Competition in Public Transport

Essays on competitive tendering and open-access competition in Sweden

ANDREAS VIGREN

Doctoral Thesis in Transport Science

Stockholm, Sweden 2017
Competition in Public Transport

TRITA-TSC-PHD 17-005
ISBN 978-91-88537-05-8

KTH Royal Institute of Technology

School of Architecture and the Built Environment
Department of Transport Science
SE-100 44 Stockholm
SWEDEN

Akademisk avhandling som med tillstånd av Kungl Tekniska högskolan framlägges till offentlig
granskning för avläggande av teknologie doktorsexamen i transportvetenskap fredagen den 8
december 2017 klockan 14:00 i Kollegiesalen, Brinellvägen 8, Kungl Tekniska högskolan, Stockholm.

© Andreas Vigren, December 2017

Supervisors:
Assoc. Prof. Svante Mandell, KTH
Prof. Jan-Eric Nilsson, VTI

Printed by Universitetsservice US-AB

Till Docksta:
“Det finns inga bojor eller band
som kan hindra mina tankar att färdas norrut
på våren, som en flygande and.

Det är här man känner vart stigarna går,

och man vet vem som är släkt med vem.
Det är här man vet var gäddorna står,
ja det är hit man kommer när man kommer hem”

Det är hit man kommer när man kommer hem – Euskefeurat.

Text av Ronny Eriksson.
 Abstract

The competitive environment in Swedish public transport has undergone major changes in recent
decades, going from a situation with private sector monopolies to one where competitive tendering
and open access competition is the rule. However, irrespective of the regime, proper analyses and
evaluations are required to determine how the market is functioning and what measures should
be taken to improve it. This thesis analyzes how contracts for tendered bus services can be better
designed for higher cost efficiency and greater competition, and it assesses the effects on ticket
prices following the first substantial entry in the Swedish open access railway market.
For such analysis and evaluation, data availability and data quality are essential. However,
the data situation in the Swedish public transport sector is poor, especially the publicly available
data, as is the case for most of Europe. For this reason, new, and in many respects unique,
data were gathered for use in this thesis on tendered contracts in bus services and the prices of
operators on the commercial railway market. These data made it possible to analyze tendered bus
contracts and open access competition in Sweden in more detail than has been done previously.
The high level of detail in these reports provides new and useful information and allows for better
advice to be given to policy makers.
The results of this work show that the cost efficiency of tendered bus services is similar
across all Swedish counties, except for the more high-density counties where efficiency is lower.
Considerably lower efficiency is also found for contracts with services run in-house by the Public
Transport Authority (PTA), compared to when the same service is run by a private actor. With
respect to the competitive environment, it was found that many contract design factors have little
or no effect on the number of bids that the PTA sees in their tenders. No measure that could be
imposed by a single PTA was found to increase the total number of bidders by more than 0.5
bidders. However, the results suggest that PTAs as a collective could try to avoid tendering too

iv
many contracts at the same time because this was shown to reduce participation by up to about
two bidders. In addition, these studies show that the local competitive environment is important
for the PTAs to consider. The way in which contract areas are defined will also affect the
participation rate as operators were found to participate in tenders to a lower extent the farther
their workplaces are from the contract area. While larger operators appear to be less sensitive
with respect to such distances, the fact that smaller operators are, and that they often bid as one
unit as members of cooperation companies, makes the competitive environment important. The
results suggest that depots could be included in the contract to stimulate participation, but this
is by no means the only nor an easy solution.
As a further investigation into competition in Swedish public transport markets, this thesis
has analyzed the entry made in 2015 by MTR Express (MTR) on the Stockholm-Gothenburg
railway line. The overall conclusion is that customers are indeed facing lower prices one and a
half years after the entry. MTR’s prices are on average 100 SEK lower than the incumbent SJ’s
prices. Furthermore, the analysis shows that the incumbent’s prices have also gone down, by
almost 13 percent, following the entry. It is important to note that these are only the results of
the short-run effects of the entry, and the more long-term development is yet to be determined.

