REVIEW OF ECONOMIC PERSPECTIVES – NÁRODOHOSPODÁŘSKÝ OBZOR
VOL. 15, ISSUE 1, 2015, pp. 49–69, DOI: 10.1515/revecp-2015-0011
Assessments of Modal Split in Long-distance
Passenger Transport1
Jakub Chmelík2
Abstract: The goal of this paper is to present basic alternative assessments of the division
of transport work (or also “modal split”) of main transport modes. For this, an
example of selected long-distance connections between centres in Czechia, including an
identification of major underlying factors, shall be exploited. The paper examines the
competitiveness of rail transport in its relation to bus and individual automobile
transport, and relations with a potential of rail transport are primarily selected. A logit
model is applied within the selected set of relations. It is entered in particular by indicators
of time (time of a ride, frequency of public transport lines) and finances (actual
transport costs) through a financial expression of generalised costs. The purpose of the
paper is to verify the selected procedure on the relations transport modes of which are
similar, and to highlight the alternatives of a comprehensive assessment of the modal
split of main transport modes in Czechia. In the conclusion, the gained results are used
to outline further alternative prospects of the topic under observation.
Key words: transport modes, modal split, competitiveness of rail, transport accessibility,
generalized costs
JEL Classification: C25, O18, R41
Introduction
The topic of identifying the division of transport work (or also “modal split”) and the
factors underlying the choice of the actual means of transport have been a major research
theme of the scientific disciplines examining the questions of transport in the
long run. The demand for a study of the topic mentioned above is primarily caused by
generally low consciousness of the share of individual transport modes in the modal
split in a given relation that is mostly influenced by the usual ignorance of direction
orientation of individual automobile transport as the main bearer of transport interactions.
The absence of the necessary data has also a considerable impact on solution to
1
This contribution was prepared thanks to research project number P404/12/1035 Spatial Dynamics
of Transport Relationship in Settlement System of Czechia granted by Grant Agency of the
Czech Republic. Author would like to thank Miroslav Marada and Viktor Květoň for valuable
comments at the beginning of the study. Author also thank two anonymous reviewers for their
helpful remarks and suggestions.
2
Charles University in Prague, Faculty of Science, Department of Social Geography and Regional
Development, Transport Geography Research Centre, Albertov 6, 128 43 Praha 2, Czech
Republic; jakub.chmelik@natur.cuni.cz
REVIEW OF ECONOMIC PERSPECTIVES
50
practical tasks, especially in the sphere of strategy of the planning of offer of public
transport as well as the planning of the influence of new transport infrastructure on
modal split. In the conditions valid in Czechia, the urgent problem of identifying the
real share of individual transport modes or individual carriers operating within the same
relation or section in the modal split is associated with the questions of competition on
the market in passenger transport within the economic competition policy. This involves
specific tasks such as finding a suitable methodological mechanism for delineation of a
relevant market (Kvizda 2011; Kvizda, Rederer 2012).
Competitiveness of individual transport modes within the modal split is influenced by a
number of factors that are associated to a great extent with real spatial links of the centres
arising from the geographic and transport position of a centre in the networks of
individual transport modes (Marada 2006). They are closely connected to the character
of settlement and the size of centres on the side of demand for transport, and with the
level of public transport and transport infrastructure on the side of offer. In general, the
factors influencing the choice of the mode with an impact on modal split can be divided
into three groups (Ortúzar, Willumsen 2001; a similar division also e.g. Strandling,
Anable 2008). The first group is constituted by a set of characteristics of the actor (passenger)
that mainly include access to a car, driving license, marital status (for example,
it can be presumed that the use of a car is influenced by the number of members of a
household), income and type of employment (such as the possibility to use a company
car) and, last but not least, the characteristics of the actor’s place of residence (population
density, urban versus rural area, transport position, etc.). The second group includes
factors that influence the journey itself, when the choice of the mode is influenced by its
purpose (such as the difference in the use of the means of transport for commuting to
work or for a weekend trip journey, more e.g. Sheppard 1995) and the time when it
occurs (such as a low offer of public transport in night hours, etc.). The third group of
the underlying factors contains the characteristics of transport offer or transport opportunities
for the population that can be divided into two categories: Quantitative and
qualitative factors. The main quantitative factors that can be assessed rather well are as
follows: The travel time (the time spent in a means of transport including walking, waiting
at the stop, change, etc.), the price of transport (fare, fuel price, fixed operating costs,
etc.) and, possibly, also the accessibility (and price) of parking. On the other hand, it is
difficult to quantify qualitative factors that are often the most important circumstances
influencing the choice of a means of transport. They can be in discrepancy with an economically
rational choice. These “soft” factors primarily include comfort, conveniences,
reliability, safety, etc. (for more see Ortúzar, Willumsen 2001).
The topic of assessment of regional differences in the distribution of modal split has
only been examined marginally in Czech and Slovak surveys, primarily within a routine
assessment of changes in the relation of spatial organisation of society and transport
links. In the past decade, the attention has mainly been paid to studies evaluating the
transport importance of centres on the basis of offer by public transport (e.g. Marada et
al. 2010; Marada, Květoň 2010) or of road traffic volume (e.g. Kraft, Vančura 2009a),
the competitiveness of individual modes on the basis of time accessibility (e.g. Kraft,
Vančura 2009b; Seidenglanz 2009; Więckowski, Michniak et al. 2012; Horňák et al.
2013; Marada et al. 2014) or a combination of the above indicators along with theoretical
intensities of interaction of centres (Chmelík et al. 2010; Horňák, Pšenka 2013).
Volume 15, Issue 1, 2015
51
Specific case of change of modal split was focused on the study by Kvizda and Seidenglanz
(2014). They analysed short-term intermodal shift from air to railway transport
after Iceland´s volcano eruption in April 2010. The described studies are mainly based
on traditional research approaches in transport arising from quantitative and predictive
methods. They are often criticized due to their considerable generalization of real social
processes and motivation of individuals whose behaviour is considered absolutely rational
in these approaches (such as the choice of the shortest route). This focus is based
on the interdisciplinary nature of the study of transport. In it, an increasing role is
played by experts with a technical rather than social background (Hanson 2006), which
is also influenced by a shortage of “soft” theoretical approaches in transport as well as
transport geography itself (Goetz et al. 2004: quoted in Goetz et al. 2009). The existing
approaches in transport were challenged by a group of researchers (primarily sociologists)
associated within the “new mobilities paradigm” (in particular Sheller, Urry 2006,
with further discussions by Shaw et al. 2008; Keeling 2008). Their attention is paid to
the effort to clarify and understand the motivation of journeys and real human mobility
as well as the interest in emotional and symbolic components of transport behaviour
(Brůhová-Foltýnová et al. 2008) with the use of qualitative research methods. This type
of research that is primarily of sociological nature widens the discussion regarding the
factors underlying transport behaviour and motivation of the choice of a means of
transport (Strandling, Anable 2008). They often have a specific focus, examining the
sex, age, social status, subjective perception of feelings during a journey, etc. In Czechia,
the research of sociology of mobility, transport behaviour and the choice of the means
of transport was conducted by Brůhová-Foltýnová et al. (2008) and Braun-Kohlová
(2010), while the issues of daily mobility and day-to-day life were dealt with by Temelová
et al. (2011) and by Mulíček et al. (2013).
