and By-Products. Transport Services are then sold for an economic return, with the aim
being that the transformation process should convert the inputs into something of a higher
economic value. The revenue generated from the sale of the output therefore should be greater
than the payment for the inputs. Also shown in Figure 5.1 is ‘By-products’. Not all of the output of
the combination process is sold for an economic return, as the process also results in the output
97
TRANSPORT COSTS
of other factors such as wastage or pollution. Often overlooked, this has to be included in the
outputs of the production process and has important implications for transport markets.
To directly relate Figure 5.1 to transport industries, ‘land’ is an important factor in the
production process, but of the three it is the one that is least under the control of transport
operators. This is because land in many cases relates to the ‘suitability’ of the prevailing geographical
environment to the provision of transport services. This was exempliﬁed above by the
urban/rural split, although many other examples exist. The provision of rail services, for example,
will be relatively easier in areas of ﬂat terrain (such as many parts of Belgium) rather than in
mountainous regions (such as many parts of Switzerland). This is because the latter will require a
far higher level of tunnels and bridging,i.e.less ‘land’ and more capital,in the production of services.
The other two factors of production are far more straightforward and more under the direct
control of the ﬁrm. Put simply, labour relates to all staﬀ involved in the production of transport
services, whether that be operational or administrative staﬀ, whilst capital relates to any goods that
have been manufactured in order to be put into the production process. This obviously not only
includes the vehicle stock, but also any other physically made equipment, e.g. terminal buildings,
infrastructure, bridges, tunnels, handling equipment, depots and IT facilities.
The transformation process outlined above is known as production, and the relationship
between the level of inputs and the level of the outputs achieved known as the production function.
This is formally speciﬁed below:
Q = f(A, L, K)
Where:
Q = quantity of output produced
f = ‘some function of’
Figure 5.1 The production process
98
TRANSPORT COSTS
A = quantity of land and raw materials used in the production process
L = quantity of labour used in the production process
K = quantity of capital used in the production process
This equation simply states that the level of output is some function of land/raw materials, labour
and capital, with the letters L and K always used in economics as the shorthand notation of these
two inputs. Where an increase in the inputs leads to an increase in the outputs, which should
virtually always be the case, then production would be said to be monotonic. Whilst appearing to
be common sense, this is actually an important theoretical consideration as real life data does not
always ﬁt with what would appear to be a very basic theoretical concept. The inputs/outputs ratio
is also one of the main bases for assessing whether a given operation can be described as ‘eﬃcient’
or not. The idea of eﬃciency in the production of public transport services is an important
concept, and as noted above has received considerable attention in the academic literature.
‘Eﬃciency’ however is an often over-used term and has diﬀerent meanings to diﬀerent people.
Within the ﬁeld of transport, it has been used in the past to describe issues such as reliability,
punctuality, pricing, costs, subsidy levels, number of passengers carried and so on. Within economics,
however, there are only three basic ‘types’ of eﬃciency although these may be seen
elsewhere under slightly diﬀerent terms. However, they are:
Technical eﬃciency – this relates to the outputs to inputs ratio, with a technically eﬃcient
operator being one that uses the minimum level of inputs to produce the maximum level of
outputs. Alternatively, this may be achieved where the minimum level of inputs is used to produce
a given level of output. Both measures are highly relevant in the study of transport industries, as in
many transport markets the output level is set by a transport authority and hence for the operator
the technical eﬃciency question is one of input minimisation.
Cost eﬃciency – sometimes referred to as productive eﬃciency, or even cost allocative eﬃciency,
cost eﬃciency arises because there may be several diﬀerent ways to produce the output, all
of which would be technically eﬃcient. For example, a high level of capital and a low level of
labour could be employed, or alternatively a high level of labour and a low level of capital
employed. Both production processes may be technically eﬃcient. The issue therefore becomes
which one is the ‘best’. In order to answer this question, the relative prices of labour and capital
are examined and the one that produces the lowest cost combination deemed to be the ‘best’ or
more exactly the most cost eﬃcient.
Allocative eﬃciency – even within the study of economics, there is much confusion over the term
‘allocative eﬃciency’, which is quite surprising given it is almost the holy grail of the economics
discipline! Allocative eﬃciency however relates to usage. As many of the former Eastern European
countries showed in the past, there is little point in producing goods and services that are
technically eﬃcient in the lowest cost combinations if no one wants them. This is a total waste of
resources and hence could never be considered as eﬃcient. Allocative eﬃciency is therefore said to
exist where goods and services are produced cost eﬃciently and in the ‘right’ quantities. Where
this exact position is found is where the price paid for the good, which can be used as in indicator
of the additional beneﬁt received from that good, equals the extra cost of producing that good.
This however is an issue which will be returned to in greater depth in the following chapter.
The ﬁrst two of these eﬃciency concepts can be illustrated graphically, which is done in
Figure 5.2.1
99
TRANSPORT COSTS
On the left, Figure 5.2 (a) illustrates the combinations of labour and capital required to
produce a single unit of output. Consequently, points nearer the origin are more eﬃcient, as
relatively speaking less of the inputs are used in the production process. Also shown are the relative
positioning of four hypothetical ﬁrms, A to D. In this example there exists a large range of
combinations of the two inputs that can be used; this varies from a high element of labour and a
low amount of capital (ﬁrm A), to a high level of capital and low amount of labour (ﬁrm C). Whilst
this may appear to have little relevance to transport services where the relative quantities of labour
and capital are unlikely to vary greatly across ﬁrms in the same industry, some variation does exist
and this is often the situation with regard to the carriage of people and freight across a number of
diﬀerent modes. These tend to employ varying levels of capital and labour in order to achieve the
same purpose. In Figure 5.2 (a) ﬁrms A, B and C have the lowest possible combinations of labour
and capital, thus an eﬃciency frontier can be drawn between these points. Consequently points
above that ‘frontier’ would represent ineﬃcient ﬁrms and points below simply unobtainable with
the level of today’s technology. You should hopefully recall that this is identical to the idea of the
production possibility curve introduced in Chapter 1, only expressed in a diﬀerent way: in this case
the focus is on minimising inputs rather than maximising the output. The frontier is curved for
theoretical reasons that surround variations in the way that one input is substituted by the other
across diﬀerent levels of the inputs, known as elasticities of substitution. As ﬁrms A, B and C
outline the actual technical eﬃciency frontier, all three would be deemed to be technically
eﬃcient. This is because there is no way of assessing which is the ‘best’ combination of the two
inputs – technical eﬃciency is simply concerned with the lowest combination of units used in the
production process, not whether this represents the best ‘mix’ of the inputs to use. Firm D
however lies above the frontier and thus would be said to be technically ineﬃcient. This is because
D is using a higher amount of capital than ﬁrm A and also a higher level of labour than ﬁrm B in the
production process.
As regards the best combination of labour and capital to use in the production process, this
can only be assessed once the prices of the individual factor inputs are introduced into the
evaluation. Assuming that all ﬁrms within the industry face identical cost conditions, this is done in
Figure 5.2(b). Added to the ﬁgure are budget lines, which are linear combinations of the costs of
employing labour and capital and are drawn as straight lines out from the origin. These are straight
Figure 5.2 Technical and cost efﬁciency
100
TRANSPORT COSTS
because the combination of prices of the inputs are ﬁxed, hence the cost of substituting one input
for another is ﬁxed over the various combinations of capital and labour that can be employed. The
slope is determined by the relative prices of the two inputs of labour and capital.
The ﬁrst ‘viable’ budget line on Figure 5.2(b) is line II, as this is the ﬁrst that is tangential to
the technical eﬃciency frontier. Firm B therefore has the lowest cost combination of inputs, as it
lies both on the technical eﬃciency frontier and the lowest budget cost line. Note that in the
case of the two other technically eﬃcient points illustrated in Figure 5.2(a), ﬁrm C is on a
higher budget line and hence would have a higher cost combination, and point A would lie on
a higher budget line again.