v
 Sammanfattning

Marknadsmiljön i svensk kollektivtrafik har förändrats över de senaste decennierna. Den har gått
från en situation med privata monopol till en där konkurrensutsatt upphandling och open ac-
cesskonkurrens råder. Oavsett vilken reglering som gäller krävs dock bra analyser och utvärderingar
på hur transportmarknaden fungerar, samt med vilka medel som bör användas för att förbät-
tra densamma. Denna avhandling analyserar hur kontrakt för upphandlad kollektivtrafik kan
förbättras för att uppnå högre kostnadseffektivitet och bättre konkurrens, och vilka effekter på bil-
jettpriser det första stora inträdet på den numera konkurrensutsatta svenska järnvägsmarknaden
haft.
För analys och utvärdering är tillgången och kvalitén på data grundläggande. Denna situation
är dock bristfällig i både svensk och europeisk kollektivtrafik. Därför har ny, och i många
avseenden unik, data samlats in för användning på upphandlade busskontrakt och biljettpriser
hos tågoperatörer. Den insamlade datan har gjort det möjligt att analysera dessa områden i mer
detalj än vad som gjorts tidigare. Även om det finns tidigare studier i dessa ämnen är litteraturen
på transportområdet inte så rik som den skulle kunna vara med bättre och mer tillgänglig data.
De fyra artiklarna som ingår i avhandlingen ger därför bidrag till forskningen genom att i mer
detalj undersöka kontraktsutformning och konkurrens i buss- och järnvägsmarknaden, vilket också
ger bättre rekommendationer till beslutsfattare.
Resultaten visar att kostnadseffektiviteten i upphandlade busskontrakt inte skiljer sig nämn-
värt mellan svenska län och regioner, utom i mer tätbefolkade områden där effektiviteten är
lägre. Betydligt lägre kostnadseffektivitet har påvisats också i områden där de regionala kollektiv-
trafikmyndigheterna (RKM) kör trafiken i egen regi. Vad gäller konkurrens har det visats att
många kontraktsrelaterade faktorer har begränsad eller ingen effekt på antalet bud RKM får i sina
upphandlingar. Inga kontraktsfaktorer kunde ensamt påverka antalet budgivare med mer än runt

vi
0.5 bud. RKM som ett kollektiv skulle däremot kunna undvika att upphandla för många kontrakt
samtidigt då detta skulle kunna minska deltagandet med upp till runt två budgivare. Slutligen
visar resultaten att den lokala konkurrensmiljön är viktig att ta i beaktande för RKM. Hur RKM
utformar kontraktsområden kan påverka utfallet i budgivningen då operatörer lägger anbud i
lägre utsträckning ju längre ifrån ett kontrakt de har arbetsplatser. Medan större operatörer är
mindre känsliga vad gäller avstånd är detta inte fallet för mindre operatörer, som inte sällan går
ihop i samverkansbolag och budar som en enhet. Att erbjuda depåer för användning i avtalet
skulle kunna deltagandet, men det är inte en okomplicerad och enda lösning på problemet.
I avhandlingen analyseras också MTR Express (MTR) inträde på den kommersiella järn-
vägsmarknaden. Inträdet gjordes i mars 2015 och utmanade den redan etablerade aktören SJ:s
monopol på sträckan. Den övergripande slutsatsen är att resenärerna möter lägre priser ett halvår
efter inträdet då MTR:s priser är i genomsnitt 100 kronor lägre än SJ:s. Vidare visar resultaten
att SJ:s priser har minskat med nästan 13 procent sedan inträdet. Det är också viktigt att
påpeka att detta är kortsiktiga effekter av den ökade konkurrensen på sträckan. Den långsiktiga
utvecklingen behöver också analyseras.