The presented contribution loosely follows up on a previous study (Chmelík et al. 2010)
that assessed the intensities of relations between Czechia’s regional towns (i.e. a set of
78 relations) and the use of passenger rail transport, evaluated by means of three indicators:
Offer of connections, the real demand, and modelled/theoretical interactions. The
results revealed some interdependencies of the above-mentioned indicators for the individual
relations, on the basis of which it was possible to identify a theoretical potential
of railway transport towards other modes. The more this was true when looked at from
the viewpoint of the relation of real demand (using the offer of rail lines) and model
interactions that characterised the intensity of theoretical links of the centres.3
All of this
was established without any deeper analysis of the modal split in a given relation or
without inclusion of rival modes, bearing in mind the fact that the gained results are
influenced by the initial set of relations with considerably differentiated structures.
The goal of the presented paper is to present basic alternative assessments of the modal
split of main transport modes. One of the chief methods will be applied on the example
of selected long-distance connections of centres in Czechia. At first, the basic alterna-
3
The basic form of a gravitation model was applied in which the distance between centres was
characterised by time accessibility on railway and the mass of centres by simple population num-
ber.
REVIEW OF ECONOMIC PERSPECTIVES
52
tives to evaluate the modal split will be examined, especially in the sphere of the database
available in Czechia, and a comparison to situation in Western-European countries
will be made. On the basis of a discussion, a technique will be chosen to evaluate the
share of main transport modes (individual automobile transport and public transport:
Passenger railway transport and regular bus transport) for each specific relation. In
accordance with a previous study (Chmelík et al. 2010) and the latest development on
the transport market, the choice will be concentrated on relations with the presumption
of a change in the potential of passenger rail transport. There will be a partial objective
of identifying specific factors underlying the modal split. On account of the availability
of the real data on modal split (see later text), the division of transport work will be
observed in 2001, 2006, and 2011. In the conclusion, the values predicted by the model
for 2011 (calibrated by the data from 2001, CSO 2003) will be compared to the real
results of intensity of commuting by the mode from the 2011 Population and Housing
Census (CSO 2013). This enables us to carry out a general assessment of whether the
model can be used for the pursued objective. In connection with general trends in
transport, one can presume in this interval an increasing share of individual automobile
transport in the total modal split, with a slower pace in the relations with a quality public
transport system. One can also presume that public transport will have the highest
share in the relations situated along rail corridors and possibly also along motorways
that provide an appropriate offer of lines. On the contrary, the lowest share can be expected
in the relations with a lower offer of public transport (frequency, necessity to
change), also determined by the quality of transport infrastructure. There is a similar
situation in the case of identifying the share of modal split of railway and bus transport
within the framework of public transport. In this sphere, one can presume a growing
role of the railway in the relations linked by quality infrastructure in connection with a
gradual modernisation of main lines and an extension of the offer of long-distance passenger
railway transport in Czechia roughly from 2005 onwards. On the other hand,
growing proportion of bus lines within the modal split of public transport can be presumed
in relations with insufficient connection to the rail network. When it comes to the
factors influencing the choice of the means of transport, in general it can be assumed
that unlike the lower order levels, the role of the size of a centre that influences the level
of transport opportunities for the population will not be an underlying factor in the
sphere of long-distance rail links between Czechia’s most important agglomerations. On
the contrary, a major influence of transport location in the networks can be expected.
The results themselves are followed up by the conclusion of the paper. Along with a
summary of the main results and confirmation/refutation of research assumptions, the
conclusion will include a critical assessment of the method used, including identification
of the main problems and a proposal of their alternative solutions in possible subsequent
research.
Alternative Assessments of Modal Split
In conditions of the Czech Republic, the basic assessment of modal split is rather limited
by the available database. In general, the data from transport yearbooks (Transport
Yearbook of the Czech Republic; Ministry of Transport 2013a; Transport Yearbook of
Prague; TSK 2013, etc.) are primarily available each year. Based on the mandatory
statistical enquiry ordered by the Transport Ministry, and on the surveys it conducts, the
Volume 15, Issue 1, 2015
53
yearbooks present values of transport output of individual modes in passenger transport.
However, neither there are any detailed data on directions, nor any additional information
(the purpose of a journey, etc.). Besides, it is not sufficiently evident in which
way the values were collected, especially in the cases in which it is said that this is an
expert estimate (such as output of individual automobile transport in the Transport
Yearbook of the Czech Republic). Due to this, results of the Population and Housing
Census (CSO 2003 and 2013) are the only comprehensive source of information on the
share of individual modes in the modal split in Czechia. Among others, the Census
records the data on the means of transport used for commuting to work and schools
down to the level of Czechia’s individual municipalities, which is an undeniable advantage.
However, the use of the data is also burdened with a number of problems.
Above all, there is an absence of a continual time series arising from a ten-year period
of observation and of additional time needed to process the results. Particularly in the
intercensal period it is necessary to be cautious when dealing with the information. This
is caused by a considerable dynamics of changes in day-to-day mobility of persons, in
the offer of public transport and transport infrastructure as well as changes in the
transport market in some relations. Moreover, there is one significant defect: Limited
time differentiation of the frequency of commuting. In the routinely available results, it
is only presented as daily and non-daily (total). Besides that, it can be presumed that a
part of the results was affected by a wrong or incomplete or deliberately false filling of
census sheets. The issue of the incomplete data is noticeable mainly in the results of
commuting flows from the 2011 census (CSO 2013), where the number of the captured
data states just 1.5 million commuters against more than 2.3 million in 2001 (CSO
2003). From the viewpoint of the study of the modal split, there is the associated problem
of the publication of the results by the Czech Statistical Office, as commuting is
presented differently in the commonly accessible outputs of the two latest censuses. The
2001 census (CSO, 2003) only published commuting by transport modes in connection
to movement of manpower. The results from the 2011 census (CSO 2013) also include
modal distribution by commuting to schools. This means that the differences eventually
reduce the data set that can be used for a comparison of both censuses. Obviously, the
data that also contain commuting to schools will differ in a number of relations, given
the assumption of a much higher share of the use of public transport in the case of this
target group. The use of the data from the censuses for comparison of the intervals mentioned
above is also affected by the fact that commuting was recorded from the place of
usual residence in 2011, but from the place of permanent residence in 2001. This change
in the census method has resulted in an apparent decrease in recorded direction of commuting
flows between centres in Czechia, which has the biggest impact just on longdistance
relations typical of non-daily (weekly) commuting. In connection with commuting
flows, also necessary is a discussion on the relevance of these data that inherently
only record regular movements within the journeys to work or school. However, a
number of other movements with a different motivation as far as the journey is concerned
is not recorded (shopping trips, visits to friends, outdoor and holiday trips). For-
REVIEW OF ECONOMIC PERSPECTIVES
54
eign surveys4
have revealed that the share of the journeys to work and to school accounts
for roughly 30% of the total mobility expressed by passenger-kilometres. On the
other hand, regular journeys (commuting and business trips) are usually the main purpose
in long-distance mobility. Travel behaviour in commuting journeys is also more
stable and homogenous than in case of others purposes (mainly leisure purposes journeys
and visits to relatives and friends) (Hubert, Potier 2003). The surveys of transport
behaviour which are conducted mainly in Western-European countries can serve to
some degree to widening of the database about the modal split because one of the main
spheres of interest is constituted by the questions about the destination (direction) of a
journey, the means of transport used, and reason of the journey. The surveys are usually
held under the auspices of the civil service, which ensures continual and guaranteed
results. Although such surveys would be certainly profitable for the decisive actors (the
Ministry of Transport, regions, and carriers) and transport researchers in Czechia, due to
the universal character of the survey, the results cannot cover all the specific needs. This
is the reason why, in the case of long-distance passenger transport,5
specific surveys are
often carried out by the carriers themselves, who thus try to identify the potential of
individual relations cutting across transport modes for their own business activities.