These are the basic principles concerning the eﬃcient production of transport services, but
note these only concern production and not actual usage, i.e. allocative eﬃciency. This could be
measured based on proﬁt, as that would account for usage and whilst incredibly naive, proﬁtable
ﬁrms must be eﬃcient as the output being produced is not only being used by consumers, but is
also valued by them at a higher price than it cost to produce – hence they make a proﬁt. It would
be very diﬃcult in practical terms however to use proﬁt as a measure of eﬃciency, due to the
presence of other (external) factors. Indeed these lead to the conclusion that it is virtually
impossible to accurately assess the level of allocative eﬃciency in transport markets. As an
example, the provision of a little used rural bus service may be essential to the daily functioning of
a local community. However, if the level of allocative eﬃciency was to be assessed based on proﬁt,
then due to the relatively large number of resources (inputs) being used to move a relatively small
number of people, this would always make a loss and hence be considered to be highly ineﬃcient.
Nevertheless, removal of the service on the grounds of allocative ineﬃciency may reduce overall
welfare because a signiﬁcant percentage of the beneﬁts arising from the service do not accrue to
the direct users of the service and hence would not be included in any such evaluation. In an
entirely free market, however, that is exactly what would happen. This raises many issues, which we
return to later in the text; however, the rest of this chapter concentrates on costs and production.
The economist’s deﬁnition of time
We begin our examination of costs by ﬁrstly considering the issue of time. This needs to be
examined in order to devise formal deﬁnitions for what is a relatively ‘short’ period of time and
what is a relatively ‘long’ period of time. The reason for doing this is that costs may behave quite
diﬀerently depending upon whether a relatively short period of time or a relatively long period of
time is being considered. Time is deﬁned in terms of the extent to which the factors of production
can be varied in order to produce a diﬀerent level of output.
The short run
In economics the short run is formally deﬁned as that period of time during which at least one of
the factors of production is ﬁxed. This could in theory be any of the inputs; however, it normally
relates to capital. The implication therefore is that variations in the level of output can only be
achieved through variation in one or more of the other inputs, normally labour. This may be
achieved through overtime working, employment of agency labour and so on.
101
TRANSPORT COSTS
The long run
In the long run, variations in the level of output can be achieved through variation of all of the
outputs, thus the ﬁrm is not restricted to using only one of them. Capital can therefore be
expanded to achieve such gains.
As regards an actual time period for the short and long runs, i.e. months, years and so on, this
can vary considerably from industry to industry. In relatively labour-intensive industries, i.e. those
that employ a relatively high degree of labour, such as the bus industry, the length of time in the
short run can probably be measured in days or weeks. Hence a few days or a couple of weeks is
roughly the length of time it would take a bus company to increase the capital stock of the ﬁrm, i.e.
purchase a second-hand bus, assuming they are available. Procurements of new buses, however,
will probably take considerably longer and would be measured in months. In more capital-intensive
industries, i.e. those that employ a relatively high degree of capital in the production process, such
as the rail industry, the short run can probably last up to a number of years. Thus for a railway
company, the gap between ordering new rolling stock and it actually being introduced into service
will take a minimum of roughly two years and can take anything up to ﬁve. There is therefore no
ﬁxed period of time as such with regard to the short and long runs, it will vary from industry to
industry dependent upon how long it takes to vary all of the factors of production.
The very long run
The ﬁnal deﬁnition of time in economics is the very long run. In simple terms the very long run is
that period of time where all factors of production are variable, including the level of technology.
Thus production levels that are not possible today may be possible in the very long run due to an
increase in the level of technology.
COSTS AND PRODUCTION IN THE SHORT RUN
Having considered what constitutes ‘eﬃcient’ production and the time dimension in production
economics, the actual production process in the short run is now considered and the underlying
economic concepts underpinning production theory examined. Short run production is considered
ﬁrst as this is the most basic form of the production function outlined earlier, as only one input is
varied to produce variations in the level of output. The total output produced is known as the total
product, with other important concepts in short run production being the average and marginal
outputs (of the variable factor), known as the average and marginal products. The average product
is simply the total product divided by the number of units of the variable factor, and hence in the
case of labour would be more commonly known as labour productivity, i.e. the average amount
produced by each person employed. The marginal product is the change in the total product that
results from adding one more unit of the variable factor into the production process. Thus, for
example, say two people are employed to drive a taxi which results in a daily mileage of 100 miles.
The average product would be 50 miles. If however daily coverage was to be expanded, in the short
run as one of the factors of production is ﬁxed this could only be achieved by either increasing the
number of drivers, or increasing the number of taxis, but not by increasing both. If therefore say
102
TRANSPORT COSTS
one further driver was employed and the daily mileage rose to 120 miles, then average product
would be 40 miles and the marginal product 20 miles. The theoretical relationship between total,
average and marginal products is illustrated in Figure 5.3.
The main issue to consider here is the shape and relative positions of these three curves. Firstly,
the total product curve – this is in the form of an ‘S’ shape. Hence, as more units of labour are
added to the ﬁxed amount of capital, production not only increases but increases at an increasing
rate. This is up to point a in Figure 5.3, which is the point of the highest marginal product as shown
in the lower diagram. Beyond point a, as more of the variable factor is added production still rises,
but at a decreasing rate, reaching the highest point of average product curve at point b before
eventually total product declines after point c. Thus in the short run some form of maximum
output is reached. This tailing oﬀ eﬀect after point a is known as the law of diminishing marginal
returns. Peppers and Bails (1987) clarify that the marginal product of the variable factor of
production will eventually decline if enough of it is combined with the ﬁxed factor. In other
words, a variable input cannot endlessly be added to a ﬁxed factor to continually achieve ever
increasing levels of output. A second point to note is that the marginal product curve cuts the
average product curve at the latter’s highest point. This relationship between the marginal and the
average is an important concept in economics and one that will appear in many other contexts. It
is however a simple mathematical relationship. In this example, while extra units of labour are
increasing total output by more than the average, then this will increase the average. If however the
Figure 5.3 Short run production, total, average and marginal product
103
TRANSPORT COSTS
last person employed should add less to the total output than the average, this will pull the average
down.2
Points b and c in Figure 5.3 can be used to break down the production process into three stages,
known conveniently as Stage 1, Stage 2 and Stage 3 production. The divisions are based upon the
total and marginal products. In Stage 1 production, the marginal product is always increasing,
hence total product is increasing at a rising rate. This would be up to point b in Figure 5.3. In Stage
2 production, diminishing marginal returns set in. Thus although the marginal product is positive it
is falling in value, and thus total product is increasing at a declining rate; in other words from point
b to point c. The earlier taxi example was a case of Stage 2 production, as the last driver added
reduced average product from 50 to 40 miles per employee. Finally in Stage 3 production the
marginal product becomes negative and total product is decreasing, as shown by all points beyond
point c in Figure 5.3.
All of these ideas are best underpinned by a practically orientated example. Table 5.1 outlines
a hypothetical case of a bus operator and the short run production of bus services. It shows what
happens to output levels, in this example measured in vehicle kilometres, as more units of labour
(between 0 and 9) are added to a ﬁxed capital stock, i.e. a ﬁxed ﬂeet size.
The ﬁrst column in Table 5.1 simply gives the number of labour units employed in the
production process. The second column is the resultant output from combining the variable units
of labour with the ﬁxed capital stock. The third column introduces the idea of the average product.
This is simply the second column divided by the ﬁrst, and as highlighted above, would usually be
referred to as labour productivity, i.e. the average units produced per employee. The last column
calculates the marginal product. In simple terms this is the extent to which output rises as a
result of employing one extra unit of labour in the production process. These three measures
therefore give the total, average and marginal products which follow the same general shape of the
theoretical curves shown in Figure 5.3.
As all output variations are produced by the addition of the variable input, all changes in the
level of output are ascribed to that variable input. Starting at zero, no output is produced, as
the only factors of production that are employed are the ﬁxed factors, in this case capital. However,
Table 5.1 Variable labour and the production of bus services
(1) (2) (3) (4)
Labour units Total product
(thousands)
Average product
(thousands)
Marginal product
(thousands)
....................................................................................................................................
0 – – 1
1 1 1.0 6
2 7 3.5 11
3 18 6.0 8
4 26 6.5 6
5 32 6.4 5
6 37 6.2 3
7 40 5.7 2
8 42 5.3 −1
9 41 4.6
104
TRANSPORT COSTS
all factors are required to produce any output – buses cannot operate without drivers. Even after a
single unit of labour is employed, however, still no output is produced. This is because by the time
the single employee has opened up the bus station/depot, cleaned the facilities, carried out all of
the administrative and maintenance tasks, there is no free time left to actually drive a bus!