vii
 Acknowledgements

First, and foremost, I want to thank my supervisor Svante Mandell for his constant support and
help with my work on this thesis. You have given me the opportunity to form the thesis with my
own ideas and approaches, while still pointing me in the right directions. I am also grateful for
you giving me such responsibility for my work. I could not have wished for a better supervisor.
Again, thank you.
I want also to thank my co-supervisor Jan-Eric Nilsson. Your knowledge of the railway and
public transport sector has been of great help in my writing, as well as your wise comments from
reading my work.
During my four years at VTI and TEK, I have come to realize that the working environment
is essential for doing good work. The newspaper Dagens Nyheter once had an online poll asking,
“How often do you laugh in your workplace?”, with the options 0, 1, 2, 3, and 4 times a day. This
poll caused a laugh at the fika because four laughs are nowhere near what we do in one day.
There is a mutual understanding and respect for one another here that I value highly. Everyone
makes a contribution to this, and everyone has therefore contributed to the joy that I have had
in coming to work every day. There are, however, some people I want to acknowledge specifically.
Hanna and Magnus, my partners in crime (and beers). Roger, for all the talks and initiated
discussions about my papers. Also, Anna, Gunilla, Inge, Johan, Mattias, and Samuel. Regarding
people outside VTI, I first want to thank Jonas and Maria (well, Maria is actually with us at
VTI now). Without you, I would not have ended up here. Thanks for seeing potential in a guy
who could not draw a proper illustration of a Pigovian tax. Also thank you to Johan Holmgren
for being the opponent at my licentiate seminar, to Sofia Lundberg for your important comments
and remarks on this thesis, Anders Karlström, and Maria Melkersson. To Ulrika, Per, Nadia, and

viii
Susanne for helping me deal with all the administrative tasks. And finally, to the Department of
Transport Science for having me over these four years, and the department of Real Estate and
Construction Management.
There is a life outside work. My various constellations of friends have been of, at least, as
much importance as my colleagues. Thanks to DD and Kent, both for our time in Umeå and
here in Stockholm. You are important to me. To my dear friends in Docksta (with vicinity):
Henke, Johanna, and Julia. Further, I cannot count the number of times we have been at Jocke’s
place on Fridays, drinking wine until five in the morning. I really enjoy these moments. To Vivi,
Bibbi, Mona, Bodil, Harriet McClain, Nicklas, Patric, Kevin, Sarah, Hassan, Rebecca, and Emmy:
LBAB says thank you, and that she loves you. To Christoffer, Johan and Johanna (and now also
Ingrid), Johnny, Niclas and Jannika, Nuno, Ronny, Simon and Pelle. To Johan Ekman, Tomas
Byberg, and Mattias for fostering my transport interest. And again, to Nicklas.
Last, and most importantly, I am most grateful to my family, whom I love incredibly much.
Thanks for your immense support and interest. To mum, dad, Mats, and mormor. Lisa, I have
counted you here for a long time also. Becka, my sister, even if I don’t show or tell you this often
enough, you are the most important person in my life. Finally, although you are not with us in
life anymore, thank you morfar, farmor, and farfar. I love you all.

Andreas
Stockhom, November 2017

ix
 List of papers

Paper I
Vigren, A (2016). Cost Efficiency in Swedish Public Transport. Research in Transportation
Economics, 59, 123-132.

Paper II
Vigren, A. (2017). How Many Want to Drive the Bus? Analyzing number of bids for public
transport bus contracts.

Paper III
Vigren, A. (2017). The Distance Factor in Bus Contracts: How far are operators willing to go?

Paper IV
Vigren, A. (2017). Competition in Swedish Passenger Railway: Entry in an open access market
and its effect on prices. Accepted for publication in Economics of Transportation.

x
 Table of Contents

Abstract iv

Sammanfattning vi

Acknowledgements viii

List of papers x

Cover Essay
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Market outcomes and regulatory environments . . . . . . . . . . . . . . . . . . . 3
3 Data availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4 Contributions and summary of papers . . . . . . . . . . . . . . . . . . . . . . . . 10
5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Paper I. Cost Efficiency in Swedish Public Transport

Paper II. How Many Want to Drive the Bus?

Paper III. The Distance Factor in Bus Contracts

Paper IV. Competition in Swedish Passenger Railway

 Cover Essay

“ ’Begin at the beginning,’ the King said, very gravely,

’and go on till you come to the end: then stop.’ ”

The King of Hearts in Alice’s Adventures in Wonderland.