However, results of such surveys are eventually not made public. The issue of transport
surveys in long-distance transport and associated spheres is examined in a great detail
by a study of the research team Axhausen et al. (2003).
If the real data are not accessible, the share of individual types of transport in the relation
under observation is examined by the means of theoretical models arising from an
aggregate of individual patterns of transport behaviour (in the case of a representative
sample) and presumed reactions of a system to their changes. This process is represented
by a group of “discrete choices” made by individuals before a journey and in its
course. Specifically, a discrete choice includes the selection between two (possibly
more) discontinuous alternatives. In the case of the topic under consideration, it is a
choice between various means of transport (Brůhová-Foltýnová et al. 2008). In the
modelling of modal split, the most frequently used methods include logit and probit
models and in the case of a choice from more than two alternatives, these are their multinomial/multiple
forms (Pas 1995; Pipkin 1995; Ortúzar, Willumsen 2001). In the
sphere of transport, the “nested logit model” (a hierarchical model) is often used for the
basic modelling of the main modes of transport. It presumes a “nested” structure of the
decision-making process. This means that first, the choice is limited to the relation of
individual and public transport (a “bimodal distribution”), and in the second step, an
4
Such as the transport survey “Swiss Microcensus on Travel Behaviour” in which transport
behaviour of the Swiss is observed in a roughly five-year period (e.g. Simma 2003). Similar
information is provided by the “National Travel Survey,” an annual survey of persons’ mobility in
Britain (DFT 2013). An overview of transport behaviour is provided for example by Marconi et
al. (2004).
5
In foreign literature, long-distance transport is usually defined as transport of persons over 80–
100 kms, or a trip with an overnight stay (Marconi et al. 2004; Frei et al. 2010; Hubert, Potier
2003).
Volume 15, Issue 1, 2015
55
evaluation of the main modes of public mass transport (train, bus) is made. The principle
of modelling is based on the assumptions of the rational choice theory in which
based on their sources (income) and limitations (primarily time and transport costs),
passengers make a decision between at least two alternatives, following a certain rule in
their decisions (Brůhová-Foltýnová et al. 2008). The multinomial form of the logit
model arises from the following relation (e.g. Pas 1995; Ortúzar, Willumsen 2001):
ܲ௡௜ ൌ
݁ఓ௏೙೔
∑ ݁ఓ௏೙ೕ
௡௝ఢ௃
where ܲ௡௜ is probability of the choice of the observed means of transport for a given
journey from the point A to the point B; ݁ is Euler's constant, ܸ௡௜ are generalised costs
of a journey from the point A to the point B; ߤ is the parameter of sensitivity expressing
the elasticity of the extent of change in the means of transport on the basis of a change
in generalised costs between alternative modes of transport; and ‫ܬ‬ is the set of all alternative
modes of transport in the observed relation A–B. There is a substantial methodological
question associated with the application of models: The construction of generalised
travel costs whose value influences the probability of the use of individual modes
(see Chmelík, Marada 2010; Chmelík et al. 2012). In the case of individual automobile
transport, these are basically all financial costs spent on a journey (i.e. the price and
average consumption of fuel, fixed costs of the car operation – maintenance, insurance,
vignette, etc.); in public transport, this is the fare (including customer applications determining
the discounts). In all evaluated modes, there is also the time spent on a journey
from point A to point B, including the walk to the car, the search for a parking place
or the journey to a public transport stop. The given travel time is subsequently converted
into a financial value. The issues of value of time are discussed by Jain and Lyons
(2008). In the construction of generalised travel costs, one can often see a constant
expressing the unrecorded qualitative aspects that can be quantified with difficulties
only. These are comfort, what one feels during a journey, the use of time during the
journey, etc., and they considerably influence the decision on the choice of a means of
transport especially in the relation “individual versus mass transport” (Riley et al., 2010).
The parameter of sensitivity to the change in the generalised travel costs is another
factor that enters the model. It is gained by calibration (most often by the maximum
likelihood method) of the real data obtained by a transport survey or by the use of available
aggregated data on the modal split in the relation under observation. This technique,
which is usually used in the classical four-step transport model (e.g. Pas 1995; Ortúzar,
Willumsen 2001), naturally includes only one of the alternatives (but the most known)
of modelling describing the choice of the means of transport or of the modal split.6
6
The method of transport resistors is another technique used in transport planning (Hrabáček
2010; Drdla 2010). In the transport and planning practice, the assessment of modal split is often
conducted in a specialised software environment such as VISUM from the company PTV AG that
has the advantage of possible link with GIS, programmes for the construction of timetables, etc.
In sociologic disciplines, “activity based approach” has been newly applied, where a journey is
considered to be only one of the attributes of activity. This is quite unlike the presented approach
REVIEW OF ECONOMIC PERSPECTIVES
56
Methodological Remarks on the Case Study
Sixteen long-distance relations in Czechia were selected for the case study evaluating
the modal split. Their travel source/destination was situated in Prague or in Brno. The
evaluation focused on the share of individual automobile transport and public transport.
The latter was further divided into the share of bus and rail passenger transport. For the
study, the relations in which major changes in time accessibility occurred in one of the
modes were primarily selected, with an assumption of impact on the modal split between
2001 and 2011. A choice was also made of the relations in which changes occurred
on the transport market that influence the potential of the use of rail transport by
passengers. As a rule, the modal split is assessed between the cores of agglomerations
themselves. In some justified cases, further agglomerated centres in which the same
transport behaviour of the population as in the core is expected were added to the core
of agglomeration. The selection of the relations is presented in Table 1 and Figure 1.
The size criterion for the selection was constituted by the value of 100 commuters according
to the means of transport in the summary of both directions, as recorded within
the 2001 Population and Housing Census (CSO 2003). Just the figures on work commuting
by modes in 2001 (school commuting is not available for the 2001 Population
and Housing Census) are also used as calibration data for the modelling of the modal
split as of the observed years 2001, 2006 and 2011. In this respect, it is necessary to
again point out the fact that the data used for the comparison with the figures from 2001
only include the movements for work that capture roughly 20–50% of all commuting
movements identified. The latter only account for a fraction of total mobility (an estimate
of less than 10%). The case study is based on the assumption that the modal split
in the sphere of commuting flows will display similar distribution in all purposes of
journeys.
Table 1 Selection of Relations for the Assessment of Modal Split
Origin Destination
Prague
České Budějovice, Tábor, Plzeň, Cheb (Cheb, Františkovy Lázně), Ústí Region
(Ústí nad Labem, Teplice, Děčín), Hradec Králové, Pardubice, Brno, Zlín+Uherské
Hradiště (Zlín, Otrokovice, Uherské Hradiště, Staré Město, Kunovice), Olomouc,
Ostrava Region (Ostrava, Opava, Havířov, Karviná), Vsetín+Valašské Meziříčí
Brno
Ostrava Region (Ostrava, Opava, Havířov, Karviná), Olomouc, Zlín (Zlín, Otrokovice),
Uherské Hradiště (Uherské Hradiště, Staré Město, Kunovice)
in which all the processes (and choices) are connected with the journey (“trip based approach”)
(Brůhová-Foltýnová et al. 2008).