When a second labour unit is employed, however, then a degree of specialisation can occur –
one can look after the bus station/depot whilst the other drives a bus. As more units are employed,
then more specialisation and/or the better scheduling of inputs can occur and hence the level of
output increases at an increasing rate. This is shown by an increase in both the average product and
the marginal product ﬁgures in Table 5.1. Output levels are therefore increasing at an increasing
rate (as evidenced by the ever increasing ﬁgure of the marginal product), i.e. we are in Stage 1
production. This continues up to where ﬁve labour units are employed, at which point the
marginal product reaches its highest value. With the employment of a sixth person, total output is
still rising but now at a declining rate. The marginal product therefore begins to fall and diminishing
marginal returns set in, which would be at point a on Figure 5.3. Although each person
employed is still increasing the total output, they are not increasing the output by as much as their
predecessor and Stage 2 production is entered, i.e. beyond point b on Figure 5.3. Note that this
has nothing to do with individual ‘productivity’, i.e. the sixth person not being as hard working
and/or as skilled as the ﬁfth person employed, but rather is related to the combination of inputs
employed in the production process. By the time nine labour units are employed, total output
actually falls, thus Stage 3 production is entered as shown at point c on Figure 5.3. This can be
for a number of reasons many of which are related to the productivity of the individual. For
example, employees may now be getting in each other’s way (remember, all other inputs are
ﬁxed). This may not be physically but rather ﬁguratively, such as in the case where one employee
has to wait until a colleague returns the bus to the depot before they can start using it and hence
being productive.
As highlighted above, there is only so much output that can be produced by the ﬁxed factor. In
this example, buses could only be run for around 20 hours a day maximum (allowing for refuelling
and maintenance). Whilst that may be the maximum level, there will also be an optimum level,
i.e. a level of the variable factor (labour) that the ﬁxed factor was ‘designed’ to be operated by. As
this level of output is approached, productivity will increase. Once that point is exceeded however
and the maximum level is approached, overall productivity will be reduced as the ﬁxed factor is
being ‘overworked’ whilst the variable factor in many respects is being underutilised.
Case study 5.1 Costs and production in transport operations
Although costs are not introduced until the next section, it is worthwhile at this stage to look at
the actual inputs used in terms of the factors of production that are employed in the production
of transport services through an examination of their respective costs. Irrespective of the mode,
at the most basic level all transport operators will employ the same inputs in the production of
transport services. At this crude level, these will consist of a combination of a vehicle, a driver/
operative and a power source. The output would normally be measured in terms of the vehicle
kilometres produced through the combination of these inputs, and the inputs would be viewed
as being complementary, i.e. all three are required to produce transport services, rather than
105
TRANSPORT COSTS
substitutable, i.e. using one rather than the other. There will nevertheless be a degree of substitutability
between these basic inputs at the margins. The Docklands Light Railway in London
for example uses driverless vehicles, which technically is a substitution of labour for capital.
Unfortunately with regard to the production inputs involved in transport modes, beyond this
basic level is where the similarity ends. Starting at the most basic split, passenger versus freight
operations, one of the major advantages of passenger services are that passengers by and large
load themselves on to the vehicle, whereas freight has to be physically loaded. Freight
operations, in terms of producing vehicle mileage therefore, will be less ‘productive’ as the level
of inputs required will be higher. Little research exists in this area, however, as few operators
serve both freight and passenger markets. This initially would suggest that there are little if any
advantages in running both passenger and freight operations; however, railways and airlines are
the two notable exceptions. In the case of railways, companies have tended to run both types of
service due to economies of scope, and airlines carry large volumes of freight in the cargo hold
of passenger aircraft due to goods in joint supply (both these concepts are examined further in
Chapter 12).The only major direct comparative study on the topic was undertaken by Professor
Chris Nash in 1985, who found that a higher labour component of around 1.45 was required
for freight as opposed to passenger rail operations (Nash, 1985). Although limited, this is
consistent with the view that freight operations will tend to be less productive in terms of
vehicle kilometres. For airlines, however, the loading of freight onto passenger aircraft will be
part of the general servicing of the aircraft whilst at the airport, and as all of this cargo is
containerised the extra inputs required are likely to be relatively low.
In order to bring out the differences in factor inputs further, however, the costs of ﬁve
different transport operators in ﬁve different markets are examined in Table 5.2. These have
Table 5.2 Mode cost comparisons
Operating Cost Airline Ferry Operator Bus Company Railway
Company
Parcels
British
Airways
Caledonian
Mac.
First
Glasgow
Virgin West
Coast
Parcelforce
2005/6 2005/6 2005/6 Value 2005/6 Value 2005/6
Value % Value % Value % Value % Value %
...........................................................................................................................................................
Labour Costs: 2346 30.0% 45.0 51.7% 45.4 69.2% 102.2 17.7% 5968 71.8%
Vehicle Costs: 1302 16.6% 18.0 20.7% 5.8 8.8% 171.7 29.8% 1392 16.7%
Infrastructure
Costs:
0 0.0% 0.0 0.0% 0.0 0.0% 141.4 24.5% 0 0.0%
Fuel Costs: 1632 20.8% 9.5 10.9% 10.0 15.2% 15.6 2.7% 0 0.0%
Terminal Costs: 1514 19.3% 12.6 14.5% 0.0 0.0% 17.5 3.0% 530 6.4%
Other Overheads: 1034 13.2% 1.9 2.2% 4.4 6.7% 128.5 22.3% 426 5.1%
Totals: 7828 100.0% 87.0 100.0% 65.5 100.0% 576.9 100.0% 8316 100.0%
Fixed Inputs: 2816 36.0% 29.0 37.7% 5.8 8.8% 330.5 57.3% 1922 23.1%
Variable Inputs: 5012 64.0% 48.0 62.3% 59.7 91.2% 246.4 42.7% 6394 76.9%
Sources – Compiled from Company Annual Reports, 2005/6
106
TRANSPORT COSTS
been adapted from the original accounts by grouping costs into six common headings. It is
important to stress even at this stage that the actual costs themselves are not what is important.
This book for example is not entitled ‘Transport Accounting’ and hence this should not be
viewed as an exercise in cost accounting. Rather, what is important is the balance between the
different types of costs and the likely impact such a balance will have on the structure of
production, the ﬁrm and the market. This exercise after all is about attempting to understand
the economics of transport operations, hence the choice of title of the book!
The above examples of transport operators clearly illustrate through cost differences
variations in the structure of production in the modes examined. What is important are not the
actual amounts,as these reﬂect the different scale of operations of each company,but rather the
relative levels spent on each input. This reveals a very high level of variation of input factors
between modes. As staff costs are a direct measure of the relative proportion of labour used in
the production of transport services, the above ﬁgures indicate that parcel and bus operations
are more labour-intensive industries than the railways, ferries and airlines. In turn, railways,
ferries and airlines would appear to be more capital intensive than the others shown. What is
also interesting to note is the relative level of fuel costs – although only shown for four out of the
ﬁve modes (due to accounting differences), these account for somewhere between around 3 and
20 per cent of all operating cost. Interestingly, fuel costs in the bus and airline companies
are around the two highest relative shares of operating costs, suggesting that the level of fuel
costs is independent of the labour and capital intensity of the production process within the
industry. Note however that all transport modes employ a high level of capital equipment in the
production of transport services, as for example few rickshaws exist as a mass mover of people
in the Western world! What is being compared here therefore is the relative levels of labour
used to work the capital equipment, and hence some transport modes are more highly capital
intensive than others.
Listed at the bottom of Table 5.2 is a very rough division where costs have been arbitrarily
split between those that relate to factor inputs that may be considered to be ﬁxed and those
considered to vary directly with production.Of the ﬁve modes shown,passenger railways appear
to have the highest level of ﬁxed inputs, at around 57 per cent, ferries and airlines have around
36 to 37 per cent of ﬁxed costs, whilst parcels (23 per cent) and the bus company (9 per cent)
appear to have by far the lowest level of ﬁxed costs. The real signiﬁcance of this division
between ﬁxed and variable is that variable inputs are only employed when transport services are
actually provided, whilst ﬁxed inputs will incur a cost even where no output is produced and
hence accumulate with the simple passage of time.If a high proportion of operating costs relate
to variable inputs, this means that in a service industry such as transport most costs are only
incurred when in revenue earning service.In many ways this considerably simpliﬁes the planning
of operations and also in theory makes entry into a new industry/market easier, as all else being
equal, start up ﬁnance requirements should be relatively smaller as revenue streams are more
evenly matched with cost outgoings.