Written by Lewis Carroll.
1 Introduction

The organization of transport markets is a constantly developing matter, not only in how the
markets and companies function, but also the extent to which regulation is needed, how and
where the public should intervene, and the extent of any interventions. Although political bias is
a factor in this that cannot be neglected when reforms are carried out (Vickerman, 2015), it is
arguable that the primary goal of any such regulations and interventions is to achieve the best
possible welfare outcome. Whether this is social welfare, corporate welfare, or the welfare of
certain individual groups can, of course, vary, but still the changes are intended to improve some
situation. In this thesis, the term welfare will refer mainly to social welfare.
Traditionally, many transport markets have been considered to require relatively strong
regulations by the ruling politicians, with little freedom for the operators, in order to deliver
high welfare. These regulations have lately come to be relaxed, and the Swedish regional bus
transport1 market might serve as an example of such a liberalization regime. While private
firms had exclusive licenses to run specific routes and had little direct competition in the early
20th century, the industry became more heavily regulated in the 1960s as profits decreased
following the rise in private car ownership (Alexandersson, 2010). Further on, the trend became
one of more expropriation of operators, more subsidies, and greater coordination of fares and
schedules in the 1970s, which was the time when the Public Transport Authorities (PTAs) were
established (Alexandersson, 2010). A decade later, competitive tendering was introduced, which
was a contract-awarding method that came to dominate the public transport bus provision. This
situation implies more private supply than in the 70s and early 80s, and it is currently the
standard way of procuring public transport services in Sweden. Today, over 90 percent of the total
supply is tendered out. Legislatively, regulations were further relaxed by the Public Transport
Act of 2012 (SFS 2010:1065), which allowed for open competition alongside subsidized traffic
within the PTAs’ jurisdictions.
1 The term “public transport” refers generally to “passenger transport services of general economic interest

provided to the public on a non-discriminatory and continuous basis”, as phrased by the European Union’s
regulation on the same (EU 1370/2007, Article 2 (a)). That is, short and long-distance bus, rail, boat, air, and
other services fulfilling these criteria.

1
 The regulatory development of the bus industry in Sweden closely follows the regulatory cycle
as formulated by Gwilliam (2008), a concept also previously discussed by Diandas and Roth (1995)
and Preston (2001). Gwilliam’s (2008) cycle includes four stages for the regulatory environment
of bus services - private sector area monopoly, regulated private local monopoly, regulated public
monopoly, and competitive private supply. This is a similar regulatory path that Sweden has
taken in other transport areas than bus traffic, such as the railway sector where supply is now
also provided through competitive procurement as well as commercial transport services by both
air, bus, and rail (Alexandersson, 2010). The tradition is today, arguably, a more liberal regime
with more market initiatives. Today, Sweden sees competition both for the market, in terms of
competitive tendering, and in the market, in terms of commercial/open-access initiatives.
Liberalization and, by extension, competition has the potential to foster innovation, which in
turn can provide new market concepts and achieve ways of producing services more efficiently. An
example is the US airline market for which the overall impression is that the deregulation in 1978
was a success in terms of lower fares as well as increased supply and efficiency (Goetz and Vowles,
2009). However, the authors of that study also argue that the deregulation has caused carriers to
underperform financially because the market is naturally oligopolistic and that less attractive
routes have not seen the same benefits from the competition in the market. The latter seems to
be the current situation in Sweden following the opening of the railway market where the only
major entry has been on the high-demand Stockholm-Gothenburg route (Fröidh and Nelldal,
2015). The entry was in March 2015 when the Hong Kong-based operator MTR Express (MTR)
started its services and thus challenged the (railway) monopoly of the state-owned operator SJ.
There are sometimes clear motivations for public intervention, for example, with the Mohring
effect (Mohring, 1972; Jansson, 1979) and when coping with non-internalized externalities of
transport, in which subsidies are often used to achieve a greater welfare. Because subsidies are an
intervention by the public sector and have the potential to distort the private market (Graells,
2015), a transparent and non-discriminatory purchasing process and awarding mechanism is
required. One of the more common in the EU is procurement through competitive tendering.
Procurement in turn also requires competition to function adequately as companies compete for
the market. Insufficient competition can cause similar market failures as in an ill-functioning free
market - including monopoly power, lower service quality, and inefficiencies - and this indicates

2
that market surveillance and policy advice is necessary both here and in a more deregulated
market. Too little information about the market structure, bid outcomes, and the impact of
contract design risks a poorer welfare outcome.
The work in this thesis is entirely shaped by the dominant regime for public transport today
with competitive tendering and open access, and it focuses on the competition in and for the
market. The research questions and purposes given in the appended papers address various
aspects of the competitive environment and performance of the tendered public transport bus
services and the long-distance rail market. These are summarized in Section 4. The overall
objectives are to investigate the competitive environment in the Swedish public transport sector.
In particular, the objectives are to assess the impacts of the recent market opening in the rail
market, the efficiency in the tendered bus market, and the competition situation in both markets.
The main contribution of the thesis is the gathering and analysis of new and unique data for
the Swedish situation, which in many respects carries over to an international context. This
allows the objectives to be addressed on a more disaggregate level than before and thus for more
detailed policy advice to be given.