Volume 15, Issue 1, 2015
57
Figure 1 Determination of selected relations and major transport network (in 2011)
Source: CSO 2013
For the sake of an estimation of the modal split, a logit model described above, specifically
the “nested logit model,” was used. At first, generalised travel costs for individual
automobile and public mass transport were constructed for the observed intervals. In the
case of the assessed set of relations, the value of generalised costs relating to a journey
by individual automobile transport was calculated by the following formula (adjusted by
Riley et al. 2010):
ܸூ஺் ൌ ൬‫ܦ‬௞௠ ∗
ܲூ஺்
ܱூ஺்
൰ ൅ ሺ‫ܦ‬ூ஺்௧௜௠௘ ∗ ܹ௧௜௠௘ሻ ൅ ሺ‫ܦ‬ூ஺்௔௖௖ ∗ ܹ௧௜௠௘ሻ
where the value ܸூ஺் represents generalised costs of individual automobile transport.
The length of a journey (‫ܦ‬௞௠), given in kilometres, was recorded as of 2012 by the
means of a journey planner (Journey planner 2012). It was estimated for the observed
intervals on the basis of the state of construction of the motorway network. In a similar
way, the distance in time (‫ܦ‬ூ஺்௧௜௠௘) between the cores of agglomerations was estimated.
In the case of multiple-core agglomerations, the value was calculated as referenced to
their geographic centre, taking into account the weight of population size. At the same
time, there was the inclusion of the indicator of accessibility (‫ܦ‬ூ஺்௔௖௖) or walk to an
automobile as well as search for the parking place etc. that also rectified the differences
between the time accessibility of centres. The bigger centres were ascribed a higher
value of accessibility due to the generally large dispersion of values (town centre versus
outskirts) in the time accessibility itself (‫ܦ‬ூ஺்௧௜௠௘). The operating costs of an automobile
per 1 kilometre of a journey (ܲூ஺்) are constructed as a sum of average consumption
of fuel of a usual car (the paper reckons with the value of 7 litres per 100 kms) plus
REVIEW OF ECONOMIC PERSPECTIVES
58
minimal average fixed costs (e.g. highway fee, car insurance, periodically repairs, etc.)
calculated in the compensation tariffs for the use of automobiles at business trips (prices
of fuel and reimbursement of travel, Ministry of Finance 2012).7
For this reason, operating
costs are considerably generalised and do not take into account any regional specificities.
Due to the accessibility of the relevant data, the values for Czechia were used in
all cases. The average occupancy of an automobile (ܱூ஺்) was set to two persons. The
value best corresponds to information from foreign studies in the sphere of longdistance
transport (e.g. Hubert, Potier 2003). However, total values for all transport
segments are usually lower as, for example, in the latest survey of transport behaviour
conducted in Switzerland (FSO 2012), where the average occupancy of an automobile
was in the interval of 1.12 to 2.05 depending on the purpose of the journey (daily commuting
to work versus leisure time activities). The model included another component:
The indicator of perceived hourly value of the time spent on a journey (ܹ௧௜௠௘), construction
of which was based on a methodological manual for modelling created for the
Transport Ministry (Riley et al. 2010). The indicator included the average sum of hourly
wages in regions (average gross monthly salary in 2001, 2006, 2011; CSO2012) which
was multiplied by the coefficient 0.24 (adopted from Riley et al. 2010, p. 78) in which
the observed agglomerations for specific relations were found.
Generalised costs for the connection by public transport (by bus and train) were subsequently
constructed for the same set of relations. The technique was the same for both
modes. When a comparison was made with automobile transport, there was a consideration
of the values for the mode that is more advantageous from the viewpoint of a rational
passenger, especially as far as time is concerned. The calculation for public
transport was based on the following relation (adjusted by Riley et al. 2010):
ܸ௉் ൌ ൣሺ‫ܦ‬௉்௧௜௠௘ ∗ ܹ௧௜௠௘ሻ ൅ ሺ‫ܦ‬௉்௔௖௖ ∗ ܹ௧௜௠௘ሻ ൅ ‫ܨ‬ ൅ ‫ܫ‬ଵ/ଶ൧ ∗ ‫ܥ‬௉்ିூ஺்
where the value of ܸ௉் represents generalised costs of public transport. Similarly to the
case of automobile, the time accessibility to a means of transport (‫ܦ‬௉்௔௖௖ – a journey to
a railway station, stop) was gained by an estimate and an indicator of perceived hourly
value of the time of a person on a journey (ܹ௧௜௠௘) was constructed. The distance in time
of centres (‫ܦ‬௉்௧௜௠௘) is based on travel times of buses and trains to the most frequented
railway and bus stations given in specific timetables (electronic timetable – IDOS 2001,
2006, 2011). In the case of multiple-core centres, the travel time was adjusted by the
geographic centre of the agglomeration, similarly to the case of automobile connection.
Modal travel time8
was used which most corresponds to the real systemic offer and is
not affected by extra links. The extent of the offer itself (‫ܫ‬ଵ/ଶ) was quantified as one half
of the typical interval of a line (lines) in a relation, which expressed the discontinuous
offer of public transport unlike that of individual transport. When it comes to the nonsystemic
offer with a minimum of lines per day, the value of a four-hour interval was
7
Fixed cost for the setting of the operating costs of an automobile per 1 kilometre was calculated
for about CZK 1 per km in average. This value is very close of the value used in the methodological
study by Riley et al. (2010).
8
Average travel time was used for the relations with differing transport offer.
Volume 15, Issue 1, 2015
59
used. The usual March Wednesday9
was the referential day for the calculations from the
timetable. Direct financial costs borne by a passenger are expressed by the price of the
fare (‫.)ܨ‬ In the case of rail transport, the fare was related to the prices of passenger fare
(i.e. after a 25% discount from the basic fare; without the inclusion of relation discounts)
for 2011 as given by the tariff of the Czech Railways (ČD TR 10 2010), the most important
carrier in Czechia. Due the inaccessibility of older data, the prices of bus tickets
are related to the values valid in spring 2012. As there was a variety of carriers, the
figure depicted for the lines in the station finder (IDOS 2012) was almost invariably
used as the most common price. Due to their difficult availability, the data for 2001 and
2006 were estimated. In principle, the latest value in a time series was always adjusted
by a coefficient that roughly respected the development of the railway tariff of the
Czech Rail. When it comes to the construction of generalised costs for public transport,
the calibration constant of the transport mode choice (‫ܥ‬௉்ିூ஺்) that ensures a concurrence
of the real and model modal splits was also included.
After the generalised travel costs were constructed, the logit model was calibrated. This
required that the value of the parameter µ be found, i.e. parameter which, based on a
change in generalised costs among alternative modes of transport within the set of observed
relations in which similar reactions to changes are presumed, expresses the elasticity
of change in the means of transport. The data on commuting by modes from the
2001 Population and Housing Census (CSO 2003) were used for the calibration. The
value µ was estimated by the maximum likelihood method, at first for the relation between
individual automobile transport and public mass transport. In the second step, it
was applied on the relation between the rail passenger transport and regular bus
transport. In the case of the first step it was also necessary to estimate ‫ܥ‬௉்ିூ஺் (the constant
of the transport mode choice) in such a way that the value of µ was as close as
possible to the interval appearing in foreign studies10
(for greater detail see Riley et al.