This division between ﬁxed and variable factors and the associated costs has major implications
on the structure of the market,as a high level of ﬁxed costs coupled with a capital-intensive
production process would suggest large ﬁrms, which would act against market entry and competition
in the market. On the other hand, it may be expected that in more labour-intensive
transport industries, such as the bus and parcel markets, competition in the market should be
107
TRANSPORT COSTS
both achievable and sustainable. To some extent this would certainly appear to be the case with
parcels, where a number of national and international couriers compete fairly intensively in the
market place. In the bus industry, however, outside of a few isolated cases, there is only limited
competition in the British deregulated market and the country tends to be divided into distinct
bus company ‘territories’, hence little direct competition exists. This suggests that the structure
of costs,whilst an important determinant,is not the only factor in determining market structure
and that other characteristics need to be considered when examining such markets. This will be
taken up in the next two chapters which examine the issues of market structure and competition
in transport markets.
Costs in short run production
Before considering how costs can be classiﬁed, it is important to stress that costs in economics
include proﬁt, or to be more exact, what is known as normal or economic proﬁt. A simpliﬁed
deﬁnition of normal proﬁt would be the opportunity cost of being in business plus some form of
risk premium in recognition of the risks that the investor is taking. Hence for example say the next
best alternative was for the investor to deposit the funds in the bank rather than invest in the
business, and for this they received a 5 per cent annual return. Then in order to be worthwhile
investing in the business, the investor would have to earn more than the 5 per cent per year earned
at the bank, as that gain is virtually guaranteed. Normal proﬁt is therefore the cost to the ﬁrm of
the investor’s outlay, and this is normally paid in the form of a dividend. As such, this has to be
included in costs in much the same way that staﬀ salaries are included in costs. Anything earned
above the level of normal proﬁt would be termed abnormal or supernormal proﬁts, as these are
rewards in excess of the risks of being in business.
Classifying costs
White (2008) highlights a number of possible ways of classifying costs in transport operations.
Firstly, costs can be classiﬁed by input, hence all labour costs are grouped together, all vehicle costs
and so on; this essentially is what was done in Case study 5.1. Secondly, costs can be classiﬁed by
the associated production of various outputs, such as passenger operations and freight operations,
or scheduled and chartered ﬂights in the case of an airline. Finally, costs can be classiﬁed by the
activity performed, thus a railway could group its costs under the headings of sales and marketing
costs, train services costs, infrastructure costs, station costs, administration costs and so on, with
each classiﬁcation of costs relating to a speciﬁc activity of railway operation. All of these enable the
analysis of diﬀerent costs to be made. Within transport economics, however, costs are generally
classiﬁed into ﬁxed and variable. As the names suggest, a ﬁxed cost is one that does not vary with
the level of output, whilst a variable cost does. As seen in Case study 5.1 above, this division and
the balance between ﬁxed and variable costs can have considerable implications on the structure of
transport markets.
Fixed costs are input costs that are sometimes classiﬁed as ‘indivisibles’ or unavoidable costs, as
these costs must be paid even if no output is produced or service provided and cannot be divided
or bought in parts, e.g. you cannot purchase half an aircraft or bus. This includes costs such as
108
TRANSPORT COSTS
leasing charges; rents and rates that relate to oﬃces, depots and stations; management salaries and
costs and administrative expenses such as telephone line rentals. Variable costs on the other hand
relate to the direct expense of providing the output, such as the wages of employees, fuel costs,
power and electricity for heat and light. Many costs however fall somewhere between the two, as
they partially vary in relation to output, hence technically these are semi-variable. Labour costs,
speciﬁcally wages, although normally classiﬁed as a variable cost, would actually be a good example
of such a semi-variable cost, as within a wage there is a basic weekly sum that can be supplemented
with overtime working. Only the overtime element would be directly variable, hence wages
technically are a semi-variable cost.
Depreciation
Depreciation is an important concept that it is worth spending some time on in order to clarify the
issues. One of the main problems associated with this cost is that ‘depreciation’ is the reduction in
economic value to the ﬁrm of using an asset in the production process. There will come a point
in time where it will be more cost eﬀective in the longer term to replace that asset rather than
continue to patch it up through maintenance and keep it running. The more it is used, the quicker
that point will be reached, hence technically depreciation as a concept is a variable cost. To return
to transport accounting, however, this reduction in value is assessed in company accounts as only an
approximation of the real reduction in value. There are two main methods used. Firstly, the
straight line method. Under this method, the scrap value is subtracted from the purchase price and
then divided by the number of expected years of usage. That then gives the value of ‘depreciation’
to be written oﬀ annually. The second approach is the reducing balance method, where a percentage
of the value is written oﬀ each year until the scrap value is reached. Both approaches are best
illustrated through the use of an example, and again a bus company is used.
Say for example a new bus costs £110,000, has a scrap value of £10,000 and has an expected
useful life of ten years. The value to be written oﬀ therefore is £100,000 and over the useful life
of ten years this would be done by subtracting a straight value of £10,000 a year. In ten years
therefore it would be shown in the books as having a value of £10,000, i.e. its scrap value. This
would be the straight line method. Alternatively, a percentage of the book value could be written
oﬀ each year, in this case 21.3 per cent, and this will achieve the same eﬀect of reducing the value
to (almost) £10,000 in ten years’ time. This would be the reducing balance method. The annual
depreciation charge is then written oﬀ against proﬁts, as this is the cost to the ﬁrm of using the
capital equipment. Both approaches are fully illustrated in Table 5.3.
Note that under the reducing balance method higher values are written oﬀ in the earlier years,
and this is said to better reﬂect the decreasing value of such assets in that earlier period. Irrespective
of what method is used, however, neither takes into account usage. For example, as illustrated
under the straight line method £10,000 is written oﬀ the value of a bus each year irrespective of
the extent to which that bus is used. In this sense, therefore, it is a ﬁxed cost, as it is one that does
not vary with the level of output produced. In the longer term, however, if that bus was little used
it would retain its economic value (obsolescence accepted) beyond the ten years, and hence would
still be useable and thus of value to the ﬁrm after it had been written oﬀ in the books. There are
many such examples of transport assets that have long been written oﬀ in the company accounts
but are still in use today, such as the Severn Rail Tunnel and the Forth Bridge.
109
TRANSPORT COSTS
Short run average and marginal costs
As total, average and marginal products are plotted against labour, costs are normally plotted
against the output produced by that labour. This then illustrates how costs vary over diﬀerent levels
of the output. These are known as the average and marginal costs curves, in this case those relating
to the short run. The ‘S’-shaped production function outlined earlier will in turn produce an
‘S’-shaped total cost curve, therefore consistent with this the average cost curve is ‘U’ shaped. In
simple terms, as total productivity increases (Stage 1 production), then average costs fall, whilst
when total productivity decrease (Stages 2 and 3 production), average costs increase. This is shown
in Figure 5.4.
This is the stylised textbook version of the short run marginal and average cost curves. As more
of the variable input is added, average costs at ﬁrst fall, are then minimised at the optimal output
level of production at point b, and then increase beyond that optimum point. Having outlined the
basic shape of these curves, what is important is why the short run average cost curve should
be U shaped. As this relates to the short run, one of the inputs in the production process is ﬁxed
and hence will be subject to the law of diminishing marginal returns, which in turn will be related
to utilisation of the ﬁxed input. Say for example in the production of bus services, the number
of buses is ﬁxed. As highlighted previously, this means that there is an optimal production level,
i.e. an ideal output size. Any production level that is under that point will lead to the underutilisation
of the ﬁxed resource and consequently higher average costs. As that optimum point is
approached, i.e. point b on Figure 5.4, these ﬁxed resources are increasingly being utilised, hence
average costs fall. Any point beyond the optimum however will lead to the over-utilisation of
resources. Furthermore, as the level of output can only be altered by varying the level of the
variable factor, in this case labour, then in order to increase labour this may involve paying overtime
rates, hiring agency labour at a higher cost and so on, thus incurring a higher overall average
unit cost.