2 Market outcomes and regulatory environments

As noted previously, the overall aim for regulatory interventions, whether it is imposing or relaxing
regulations, is arguably to enhance the welfare outcome. In order to formalize this way of thinking,
a “regulatory curve” as suggested by van de Velde and Preston (2013) is useful. The curve is
illustrated in Figure 1, and was based on Figure 2 in van de Velde and Preston (2013). The
basic idea is that there exists some “optimal” amount and type of prescription to the market
so that the best outcome (welfare) is achieved. This prescription could be different types of
regulation, industry standards, or contract design elements. Interventions in the market can then
be benchmarked to this curve to assess where one stands today and where one is heading. This
sort of benchmarking can be made in terms of the amount of prescription, but more importantly
in terms of the outcome. An important point is also that going from a very regulated market,
often with many prescriptions, to a more deregulated market does not necessarily mean a better
outcome (higher welfare). One might just arrive at the same outcome level on the curve, but on the
other side of the optimal amount of prescription. It is important to note that interventions could

3
 Outcome
(welfare)

Too little Right amount Too much

prescriptions of prescripitons prescriptions

Figure 1: The regulatory curve re-illustrated from Figure 2 in van de Velde and Preston (2013).

in fact have counteracting effects, for example, the decision by politicians to run public transport
services in-house, or give different outcomes in different counties as shown by Nash et al. (2013).
While studies have shown that in-house operations are generally less cost efficient in delivering
output (Roy and Yvrande-Billon, 2007; Ottoz et al., 2009), thus getting less traffic for the money,
such a regime could also deliver other values such as higher customer satisfaction (Fiorio et al.,
2013) that might affect welfare positively.
The outcome is not only shaped in the longer run by legislative regulatory changes, but also in
the shorter run by how contract design factors are used. In the case of bus tenders, contract design
factors can be viewed as a form of regulation or prescription imposed by the PTAs on the operators.
At the same time, this contract design is probably prescribed in order to achieve better outcomes
for the public transport service. While these factors can be an artifact of legislation, for example,
requiring bus operators to equip buses running in a contract with seat belts, many factors imposed
voluntarily by the PTAs can affect the outcome, not seldom in unknown directions if ex-post
information is uncertain. This issue relates to the theory on incomplete contracts (Hart and
Moore, 1988), and the observation that auctions (like competitive tendering) do not necessarily
perform well with complex or incomplete contract designs (Bajari et al., 2008). Services provided
by contracted actors can be further problematic when non-contractible cost reductions have

4
serious negative effects on non-penalized quality factors (Hart et al., 1997). In addition, concepts
such as information asymmetries between the PTA and the operator, and between operators,
could well affect the outcome in terms of cost or service provision.
Because the outcome of a regulatory environment or a contract is affected by its design, the
design is important to evaluate and study both by the authority and by researchers. Policy
implications for public transport can certainly be drawn from studies in other sectors, and
outcomes to some extent can be inferred. For example, the study by Lundberg et al. (2015)
concluded that the inclusion of green criteria has limited effectiveness as an environmental policy
measure when studying tenders of Swedish internal regular cleaning contracts. Or that favoring
targeted companies (for example, small or local companies) reduces efficiency and revenue as Athey
et al. (2013) show using timber sales data. Policy makers in transport can indeed learn from
this. Still, while policy advice can be inferred from other sectors, evaluation in the relevant sector
is still necessary. In the end, research and evaluation is needed to know in which direction the
transport market is moving on the outcome axis, and what interventions might be necessary to
achieve better outcomes in the future. For such empirical studies, transparency is required and
data need to be available and of good quality.