2010, p. 88). A weighing of the costs of public transport by the constant ‫ܥ‬௉்ିூ஺் was
necessary because simple costs of public transport were always considerably lower than
those of the automobile. In reality, this would mean that individual automobile transport
has a negligible share in the modal split. This disproportion is most affected by the
difficult quantification of soft, non-financial influences (quality, comfort, privacy, security,
etc.) on transport behaviour of population. The most suitable parameter for the
relation between individual automobile and public transport ߤ ൌ െ0.0031 was reached
with the constant ‫ܥ‬௉்ିூ஺் ൌ 2.2. When it comes to the relation between rail passenger
and regular bus transport, the parameter reached the value ߤ ൌ െ0.0123. The values of
the parameters were subsequently used for an estimate of the modal split model for the
intervals under observation.
9
The usual March Wednesday was selected because it poses no limitations (weekends, Easter
holiday were in April in 2001, 2006 and 2011) or additional measures (for example additional
trains in peaks on Fridays and Sundays) (see also Chmelík et al. 2010).
10
Interval between -0.1 and -0.01.
REVIEW OF ECONOMIC PERSPECTIVES
60
Results
A comparison of the share of individual automobile transport and public transport in
individual relations has yielded interesting results. In general, there is an obvious falling
share of public transport in the modal split in the period under observation, while higher
dynamics of changes is predicted for its second half. This is apparent from Table 2,
where the value of the share of public transport in the modal split is compared to the
average value of public transport for the whole Czechia calculated from the statistics of
the Ministry of Transport11
(Ministry of Transport 2013a). According to the model, out
of the 16 monitored relations, 13 were above the average in 2001 and 2006, but only 10
in 2011. A general growth in the use of automobile in the model was enhanced, among
others, by a large number of motorways and further road projects having been put into
operation, which influenced the time accessibility of the observed centres with an impact
on the lowering of generalised travel costs. There was only the following exception
in the survey: the relation Prague–Vsetín/Valašské Meziříčí, where the proportion of
public transport increased slightly between 2006 and 2011, probably due to the high
level of the train connection in 2011. In the relation Prague–Pardubice where, according
to the model, public transport accounted for the biggest share (mainly high share of
railways), for over 40% of the transport demand between 2001, 2006 and 2011. As far
as all other relations are concerned, there was always a predominance of the use of
automobile in 2001, 2006 and in 2011, usually exceeding 60%. However, it must be
noted in this respect that the decreases cannot be appropriately assessed and they are
rather questionable in a number of cases. Given the large variety of offer and travel time
on rail, for example, the 9% fall in the relation Prague–Ostrava Region between 2006
and 2011 is unlikely. This is even more unlikely at present, when three carriers operate
in the relation Prague–Ostrava Region/Olomouc, while its representatives stated that the
market share of the railways has increased here. However, due to the time limit as of
2011, the latest change in offer was not included in the estimate. As a rule, the most
important values above the described national average of the share of public transport
were reached among the relations with the location on rail corridors or with an adequate
offer of the connections of both main modes of public transport. The model ascribed the
lowest proportion of the use of public transport to the relations that are less frequented
from the viewpoint of intensity of transport (Prague–Cheb, Vsetín/Valašské Meziříčí) or
are described by a rather small set of the real data entering the calibration, which affects
their informative value. Moreover, the real values of the share of observed modes correspond
to the model data just in these relations. A similarly considerable difference between
real and model data is obvious in the relation Prague–Pardubice, where the model
predicted a higher share of the use of individual automobile transport than that found in
the real data. In general, it is necessary to stress again in this connection that the real
data presented in the Table 2 and that were used for the calibration of the model only
include the journeys to work. In a number of relations, this fact can strongly affect the
11
The value of the share of public transport (33% in 2001, 31% in 2006, and 32% in 2011) arises
from the sum of transport outputs for railway, bus and urban mass transport.
Volume 15, Issue 1, 2015
61
share of public transport that is often the main bearer of transport demand in commuting
to schools.
Table 2 Resulting Estimates of the Share (%) of Individual Automobile Transport and
Public Transport in Modal Split (2001, 2006, 2011)
Relation/Year
2001 R 2001 E 2006 E 2011 E 2011 R
IAT PT IAT PT IAT PT IAT PT IAT PT
Prague – České Budějovice 67 33 60 401 64 361 70 30 67 33
Prague– Tábor 65 35 58 421 61 391 64 361 59 41
Prague – Plzeň 63 37 59 411 62 381 63 371 64 36
Prague – Cheb Region 55 45 72 28 76 24 81 19 61 39
Prague – Ústí Region 53 47 61 391 63 371 66 341 55 45
Prague – Hradec Králové 70 30 57 431 60 401 63 371 61 39
Prague – Pardubice 50 50 54 461 55 451 59 411 40 60
Prague – Brno 72 28 64 361 67 331 72 28 68 32
Prague – Zlín+Uh.Hradiště 67 33 71 29 77 23 80 20 72 28
Prague – Olomouc 57 43 60 401 63 371 67 331 50 50
Prague – Ostrava Region 59 41 64 361 65 351 74 26 56 44
Prague – Vsetín+V.Meziříčí 68 32 75 25 81 19 79 21 53 47
Brno – Ostrava Region 67 33 59 411 59 411 67 331 67 33
Brno – Olomouc 61 39 60 401 62 381 67 33 75 25
Brno – Zlín 70 30 61 391 63 371 67 33 79 21
Brno – Uherské Hradiště 58 42 58 421 59 411 62 381 72 28
Note: 1) the value of the share of public transport is higher than the average value of the share of
public transport for the whole Czech Republic, R = real data, E = model estimate
Source: CSO 2003, 2013, author’s calculations
The results of the second model that estimated the shares of railway and bus transport in
the modal split in public transport deserve a comprehensive discussion. As generalised
costs are constructed in the same way in both cases, one can expect it to have higher
informative value. Besides, unlike the previous model, the real values of the shares can
be more easily estimated thanks to the knowledge of the real offer of connections that is
supposed to better reflect the real demand arising from the time competitiveness of both
observed modes. Based on the results presented in Table 3, it can be stated that according
to the estimate of the model, there were no major transformations of the prevailing
mode in the individual relations in the period under observation. In 2001, railway
transport was more used in 7 out of 16 relations, and in 8 relations in 2006 and 2011.
From the viewpoint of the development of the shares, one can delineate several groups
of relations with similar characteristics.
The first group is constituted by the relations where a higher share of railway transport
was estimated by the model in all observed intervals. Higher dynamics of changes was
REVIEW OF ECONOMIC PERSPECTIVES
62
rather predicted for the second half of the examined period. This fact is primarily influenced
by the extended variety of offer of rail transport that was associated with the use
of modernised railway corridors. In fact, a higher share of railway transport was also
recorded in the case of real values from 2011. When it comes to spatial distribution, this
group primarily included the relations heading from the east to Prague, which means the
lines from Pardubice, Olomouc, the Zlín Region and the Ostrava Region. Despite a high
share of railways in these relations, the results are questionable because due to the negligible
offer of bus connections, even lower use of buses can be expected in reality. This
is apparent for example in the relation Prague–Pardubice, where the model for 2011
predicts a 13% share of bus, but the real data on commuting say it only amounts to 6%.
In fact, the bus connection is roughly twice as long as far as time is concerned, and it is
also necessary to change in Hradec Králové. In this respect, one can discuss the methodological
question of time disadvantage of interchange lines (at least of the bus lines
that are not common, unlike the railway network offer) or the highest possible interval
of offer (the study reckons with one half of the four-hour interval). Along with the
above, i.e. typical railway relations, the group can also include the relations Brno–
Ostrava Region and Prague–Cheb in which a higher share of railway transport use was
recorded in all observed years, but in the first half of the observed period it was decreas-
ing.