In order to examine average and marginal costs further, Table 5.4 expands the previous
Table 5.3 Illustration of depreciation by straight line and reducing balance methods
Year Straight line method Reducing balance method
Value at the
beginning of
year
Annual
depreciation
charge
Value at the
end of year
Value at the
beginning
of year
Annual
depreciation
charge
Value at the
end of year
....................................................................................................................................
1 110,000 10,000 100,000 110,000 23,430 86,570
2 100,000 10,000 90,000 86,570 18,439 68,131
3 90,000 10,000 80,000 68,131 14,512 53,619
4 80,000 10,000 70,000 53,619 11,421 42,198
5 70,000 10,000 60,000 42,198 8,988 33,210
6 60,000 10,000 50,000 33,210 7,074 26,136
7 50,000 10,000 40,000 26,136 5,567 20,569
8 40,000 10,000 30,000 20,569 4,381 16,188
9 30,000 10,000 20,000 16,188 3,448 12,740
10 20,000 10,000 10,000 12,740 2,714 10,026
110
TRANSPORT COSTS
example regarding the production of bus services in the short run by including cost data. In this
case it has been assumed that ﬁxed costs are £80,000 and the variable factor, labour, costs £30,000
per unit.
The ﬁrst four columns are taken from Table 5.1, and to this have been added ﬁve new columns
to include the relevant cost data. Total Fixed Costs, as outlined above, do not vary with the level
of output, hence remain ﬁxed at £80,000 irrespective of the level of output produced. These are
the costs associated with the ﬁxed factor, in this case capital. Total Variable Costs are simply the
number of labour units employed times the cost of each unit, and total costs the addition of ﬁxed
Figure 5.4 Short run average and marginal cost curves
Table 5.4 Variable and ﬁxed costs of short-run production of bus services
Production Costs
Labour
units
Output
produced
(000s)
Average
product
(000s)
Marginal
product
(000s)
Total
ﬁxed
costs
Total
variable
costs
Total
costs
Average
total
costs
Marginal
costs
....................................................................................................................................
0 – – 1 80000 – 80000 – 30.00
1 1 1.0 6 80000 30000 110000 110.00 5.00
2 7 3.5 11 80000 60000 140000 20.00 2.73
3 18 6.0 8 80000 90000 170000 9.44 3.75
4 26 6.5 6 80000 120000 200000 7.69 5.00
5 32 6.4 5 80000 150000 230000 7.19 6.00
6 37 6.2 3 80000 180000 260000 7.03 10.00
7 40 5.7 2 80000 210000 290000 7.25 15.00
8 42 5.3 −1 80000 240000 320000 7.62 –
9 41 4.6 – 80000 270000 350000 8.54
111
TRANSPORT COSTS
and variable. Average total cost is therefore the total cost divided by total output, and marginal
cost the cost of the last unit produced. When for example labour is increased from one to two
units, output increases by 6,000 units and costs rise by £30,000. The marginal cost of the last unit
therefore is the diﬀerence in output divided by the diﬀerence in costs and hence is £5.
Notice also that when only production is examined the highest level of labour productivity
occurs at 4 units; however, the lowest average cost occurs when 6 labour units are employed. The
reason for this diﬀerence is because the average cost also includes the capital cost, whilst no
account is taken of capital inputs under the previous calculation as all variations in output were
apportioned to the labour factor. This is one reason why productivity measures based on a single
input, such as labour in this case, can be misleading.
The importance of the idea of the average cost is illustrated in the next case that looks at the
operational characteristics of low-cost airlines.
Case study 5.2 The importance of average cost in the business
model of low-cost airlines
It is difﬁcult, if not impossible, to have a chapter on transport costs without having a speciﬁc
look at the model of the low-cost airline.As the name of these carriers strongly suggests,this is a
business model of airline operation that is heavily based on achieving low average costs in the
operation of services. It should be stressed however that whilst ‘low cost’ is a philosophy that is
central to these companies’ operations, this is only one part of a complete business model that
is geared towards achieving ‘low cost’ through not only cutting costs directly but also by a
number of other measures as well.
The ﬁrst acknowledged low-cost carrier was the American airline, Southwest, which began
operating along low-cost principles in 1978, not long after deregulation of the US Domestic
airline market. This business model however did not spread to Europe owing largely to the
heavily regulated air market that had been exempt from both the initial Treaty of Rome in 1957
and the Single European Market in 1986. As a result, the European Air Market operated
restricted capacity on routes and regulated prices and was largely dominated by state-owned
National Airlines.In the early to mid 1990s,however,the EU passed a number of packages of air
regulation reforms that opened up the market to potential new entrants. By doing so, this also
presented an opportunity for the entry of low-cost operations, and this was initially taken up
with considerable success by the Irish operator Ryanair. This model of operation has since
been followed by a large number of other airlines, including easyJet, Flyglobespan.com and
Flybe. These airlines operate on the basis of offering low priced fares based on a low average
cost. Note that this is slightly different from an out and out low cost, as in this business model
it is the achievement of low average cost that matters, not just ‘low cost’. This obviously has
implications on the operations of the airline, and in particular also puts great emphasis on
aircraft utilisation. The general ‘model’ of the low-cost carrier can be summarised as consisting
of the following main elements:
᭿ Low staff costs
᭿ Low aircraft turnaround times
112
TRANSPORT COSTS
᭿ Route network based around secondary or regional airports rather than major hubs
since they are usually less congested
᭿ On line ticket sales only with ticketless travel
᭿ Cabin crew perform other duties during turnarounds, such as cleaning the inside of the
aircraft
᭿ Simpliﬁed point-to-point operations rather than complex hub and spoke
᭿ All ‘extras’ beyond the basic seat incur an additional charge, e.g. in ﬂight catering or
priority in boarding
᭿ No spare aircraft capacity held in reserve to cover in the case of unforeseen breakdowns
or delays
᭿ Fleet based on a single aircraft type to reduce maintenance costs.
In order to tease out these differences, the costs of three airlines are examined below in Table
5.5.These have been grouped under ﬁve common headings in order to attempt to bring out these
differences. The ﬁrst airline, British Airways, can be viewed as a ‘traditional airline’, whilst the
other two,easyJet and Ryanair,operate largely along low-cost principles,the latter more so than
the former.
All costs are taken from the respective 2006 Annual Reports; however, not all costs are
directly comparable. Despite the existence of Standard Statements of Accounting Practice
designed to ensure consistency across company accounting practices, differences will always
remain. Comparison differences therefore cannot be taken as ‘exact’. As an example, easyJet
only include Crew Costs (cabin and ﬂight deck) as speciﬁc staff costs in their annual reports,
whilst British Airways (BA) include all staff costs, e.g. cabin and ﬂight deck crew, administration
staff, sales and marketing staff and so on. Nevertheless, the above costs do give some basis
for an overall appraisal of different costs.
BA provides an interesting starting point to examine ﬁrstly the overall cost structure of
airlines and secondly where any savings in operating costs could be made in what would be
perceived as a traditional airline. Starting with labour costs, these account for around 30 per
cent of all operating costs. Whilst this is a signiﬁcant proportion, in transport operating terms
this suggests a relatively capital intensive industry – as was seen earlier, for bus companies staff
costs account for around 65 per cent of operating costs. Examining the other costs for BA,
direct operational costs such as fuel and oil, engineering and maintenance costs, landing fees
Table 5.5 Operating costs, British Airways, easyJet and Ryanair, 2006
Airline: British Airways easyJet Ryanair
Actual % Actual % Actual %
...........................................................................................................................................................
Staff costs 2346.0 30% 75.2 11% 171.4 13%
Selling costs 449.0 6% 26.0 4% 13.9 1%
Aircraft costs 2446.0 31% 366.8 54% 590.1 45%
Fuel costs 1632.0 21% 165.9 25% 462.5 35%
Other costs 955.0 12% 42.2 6% 85.6 6%
Source: Adapted from the respective company accounts
113
TRANSPORT COSTS
and so on account for just under 50 per cent of operating costs. It is in these two areas, staff
costs and operational costs, where the biggest differences between BA and the two LCAs
emerge. BA has a far higher proportion of operating costs ascribed to staff and a far lower
proportion accounted for by direct operational costs than either easyJet or Ryanair. The other
area of difference is in selling costs, although these on the whole are a small proportion of total
operating costs.