The second group is formed by the relations in which higher proportion of bus transport
in the modal split of public transport was recorded in all observed intervals. Basically,
this referred to the selected radial link to Prague (from Brno, Plzeň, České Budějovice,
Tábor and Hradec Králové) and the connection to Brno from central and eastern Moravia
(Olomouc, Zlín and Uherské Hradiště). The result in these relations was mainly
determined by railway infrastructure which – when it comes to time accessibility – is
able to compete with road transport only to a limited degree. In the case of Czech relations,
apart from a lower model share (though growing between 2001 and 2006) in the
connection Prague–Plzeň/České Budějovice, one could see a steep fall in real values of
the share between 2001 a 2011, which occurred, according to the model, in the second
half of the observed period. This trend can be associated with the ongoing construction
of the third and fourth transit rail corridors that had a negative impact on travel comfort
(prolongation of travel times, instability of the timetable, or frequent closures). On the
other hand, the bus competition offers here a relatively wide and systemic offer of links.
The most significant discrepancy between the model and the real shares was found in
the relation Prague–Plzeň, where the model ascribed to the railway an almost double
proportion compared with the data on commuting by mode in 2011. In the case of Moravian
“bus” relations, the model slightly underestimates the share of the rail in the
connections of Brno and Olomouc, as compared with the real shares, while it overestimates
it in the relations Brno–Uherské Hradiště/Zlín. Due to a longer travel time, the
systemic offer of railway connection for passengers probably does not offer any adequate
competition here.
The relation Prague–Ústí nad Labem remained in the third group. Partial changes in the
preferred mode occurred in it between the observed intervals. In 2001, bus transport had
a slightly higher share, but it was railway transport in the following intervals. A significant
increase was recorded between 2006 and 2011. The result can be attributed to the
reconstruction of the line connecting Prague with Ústí nad Labem, and the subsequent
Volume 15, Issue 1, 2015
63
widening of the offer of connections after 2006. Despite this, the resulting model share
of rail in 2011 is still much lower than the real values by the mode used for commuting.
Table 3 Resulting Estimates of the Share (%) of Train and Bus in Public Transport Modal
Split (2001, 2006, 2011)
Relation/Year
2001 R 2001 E 2006 E 2011 E 2011 R
T B T B T B T B T B
Prague – České Budějovice 52 48 46 54 48 52 45 55 44 56
Prague– Tábor 51 49 44 56 46 54 45 55 52 48
Prague – Plzeň 39 61 43 57 43 57 36 64 21 79
Prague – Cheb Region 66 34 54 46 53 47 65 35 49 51
Prague – Ústí Region 58 42 49 51 51 49 68 32 81 19
Prague – Hradec Králové 53 47 41 59 45 55 42 58 35 65
Prague – Pardubice 87 13 77 23 84 16 87 13 94 6
Prague – Brno 26 74 37 63 34 66 37 63 34 66
Prague – Zlín+Uh.Hradiště 44 56 59 41 62 38 73 27 58 42
Prague – Olomouc 85 15 82 18 82 18 93 7 89 11
Prague – Ostrava Region 65 35 76 24 71 29 89 11 85 15
Prague – Vsetín+V.Meziříčí 56 44 67 33 71 29 82 18 79 21
Brno – Ostrava Region 73 27 63 37 53 47 66 34 78 22
Brno – Olomouc 55 45 49 51 47 53 49 51 51 49
Brno – Zlín 20 80 44 56 44 56 45 55 30 70
Brno – Uherské Hradiště 31 69 39 61 36 64 36 64 24 76
Note: T = train, B = bus R = real data, E = model estimate
Source: CSO 2003, 2013, author’s calculations
In general, the results of both models mainly confirm the changes in significance of
modal split of public transport caused by important infrastructure improvements of
railway corridors. In the field of long-distance rail transport were in this context increased
number of trains which were able to compete road transport in the travel time.
On the other hand, the real results could perhaps be better because only selected railway
infrastructure projects had been finished in the past decade. This situation is also influenced
by quite a low level of the infrastructure planning in Czechia after 1989, when
many projects were conceived by the old transport strategy and did not reflect the
change of transport demand and traffic behaviour. The hierarchy of priorities of new
infrastructure and also comprehensive study based on the transport model including
modelling of the changes of modal split was not made in Czechia in the past. (Chmelík,
Marada 2014). Absence of complex view of transport infrastructure strategy is newly
dealt by the Ministry of Transport by the project "Transport Sector Strategy, 2nd Phase".
This strategy is based on the new national multi-modal transport model (Vachtl et al.
2013).
REVIEW OF ECONOMIC PERSPECTIVES
64
Conclusion
The goal of this paper was to present basic alternatives of the assessment of modal split
by the main transport modes. One of the chief methods was subsequently applied on a
delineated set of long-distance relations in Czechia. Before the case study was drafted,
the issue of accessibility of the real data on the use of individual transport modes had
been discussed. In this connection, the situation in the European countries with advanced
transport conditions (Switzerland, Germany, the United Kingdom, etc.) was
briefly outlined. In these countries, rather comprehensive transport surveys have been
made that considerably expand the information base with the knowledge transport behaviour
of the population. It is important to note that transport surveys are usually conducted
by decision-making bodies as well as by organisations that report to them. In this
respect, it may be desirable to stage a relevant discussion on the issue in the Czech Republic
and to define the need of the data and end users. As the need of any surveys of
transport behaviour and consumer polls is not mentioned even marginally in the basic
strategic document named Transport Policy of the Czech Republic for 2014–2020 With
a Prospect Until 2050 (Ministry of Transport 2013b), the Ministry of Transport cannot
be expected to take any major initiative in this field. In fact, some information, primarily
that gained by a survey, can have vital impact on many spheres of transport planning.
In the case study, this can be exemplified by the used weight of the perceived value of
the time spent on a journey (by the type of vehicle) or while waiting for a connection
that is usually adopted from foreign studies in the conditions of Czechia. The value is
further used for economic assessment, for example, but it is possible that due to different
transport behaviour of population and its economic situation, etc., the results of the
studies in Western Europe do not reflect the assumptions expected in the case of
Czechia.
The core of the contribution was formed by a case study within which the used methodology
was tested. It involved the construction of “universal” generalised travel costs and
a subsequent application of the logit model. This resulted in a rather interesting comparison
of the shares of individual transport modes in the relations under observation. In
general, the assumptions were confirmed. The model estimate revealed that the share of
individual automobile transport was rising in the observed intervals and, in a number of
cases, even more than in the case of the development of the real data, although they
included one part of common mobility. As suggested, when it comes to the share of rail
and bus transport, higher growth was recorded in the relations along rail corridors, especially
in the second half of the observed period. The development was not so dynamic in
the relations with a prevailing proportion of bus and a in a number of cases it even stagnated.
It is important to note that the model results are influenced by the usage of a set
of calibration data as of 2001, and furthermore significantly follow the definition of
travel costs. These are far from being able to incorporate all the relevant factors influencing
the decisions of actors of the transport process in their transport behaviour.
However, in the case of long-distance relations, it is difficult to thoroughly verify the
informative value of the model due to the size of the transport market, especially in
individual journeys. This suggests verification by a similar survey on a lower regional
order that could then be confronted with a detailed field survey in the area. Nevertheless,
the results confirmed that the application of the basic logit model can give very valuable
outputs.
Volume 15, Issue 1, 2015
65
References
AXHAUSEN, K. W., MADRE, J.-L., POLAK, J. W., TOINT, Ph. (eds.). (2003). Capturing
Long-Distance Travel. Research Studies Press, Baldock, 342 p.
BRŮHOVÁ-FOLTÝNOVÁ, H. a kol. (2008). Analýza každodenního dopravního chování
dospělého městského obyvatelstva a nástroje regulace dopravy. Závěrečná zpráva
z projektu MD 24/2006-430-OPI/3 z OP „Infrastruktura“ – Priorita 2 (2.4). Univerzita
Karlova v Praze, Centrum pro otázky životního prostředí, 281 p.
BURGE, P., KIM, CH. W., ROHR, CH. (2011). Modelling Demand for Long-Distance
Travel in Great Britain: Stated preference surveys to support the modelling of demand
for high-speed rail. Prepared for the UK Department of Transport, RAND Europe, 65 p.
ČD TR 10 (2010). Czech Railways Tariff for domestic passengers and baggage
transport, change no. 4 valid from 12.12.2010.
CHARLTON, C., VOWLES, T. (2008). Inter-urban and regional transport. In:
Knowles, R., eds.: Transport geographies: Mobilities, Flows and Spaces. Blackwell
Publishing Ltd., Oxford, pp. 120−136.
CHMELÍK, J., KVĚTOŇ, V., MARADA, M. (2012). Možnosti hodnocení “modal
splitu” hlavních dopravních módů v Česku: příklad vybraných dálkových spojení v
Česku. In: Seminář Telč 2012: Konkurence na železnici – budoucnost pro 21. století
nebo destrukce sítě? (presentation), 12 p.
CHMELÍK, J., MARADA, M. (2010). Konkurence dopravních módů na vybrané relaci
v Česku. In: Seminář Telč 2010: Konkurence na evropských železnicích − ekonomické,
právní a regionální faktory (presentation), 15 p.
CHMELÍK, J., MARADA, M. (2014): Assessment of the impact of a new motorway
connection on the spatial distribution and intensity of traffic flows: A case study of the
D47 motorway, Czech Republic. Moravian Geographical Reports, Vol. 22, No. 4, pp.
14–24. DOI: 10.1515/mgr-2014-0020
CHMELÍK, J., KVĚTOŇ, V., MARADA, M. (2010). Evaluation of competitiveness of
rail transport on example of connection among regional capitals in Czechia.
Národohospodářský obzor / Review of Economic Perspectives, 10, No 1, Masarykova
univerzita v Brně, Brno, pp. 5–20.
CSO (2003). Population and Housing Census 2001: Commuting to work and schools:
The Czech Republic. Czech Statistical Office, Prague.
CSO (2012). Average gross salary 2001, 2006, 2011: Regional Time Series. Czech
Statistical Office. Retrieved April 17, 2012, from http://www.czso.cz/csu/redakce.nsf/i/
regionalni_casove_rady
CSO (2013). Population and Housing Census 2011: Commuting to work and schools:
The Czech Republic. Czech Statistical Office, Prague.
DFT (2013). National Travel Survey. Department for Transport. Retrieved October 15,
2013 from https://www.gov.uk/government/ collections/national-travel-survey-statistics
REVIEW OF ECONOMIC PERSPECTIVES
66
DRDLA, P. (2010). Posouzení využití metodiky dopravních odporů. Vol. V, No 1,
Perner´s Contacts, pp. 47–56.
FREI, A., KUHNIMHOF, AXHAUSEN, K. W. (2010). Long distance travel in Europe
today: Experiences with a new survey, paper presented at the 89th Annual Meeting of
the Transportation Research Board, Washington, D. C., January 2010. Retrieved October
15, 2013, from http://e-collection.library.ethz.ch/eserv/eth:715/eth-715-01.pdf
FSO (2012): Mobilität in der Schweiz: Ergebnisse des Mikrozensus Mobilität und
Verkehr 2010. Neuchatel, Federal Statistical Office, 115 p.
GOETZ, A. R., VOWLES, T. M., TIERNEY, S. (2009). Bridging the QualitativeQuantitative
Divide in Transport Geography. Professional Geographer, 61, No 3, pp.
323−335.
HAMPL, M. (2005). Geografická organizace společnosti v České republice: Transformační
procesy a jejich obecný kontext. PřF UK, Praha, 147 p. + map appendix.
HORŇÁK, M., PŠENKA, T. (2013): Verejná doprava jako indikátor medzisídelných
väzieb medzi mestami Slovenska. Geografický časopis. Vol. 65, No. 2, pp. 119–140.
HORŇÁK, M., PŠENKA, T., KRIŽAN, F. (2013): The competitiveness of the longdistance
public transportation system in Slovakia. Moravian Geographical Reports. Vol.
21, No. 4, pp. 64–75.
HRABÁČEK, J. (2010). Dopravní odpor a možnosti jeho využití při posouzení modal
split (metoda RWTH Aachen). In: Seminář Telč 2010: Konkurence na evropských
železnicích − ekonomické, právní a regionální faktory. Recenzovaný sborník příspěvků.
ESF MU v Brně, pp. 120−135.
HUBERT, J. P., POTIER, F. (2003). What is known? In: Axhausen, K. W., Madre, J.L.,
Polak, J. W., Toint, Ph. (eds.): Capturing Long-Distance Travel. Research Studies
Press, Baldock, pp. 45–70.
IDOS (2001): Electronic timetable. CHAPS spol. s.r.o.
IDOS (2006): Electronic timetable. CHAPS spol. s.r.o.
IDOS (2011): Electronic timetable. CHAPS spol. s.r.o.
IDOS (2012): Electronic timetable. CHAPS spol. s.r.o.
JAIN, J., LYONS, G. (2008): The gift of travel time. Journal of Transport Geography,
16, 2, pp. 81–89.
Journey planner (2012). Retrieved April 12, 2012, from www.mapy.cz
KEELING, D. (2008). Transportation geography – new regional mobilities. Progress in
Human Geography, 32, 2, pp. 275–283.
KRAFT, S., VANČURA, M. (2009a): Geographical organization of the transport system
in Czechia and its development in the transformation period. Geografie, 114, 4, pp.
298−315.
Volume 15, Issue 1, 2015
67
KRAFT, S., VANČURA, M. (2009b). Dopravní systém České republiky: efektivita a
prostorové dopady. Národohospodářský obzor / Review of Economic Perspectives, 9,
No. 1, Masarykova univerzita v Brně, Brno, pp. 21–33.
KVIZDA, M. (2011). Vymezování relevantního trhu v odvětví železniční dopravy. In:
Kvizda, M., Tomeš, Z. (eds.): Regulace konkurenčního prostředí na železnici – teorie
v centru a praxe v regionech. Seminář Telč 2011 – sborník příspěvků. Ekonomickosprávní
fakulta, Masarykova univerzita v Brně, pp. 49–64.
KVIZDA, M., REDERER, V. (2012). Použití spotřebitelského šetření k vymezování
relevantního trhu v železniční dopravě – možnosti a problémy. In: Kvizda, M., Tomeš,
Z. (eds.): Konkurence na železnici – budoucnost pro 21. století nebo destrukce sítě?
Seminář Telč 2012 – sborník příspěvků. Ekonomicko-správní fakulta, Masarykova
univerzita v Brně, pp. 32–45.
KVIZDA, M., SEIDENGLANZ, D. (2014): Out of Prague: a week-long intermodal
shift from air to rail transport after Iceland’s Eyjafjallajökull erupted in 2010. Journal of
Transport Geography, 37, pp. 102–111. DOI:10.1016/j.jtrangeo.2014.04.014.