This is where an understanding of the differences between ﬁxed and variable costs becomes
useful,as also does knowledge of how employment contracts are drawn up.Labour,for example,
would virtually always be classiﬁed as a variable cost.Where employees are salaried,however,as
in this case, then in the short run this is a ﬁxed cost. A similar argument applies to depreciation
and aircraft lease costs, as again in the short run capital equipment will be ﬁxed and hence the
depreciation or lease charge relatively ﬁxed. Following this logic then, in fairly rough terms all
operational costs are variable whilst all other costs, including labour and aircraft costs, are
ﬁxed. In the case of BA, therefore, around half of the costs are variable whilst for the other two
airlines this is far higher at nearer 70 per cent.
Whilst the airline can do little about variable costs such as fuel and oil as these vary directly
with output, these costs are zero while the aircraft is stationary. Fixed costs however are not.
For example, cabin crew are still employed (being paid) after the aircraft has reached the
airport gate and all passengers have disembarked, hence such costs are incurred even although
the aircraft is technically not in revenue earning service. This is achieved only when the aircraft
is in ﬂight, therefore what becomes crucial in any such business model of low-cost operation is
aircraft utilisation – the higher the utilisation, the more ﬁxed costs are spread over output, the
average cost reduces and the proportion of ﬁxed to variable cost also reduces. Thus high
turnaround times become the vital element in this type of operation.easyJet for example operate
Figure 5.5 Percentage breakdown, operating costs, British Airways, easyJet and Ryanair,
2006
Source: Compiled from the respective company accounts
114
TRANSPORT COSTS
something like an absolute maximum of a 45 minute aircraft turnaround time and in most cases
considerably less than that. Also what is important is that staff while on the ground, such as
cabin crew,are actively employed to service the inside of the aircraft while it is stationary,hence
they are being ‘productive’ during what would normally be ‘dead’ time. Note also that ‘dead
time’ as such is far shorter due to fast turnaround times.
Another important aspect in reducing average cost is the lack of any slack in the whole
system of operation. Spare aircraft are simply not an option, as this would constitute a ﬁxed
cost with zero output, hence any faults that develop in the ﬂeet during the daily operation can
have considerable and long-lasting (i.e. all day) knock-on effects on the whole system. Those
with some experience of LCAs for example will be familiar with the announcement of ‘ﬂight
delayed due to the late arrival of the incoming aircraft’ which although a completely meaningless
statement does highlight the lack of any slack in the whole system.Why LCAs can get away
with such practices of course is due to the charging of low air fares.
The key to success in the operation of LCAs, therefore, is not ‘low cost’ per se but rather the
attainment of a low average cost (per passenger carried). Few pilots for example work for low
wages; however, the key to effectively reducing a pilot’s wage is through increasing the number
of hours worked (within regulatory limits of course).
In many senses, LCAs have changed the whole economics of airline operation where traditional
thinking was in terms of an industry with a high proportion of capital costs and a
relatively low level of variable costs. Furthermore, most variable costs were associated with
take-off and landing, with respect to landing and terminal charges and most importantly fuel
costs – a high proportion of the fuel consumed on a ﬂight is during take-off and landing phases,
sometimes as high as 50 per cent. All of these factors would suggest far higher unit prices on
short-haul ﬂights as the average cost of such a ﬂight would be far higher.
All of these factors surrounding air transport economics however were known for some time,
particularly the aspect of aircraft utilisation being the key to success in the airline business.
Why they had not been fully exploited before is due to other factors, most notably market
conditions and regulatory regimes. Facing such conditions, it made economic sense for the
operator to restrict supply as this increased proﬁts. In the LCA model, however, proﬁts are
maximised through low proﬁt margins and high passenger volumes.
COST AND PRODUCTION IN THE LONG RUN
The ﬁrst point to note about costs and production in the long run is that because all input factors
are variable there is no division between ﬁxed and variable costs. To brieﬂy recap, whilst a ﬁrm
may be planning a new production facility or entering a new market, it is operating in the long run.
Once however it builds the new factory or sets up the new depot to service the new market, it is
then operating in the short run (because at least one of the factors of production is ﬁxed).
How costs behave in the long run is closely related to the behaviour of production and thus
what happens when more inputs are added to the production process. There are some important
diﬀerences between production and costs; however, these will be further developed later. To begin,
Figure 5.6 illustrates a long run production function.
115
TRANSPORT COSTS
As with the short run, the long run production function is S shaped in nature. At ﬁrst there are
large gains when ﬁrm size increases – the relative percentage gain in output is greater than the
relative percentage increase in inputs. Note also that this eﬀect increases as ﬁrm size increases.
These gains in total productivity, or increasing returns to scale, continue up to point b on
Figure 5.6. Once ﬁrm size (as measured by the level of inputs) rises past point b, however, the
proportionate gains from adding more inputs are not as large as before, hence the ﬁrm experiences
decreasing returns to scale. This decline continues until increasing ﬁrm size has very little eﬀect on
the level of output. The main reasons for this general pattern of increasing and decreasing returns
to scale that deﬁne the shape of the long run production function are outlined below.
Sources of increasing returns to scale
Specialisation of labour – larger ﬁrms allow more specialisation of the workforce to occur. If this is
considered at the most basic level of a one-person sole-trader business, then the owner has to
undertake all of the tasks involved in the running of the business. They therefore become a ‘jack of
all trades and master of none’. As ﬁrm size increases, more labour can be employed in specialised
tasks and thus become more proﬁcient at those tasks. Consequently productivity would be
expected to increase. Note also that as this is the long run, there is potentially no ceiling on this
eﬀect and is probably best exempliﬁed in Fordism large-scale production, where individuals can
become ‘experts’ at very speciﬁc tasks.
Scheduling of inputs – this is similar to the specialisation of labour source but refers to the
scheduling of all of the inputs, a factor that is particularly prevalent in the transport industries. As
ﬁrm size increases, there exists greater ﬂexibility in how the inputs can be combined and hence
better utilisation of all of the inputs may be expected, i.e. higher total productivity. For example, in
larger road haulage companies, there may be more ﬂexibility in the scheduling of drivers to ensure
that the vehicle stock is operated over the longest possible number of hours. This would ensure
higher total productivity.
Figure 5.6 The long run production function
116
TRANSPORT COSTS
Capital inputs – this concerns a number of issues that are broadly grouped together here under
the title ‘capital inputs’. Firstly, some capital inputs can be very expensive to purchase. Only larger
ﬁrms can aﬀord to spend on these inputs, but they only do so on the basis that this will lead to
improved eﬃciency (i.e. higher productivity) in the longer term. For example, increasing a railway
line from single to double track increases capacity by a factor of four, hence potentially signiﬁcantly
increasing the productivity of rail services.
The second issue also relates to specialisation. Using the sole-trader example, then it may well
be that the company vehicles consist of a single solitary van. This van will have to fulﬁl all of the
transport requirements of the company, some of which it may be better suited to than others. As
ﬁrm size increases, however, then the company ﬂeet can be increased not only in size but also in
scope, hence more suitable vehicles can be used for more suitable tasks. This should lead to higher
productivity.
Thirdly, it may make more sense for larger companies to carry a larger number of spares and
maintenance facilities, hence downtime of capital equipment as such should be reduced and
consequently higher productivity achieved. It should be noted that even with ‘just-in-time’ production
methods and the contracting out of maintenance and servicing facilities (as has happened
with many bus companies), certain advantages with regard to this issue may still be expected.
It may be anticipated for example that larger companies will have far more inﬂuence with the
suppliers of such services that results in quicker response rates to service requests. Capital downtime,
therefore, may be expected to be lower.
Indivisibilities – the standard example of an indivisibility is a telephone line. When setting up in
business, a company will need to install and rent a phone line. With small expansions in size, there
will probably be no need to install a second phone line, hence this ‘input’ is spread over a larger
output. There will obviously come a point where a second phone line will have to be installed, but
this would only be done if it was advantageous to do so, e.g. if it improved ‘productivity’. Another
example of an indivisibility of course is our sole trader’s van. As ﬁrm size increases, the van’s
overall utilisation would improve.
All of these sources of increasing returns to scale should be reasonably clear, but what may
cause this increase in productivity to tail-oﬀ, i.e. sources of decreasing returns to scale? This is
particularly prevalent in the long run as there is no upper constraint in the form of a ﬁxed input.
Sources of decreasing returns to scale
Loss of control – as ﬁrm size increases, there is a loss of control over the whole organisation. This
loss of control decreases overall productivity. Under this heading it is worth highlighting
the concept of ‘X-ineﬃciency’, originally devised by Leibenstein (1966). In simple terms,
X-ineﬃciency relates to general management slack, and large (publicly owned) organisations are
said to be particularly vulnerable to this concept. In less academic terms, in larger organisations
there may be a loss of the ‘sense of the individual’ and more opportunities for general slack
working practices, hence leading to lower productivity levels.
Geographical location – particularly prevalent in the bus industry, but true of other transport
industries, when a ﬁrm initially sets up in business it will probably be on or near to the optimal
location. Increasing size in the longer term means building other production facilities, such as depots,
and these will not necessarily be at the best location. This can result in fairly long distances between
117
TRANSPORT COSTS
the depot and the market served, hence a signiﬁcant proportion of time is spent in driving vehicles
between the two and not actually providing transport services. As a result, productivity decreases.
Administration procedures – large ﬁrms often require many more layers of middle and upper
management, plus administration procedures, in order to control costs and processes within the
organisation. This is commonly known as ‘bureaucracy’; however, in this situation this should not
be confused with ‘red tape’, as that is considered in the next section. More speciﬁcally, this refers
to the time dimension that such ‘form ﬁlling’ requires and hence the opportunity cost of this form
ﬁlling is the distraction of employees from the production process. When measured in terms of
overall output, therefore, it requires on average a higher number of employees to produce a higher
level of output.
Average and marginal costs in the long run
Having considered production in the long run, it is possible to see how costs would be expected to
behave in the long run. These are graphed in Figure 5.7.
As can be seen from Figure 5.7, average costs at ﬁrst fall as ﬁrm size (as measured by output)
increases. This continues up to the point where average costs are minimised at the optimum level
of production, known as the minimum eﬃciency scale (MES). After this point the trend is reversed
and average costs rise as ﬁrm size increases. Along the part of the curve where the average cost is
falling the ﬁrm would be said to be experiencing economies of scale. Very often these are
incorrectly termed increasing returns to scale; however, returns to scale relates to production
output while economies of scale relates to production costs. Along the part of the curve where
average costs are rising the ﬁrm would be said to be experiencing diseconomies of scale, or again
often incorrectly referred to as decreasing returns to scale.
This does once again pose the question as to why long run average costs should ﬁrst fall and
then rise as output (and hence ﬁrm size) increases. This is explained below.
Figure 5.7 Long run average and marginal costs
118
TRANSPORT COSTS
Sources of economies of scale
Increasing returns to scale – as seen above, at ﬁrst as ﬁrm size increases it encounters increasing
returns to scale i.e. ever higher levels of productivity. This increased productivity therefore means
that relatively lower levels of the inputs need to be employed to produce higher levels of output,
hence the average costs per unit of output falls.
Bulk buying – larger ﬁrms can normally obtain some form of discount for buying capital
equipment and supplies in larger numbers, and hence average costs would be expected to be lower
for larger ﬁrms. Bus companies are a good example of this, where large group holding companies
can negotiate discounts on fuel and tyres due to the sheer volume that the company will use in the
course of its normal operations. Due to this volume, suppliers can aﬀord to concede larger
discounts and still return acceptable proﬁts.
High cost inputs – these have already been examined under returns to scale, hence most are
associated with improving productivity rather than directly reducing average costs. It is worth
highlighting however that advertising is also sometimes cited as a high cost input, but it is debatable
if this really leads to economies of scale. In the Cola market, for example, it almost certainly does.
By advertising extensively, ﬁrms seek to increase the size of the market and hence allow them to
increase production in order to take advantage of other economies of scale.
Financial economies – larger ﬁrms are normally better placed to secure additional ﬁnance as they
can oﬀer greater security. Interest rates therefore may be lower as there is a lower risk involved to
the ﬁnance company, hence average costs are lower.
Sources of diseconomies of scale
Decreasing returns to scale – as highlighted above, there are a number of sources of technical
ineﬃciencies that lead to reduced productivity for larger ﬁrms, thus the average output per unit of
input falls. In order to produce higher levels of output, therefore, relatively higher numbers of
inputs need to be employed, and this adds to costs causing diseconomies of scale.
Red tape – as noted above, larger ﬁrms often require many more layers of middle and upper
management, plus administration procedures in order to control costs and processes within the
organisation. This is ‘red tape’ and the actual cost of this added administration burden will add to
costs, hence increasing the average cost per unit. Note however if we are being absolutely correct,
the extra staﬀ results in decreasing returns to scale (as proportionally less staﬀ are directly involved
in production), whilst the added costs associated with red tape such as stationery costs, the need
for greater oﬃce space and more IT facilities and so on result in diseconomies of scale.
The coverage of costs in the long run is completed by examining economies of scale in railway
operations and the impact that this can have upon how services are actually provided to the market.
Case study 5.3 Economies of scale and reform in railway
operations
The general view of economies of scale within the rail industry is that, due to a high capital
requirement in the provision of rail services, economies of scale are signiﬁcant and hence
company size needs to be large in order to capture all of these effects.In the past this was one of
119
TRANSPORT COSTS
the main reasons which led to the nationalisation of railway industries across Europe (the ﬁrst
being Switzerland in 1901 and the last Britain in 1948), where most of the main-line railways
were taken under the control of a single company so that economies of scale could be achieved.
There were also a number of other important reasons for nationalisation; however, here we only
concentrate on the economies of scale argument. This particular view of railway economics has
become known as the traditional view (Preston, 1994), in which infrastructure and services are
part of an integrated system and economies of scale in both are signiﬁcant. Organisationally,
therefore, services and infrastructure should be part of the same (large) company.
In recent years, however, virtually every European country has re-organised their railway
systems, with many separating both organisationally and ﬁnancially infrastructure from services,
i.e. one company owns and operates the infrastructure whilst a different company owns
and operates the rolling stock. This is known as a vertically separated railway. In Europe,
Sweden was the ﬁrst country to organise its rail system along these lines. All the infrastructure
functions were separated out into the Swedish National Rail Administration (Banverket) whilst
services remained within the Swedish State Operator (Statens Järnvägar). In contradiction to
the traditional view outlined above, however, such a division is consistent with what has become
known as the revisionist view of railway economics. Under this belief, the central premise is that
economies of scale are only associated with the infrastructure and not with services. Therefore,
scale effects will still be taken advantage of as long as infrastructure is retained as a single
entity.As regards the size of operating companies,this is unimportant as there are no economies
of scale associated with this activity.The advantage of such a system over the traditional railway
is that different operators can operate on the network and hence some form of competition
introduced into the provision of rail services.
Note that both the traditional and revisionist views are merely matters of opinion, and for a
better understanding we need to examine the empirical research carried out on the topic. This
would suggest that the theoretical U-shaped cost curve (and hence S-shaped production
function) applies to railway operations. For example, Preston (1994) in a study of 15 Western
European (integrated) railways found diseconomies of scale for larger train systems such as
(West) Germany and Britain, and increasing returns for smaller rail systems such as Ireland,
Switzerland and Belgium. In an updated study, Shires and Preston (1999) found that the
MES (minimum efﬁciency scale) for integrated railways was around about the size of the
Danish and Belgian rail networks. The implication, if only economies of scale are considered, is
that countries with large networks such as Britain should organise their rail system into three
to four integrated railways (perhaps on a regional basis) rather than as one single national
operator. Such a structure would place each system close to the MES point and hence eradicate
diseconomies of scale.
Within smaller railway systems, scale effects have been found to be even more substantial.
Filippini and Maggio’s (1992) study of the Swiss ‘private’ rail network revealed scale effects to
be considerable, leading the authors to conclude that there were potentially substantial beneﬁts
to be gained from end-to-end and parallel mergers within the Swiss industry.Cowie (1999) also
found scale effects to be signiﬁcant in the Swiss industry. Switzerland is made up of a single
national operator, CFF, on the main lines, and around 60 to 70 mainly publicly owned local
railways on the local lines, varying in size from as small as 4 kilometres up to around 400
kilometres in length. Signiﬁcant scale effects in such small systems would again be consistent
120
TRANSPORT COSTS
with a U-shaped cost curve.This is because the gains that could be achieved in terms of lowering
average cost for smaller systems by increasing output would be higher than for medium-sized
operators (have a look at Figure 5.7 to conﬁrm this).