MARADA, M. (2006). Vertikální a horizontální dopravní poloha středisek osídlení
Česka. In: Kraft, S., Mičková, K., Rypl, J., Švec, P., Vančura, M.: Česká geografie
v evropském prostoru, elektronický sborník příspěvků (CD-ROM) z XXI. sjezdu České
geografické společnosti, katedra geografie, Pedagogická fakulta, Jihočeská univerzita
v Českých Budějovicích, pp. 169–174.
MARADA, M., KVĚTOŇ, V., MATTERN, T., ŠTYCH, P., HUDEČEK, T. (2014):
Accessibility patterns: Czech Republic Case Study. EUROPA XXI, Vol. 24, 2013, pp.
61–76.
MARADA, M., KVĚTOŇ, V., VONDRÁČKOVÁ, P. (2010). Doprava a geografická
organizace společnosti v Česku. Edice Geographica, ČGS, Praha, 165 p.
MARADA, M., KVĚTOŇ, V. (2010). Diferenciace nabídky dopravních příležitostí
v českých obcích a sociogeografických mikroregionech. Geografie, 115, 1, pp. 21–43.
MARCONI, D., SIMMA, A., GINDRAUX, M. (2004). The Swiss Microcensus 2005:
An International Comparison on Travel Behaviour. Paper for 4rd Swiss Transport Research
Conference, Ascona, March 2004. 19 p. Retrieved October 11, 2012, from
http://www.strc.ch/conferences/2004/Marconi_Simma_Gindraux_SwissMicrocensus20
05_STRC_2004.pdf
Ministry of Finance (2012). Prices of fuel and reimbursement of travel. Ministry of
Finance of the Czech Republic. Retrieved April 19, 2012, from
http://www.mfcr.cz/cs/legislativa/cestovni-nahrady
Ministry of Transport (2013a). Transport Yearbook of The Czech Republic (1998-
2012). Ministry of Transport of the Czech Republic. Retrieved October 15, 2013 from
https://www.sydos.cz/cs/rocenky.htm
Ministry of Transport (2013b). Transport Policy of the Czech Republic for 2014–2020
With a Prospect Until 2050. Ministry of Transport of the Czech Republic. Retrieved
REVIEW OF ECONOMIC PERSPECTIVES
68
October 20, 2013 from http://www.mdcr.cz/NR/rdonlyres/6771FC27-DCCC-4B72-
BD0E-3EF7E6118704/0/Dopravnipolitika20142020schvalena.pdf
MULÍČEK, O., OSMAN, R., SEIDENGLANZ, D. (2013): Imaginace a reprezentace
prostoru v každodenní zkušenosti. Sociologický časopis / Czech Sociological Review,
AV ČR, Sociologický ústav, Vol. 49, No. 5, pp. 781–810.
ORTÚZAR, J. D., WILLUMSEN, L. (2001). Modelling Transport, Third Edition. John
Wiley, Chichester, 499 p.
PAS, E. I. (1995). The Urban Transportation Planning Process. In: Hanson, S., ed.: The
Geography of Urban Transportation – second edition. The Guilford Press, London, pp.
53−77.
PIPKIN, J. S. (1995). Disaggregate Models of Travel Behaviour. In: Hanson, S., ed.:
The Geography of Urban Transportation – second edition. The Guilford Press, London,
pp. 188−218.
PRESTON, J., O´CONNOR, K. (2008). Revitalized transport geographies. In: Knowles,
R., eds.: Transport geographies: Mobilities, Flows and Spaces. Blackwell Publishing
Ltd., Oxford, pp. 227−237.
RIETVELD, P., BRUINSMA, F. (1998). Is Transport Infrastructure Effective? Transport
Infrastructure and Accessibility: Impacts on the Space Economy. Springer-Verlag,
Berlin – Heidelberg, 383 p.
RILEY, P. et al. (2010). Lokálně ověřený rámec pro multimodální modelování poptávky
po veřejné dopravě osob v souvislosti s interními a externími kvalitativními a
ekonomickými faktory dopravní obsluhy. Závěrečná zpráva z projektu VaV MD č.
CG721-045-190. Jacobs Consultancy, 162 p.
RODRIGUE, J.-P., COMTOIS, C., SLACK, B. (2006). The Geography of Transport
Systems. Routledge, London, 284 p.
SEIDENGLANZ, D. (2009). Konkurenceschopnost železniční a letecké dopravy. In:
Kvizda, M., Tomeš, Z. (eds.): Konkurenceschopnost a konkurence v železniční dopravě
– ekonomické, právní a regionální faktory konkurenceschopnosti železnice. Seminář
Telč 2009 – recenzovaný sborník příspěvků. Ekonomicko-správní fakulta, Masarykova
univerzita v Brně, pp. 72–86.
SHAW, J., KNOWLES, R., DOCHERTY, I. (2008). Introducing Transport Geographies.
In: Knowles, R., eds.: Transport geographies: Mobilities, Flows and Spaces.
Blackwell Publishing Ltd., Oxford, pp. 3−9.
SHELLER, M., URRY, J. (2006). The new mobilities paradigm. Environment and
Planning A, 38, Pion, pp. 207−226.
SHEPPARD, E. (1995). Modelling and Predicting Aggregate Flows. In: Hanson, S., ed.:
The Geography of Urban Transportation – second edition. The Guilford Press, London,
pp. 100−128.
SIMMA, A. (2003). History of the Swiss Travel Surveys, Swiss Federal Office for Spatial
Development, paper for 3rd Swiss Transport Research Conference, Ascona, March
Volume 15, Issue 1, 2015
69
2003. 23 p. Retrieved October 11, 2012, from
http://www.strc.ch/conferences/2003/simma.pdf
STRANDLING, S., ANABLE, J. (2008). Individual transport patterns. In: Knowles, R.,
eds.: Transport geographies: Mobilities, Flows and Spaces. Blackwell Publishing Ltd.,
Oxford, pp. 179−195.
TEMELOVÁ, J., NOVÁK, J., POSPÍŠILOVÁ, L., DVOŘÁKOVÁ, N. (2011).
Každodenní život, denní mobilita a adaptační strategie obyvatel v periferních lokalitách.
Sociologický časopis / Czech Sociological Review, Vol. 47, No 4. pp. 831–858
TSK (2013): Transport Yearbook – Prague (1988-2012). Technical Administration of
Roadways of the Capital of Prague (TSK). Retrieved October 15, 2013 from
http://www.tsk-praha.cz/wps/portal/doprava/web/pro-odborniky/rocenky
VACHTL, M. et al. (2013): Dopravní sektorové strategie 2. Fáze (souhrnný dokument).
Ministerstvo dopravy, Praha, 305 pp. Retrieved October 15, 2013 from
http://www.dopravnistrategie.cz/.
WIĘCKOWSKI, M., MICHNIAK, D., BEDNAREK-SZCEPAŃSKA, M., CHRENKA,
B., IRA, V., KOMORNICKI, T., ROSIK, P., STĘPNIAK, M., SZÉKELY, V.,
SLESZYŃSKI, P., ŚWIĄTEK, D., WIŚNIEWSKI, R. (2012): Poľsko-slovenské
pohraničie z hľadiska dopravnej dostupnosti a rozvoja cestovného ruchu. Varšava: Instytut
Geografii i Przestrzennego Zagospodarowania Polska Akademia Nauk; Bratislava
: Geografický ústav Slovenskej akadémie vied. 283 p.