To date, however, very little research has been carried out on the impact of separating
operations from infrastructure on economies of scale. In one of the few such studies, Cowie
(2002) examined the British train-operating companies and found that scale effects were
signiﬁcant, hence suggesting that the pure revisionist view, that there are no economies in
operation, was not true. Size, therefore, with regard to how train companies are organised, is
important.Scale effects however were found to be smaller than for integrated railways,with the
MES point found to be around two thirds of the output level of Preston’s earlier study based
upon integrated railways. It nevertheless suggests that in Britain there should be around four to
ﬁve train-operating companies rather than the current number of seventeen. The research also
suggested by implication that there existed signiﬁcant economies of scale in the provision of
infrastructure.
Economies of scale have a major impact on the consideration of the best size of railway to
produce rail services, as clearly if scale effects are considerable then costs can be signiﬁcantly
reduced by having a very large operator. Costs however are only one half of the proﬁt equation,
and hence where railways are heavy loss makers and thus heavily subsidised (as is the case in
most of Europe), the attainment of economies of scale in the production of services whilst
important is only one aspect amongst many other considerations when policy makers consider
how ‘best’ to organise a rail system. For example there may be beneﬁts in having a higher
number of rail operators than the ‘optimal’ level due to an increase in competitive pressures,
particularly for the market.
SHORT AND LONG RUN AVERAGE COSTS
We brieﬂy end the chapter with a look at the relationship between the short and the long run
average and marginal cost curves. This is shown in Figure 5.8, and is useful to further underline the
relationship between all of these concepts.
In Figure 5.8, the long run average cost curve is a summation of a series of short run average
cost curves. Beginning at an output level of Q1, the ﬁrm is operating at point a on both the short
run and long run average cost curves. Note that the short run average cost curve is tangential to
the long run average cost curve at that level of output. Note also that at each point on the long run
average cost curve there is a similar short run average cost curve which is tangential to it. If the
ﬁrm was to increase production from Q1 to Q2 in the short run, however, i.e. where at least one of
the inputs is ﬁxed, average costs would increase to point c on the short run average cost curve. This
would be considerably higher than if production was increased to Q2 in the long run, in which case
the average cost would be found at point b on the long run average cost curve. This is because in
the short run the ﬁrm would encounter the law of diminishing marginal returns, and hence
because one of the factors is ﬁxed, say vehicles, the variable factor, labour, cannot be fully utilised.
If however production is increased in the long run, then diminishing marginal returns will not be
encountered and in this example, the ﬁrm will experience economies of scale as it approaches the
121
TRANSPORT COSTS
MES point. An alternative way of viewing this is that short run costs are higher because when
operating in the long run the ﬁrm eﬀectively has a blank sheet of paper (in theory at least), and
hence can plan for the lowest cost level of production. Once one of the factors is ﬁxed, however,
then it is more conﬁned in what it can do and has to work around this constraint, which will
incur a higher average (short run) cost. This relationship between costs in the short and long
runs is important, particularly when examining transport markets. Very short increases in demand
have to be met by a short run increase in supply, which in turn would incur higher costs. Given
the peak nature of most public transport markets, the result is that ﬁrms can very seldom
‘optimise’ on the long run average cost curve, as supply needs to be ﬂexible to meet these peaks in
demand.
CHAPTER SUMMARY AND REFLECTION
In this chapter the focus has been on transport costs. This began with a deﬁnition of production, in
which it was revealed that the deﬁnition of the short run is where at least one factor of production
is ﬁxed. We then examined the behaviour of costs both in the short and long runs, and found that
both the short and long run cost curves were U shaped. In the case of the former this was due to
the law of diminishing returns and the latter due to economies/diseconomies of scale. We ﬁnished
the chapter by detailing the relationship between costs in the short and long run. A case study of
the rail industry revealed that costs in practice would appear to follow the theoretical concepts,
with studies indicating that average cost curves are U shaped in both vertically and integrated
railway systems. Whilst not explicitly stated, it also highlighted the eﬀect of costs on the structure
of the industry, where for example if other things remained equal then it would be expected that
more ﬁrms would operate in say the road haulage or bus markets than the passenger or freight rail
markets primarily due to the diﬀerences in cost structures and the eﬀects of economies of scale.
What has also been highlighted is that costs are only one part of calculating the proﬁt level and
consequently only one factor in the planning of transport operations. Revenues, both in terms of
Figure 5.8 Long and short run average and marginal cost curves
122
TRANSPORT COSTS
that collected directly from the passenger and also sums paid by transport authorities in the form
of transport subsidies, are also required since proﬁt is revenue minus cost. The following chapter
builds directly on the ideas outlined in this chapter and introduces revenue into the analysis in
order to provide a complete insight into both sides of the transport market. Subsequent chapters
will then consider the organisation of transport services and transport subsidy, both of which are
heavily inﬂuenced by the structure of the costs of provision of transport services. From a public
perspective, however, the ﬁnancial costs, both in terms of capital and operating, need to be oﬀset
against not only the ﬁnancial gains but also wider public beneﬁts, and this topic is taken up further
in Chapter 14 on transport appraisal
CHAPTER EXERCISES
Exercise 5.1 Technical,cost and allocative efﬁciency in bus operations
The following table gives some basic information relating to two small-scale bus operators:
a) Consider the following questions:
Which of the two companies is more technically efﬁcient in the production ofi
bus services (note: you will need to separately calculate labour productivity and bus
productivity and compare the two ﬁgures)?
Which of the two companies is more cost efﬁcient in the production of bus services?ii
Which of the two companies is more allocatively efﬁcient?iii
b) In your answer to part a. (i) you should have found that Company A had an advantage in both
productivity ratios,hence by implication is more technically efﬁcient.However,say this had not
been the case.For example,if Company B had only 16 staff rather than 21 then it would have a
superior labour productivity but an inferior bus productivity, hence there would be no clear
answer (you should check this), as that would be dependent upon the balance of the two inputs
used in the production process. In other words, such evaluations need some way of combining
these ratios to come up with a single answer. One fairly simply way of doing this would be to
weight each productivity ratio by that factor inputs share of costs. This would be a basic form
of what is known as a Tornqvist productivity ratio.You should now calculate this index for both
Company A and Company B using B’s revised labour ﬁgure of 16 staff to determine which
company is more technically efﬁcient.
Company A Company B
Number of buses: 3 5
Number of employees: 13 21
Average wage: 19000 16000
Vehicle kilometres run: 210000 300000
Bus cost per kilometre (including fuel costs): 0.36 0.30
Annual number of passengers carried: 370000 460000
123
TRANSPORT COSTS
Exercise 5.2 Total,average and marginal products and costs
This exercise concerns the provision of rail services, and the task is comparatively straightforward
if slightly involved. Quite simply, you have to ﬁll in all the blanks, for which you will need the following
information:
Fixed Costs: 100000
Price of a variable factor: 50000
You should round all ﬁgures to two decimal places.
Once you have completed this table, you should use your calculations to answer the following
questions:
a) At what level of output should the ﬁrm operate at?
b) What is the most ‘efﬁcient’ level of output in terms of:
Technical efﬁciency?i
Cost efﬁciency?ii
In terms of measuring the ﬁrm’s ‘efﬁciency’,which of these two measures should be usediii
and why?
c) What units is the level of output measured in?
Table 5.2a
Labour Output(000s) Costs
Units TP AP MP TFC TVC TC ATC MC
....................................................................................................................................
0 0
1 50
2 110
3 180
4 260
5 350
6 420
7 480
8 530
9 570
10 590
TP = total product AP = average product MP = marginal product
TFC = total ﬁxed costs TVC = total variable costs TC = total costs
ATC = average total costs MC = marginal cost
124
TRANSPORT COSTS
Exercise 5.3 Economies of scale in railway operation
Re-examine Case study 5.3 and answer the following questions:
1 Brieﬂy outline your understanding of the traditional and revisionist views of railway economics.
2 List what you believe to be the main sources of economies of scale in the rail industry. Once you
have produced this list, indicate which arise as a result of returns to scale and which are cost
savings.
3 What on the other hand do you believe are the main sources of diseconomies of scale in larger
integrated railways?
4 If you were a rail industry regulator in Britain today, what other factors apart from economies of
scale would you take into account when deciding on the number of operators to have in the market?
125
TRANSPORT COSTS