3 Data availability

Statistical methodology and computing capacity are hardly the main limitations for performing
policy evaluation and research today. Rather, it is the availability of data, which varies depending
on country and transport mode. van de Velde (2015) notes that the data situation for public
transport in most countries is generally poor in terms of availability, extent, and quality. This in
turn hinders good analysis.
While quantitative analysis has indeed been conducted both on contractual variables and
efficiency in subsidized tendered public transport, it is clear that data availability depends on the
intended research. Generally, in many countries there seems to be good availability of operator-
related data such as company expenses and to some extent their kilometers produced. These
operator data are often used in econometric studies on technical efficiency (Farsi et al., 2005; Roy
and Yvrande-Billon, 2007; Obeng, 2013; Jarboui et al., 2014) and cost efficiency (Filippini and
Prioni, 1994; Farsi et al., 2006; Walter, 2011; Sakano and Obeng, 2012; Jarboui et al., 2013),

5
all of which almost exclusively use operator data. These data are to a large extent available
through published surveys and public statistics in Germany, France, Italy, and the US. Dalen and
Gómez-Lobo (2003) and Piacenza (2006) compiled operator data themselves in order to analyze
competition and ownership efficiencies. The number of studies including more contract-related
variables is smaller, but there are a few examples on contract data for Germany, Italy, and the
Netherlands (Amaral et al., 2013; Augustin and Walter, 2010; Beck and Walter, 2013; Veeneman
et al., 2014; Avenali et al., 2016, 2017), and on both contract and award data for London (Cantillon
and Pesendorfer, 2006; Iossa and Waterson, 2017). Common for most of these studies is, however,
that the data is limited in scope; the data is mostly collected through a survey and/or only for
some variables, which means that not all information on the tender is necessarily available. This
limitation is not unsurprising, however, because the collection of such data is time consuming
and difficult work.
For the commercial markets in most of Europe, including Sweden, data is arguably even
scarcer than for tendered public transport and are often next to non-existent. This scarcity
includes both information on ticket prices (fares) and numbers of passengers. However, at least for
Sweden, there is detailed information regarding supply in terms of timetable data on departures
and arrivals that is available from Samtrafiken’s GTFS data (General Transit Feed Specification).
Compared to the case with subsidized transport, this lack of detailed information is perhaps not
very surprising because such information is potentially sensitive corporate information. Still, the
data is needed in order to monitor the market and, perhaps more importantly, to evaluate policy
reforms. The liberalization of passenger railway services in Europe has meant more open access
competition in some countries, but empirical evaluations on the impacts and welfare effects of
the same are limited. Some literature simulating competition exists and indicates positive social
welfare effects of increased competition (Preston et al., 1999; Preston, 2008; Preston and Robins,
2013; Johnson and Nash, 2012). The applied literature looking at the effects of price competition
is limited, and literature on the effects on demand are arguably non-existent on more disaggregate
levels. Because train operators sell their tickets online, some recent papers have used web crawlers
to obtain price data for analyses using both descriptive (Tomeš et al., 2014, 2016) and regression
techniques (Bergantino et al., 2015; Beria et al., 2016).

6
 300

Days to complete documentation 250

200

150

100

50

0
0 5 10 15 20
Number of respondents

Figure 2: Days until complete documentation

For this thesis, two main data sources were collected and compiled - one with contract
information on bus tenders, and one with ticket prices of rail operators. In relation to the previous
data used in the two transport markets, these new datasets are unique both in their extent and
their level of detail. These data have also allowed for the more detailed investigation of the
Swedish public transport market than has previously been possible.
The contract data contain detailed information on bus tenders carried out by the Swedish
PTAs, both with regard to the contract setup and the tender awarding. These data were used in
Papers II and III, and a more detailed description is found in Paper II. The collected material is
essentially the enquiry documentation and the awarding protocol of all tenders for bus contracts
that were active in Sweden in December 2015. That is, the ex-ante information from the tenders.
Because the enquiry documents were collected, all information about the setup of the contract and
the traffic to be run is available, which allowed for the specification of a wide range of variables.
Similar efforts for Sweden were made in Nilsson (2011) and Hultén (2015), but those attempts fell
short, and obstacles like the ones encountered in this paper were met. One of the major obstacles,
and perhaps the most serious, was poor compliance with the principle of public access to official

7
documents2 . Because enquiry documents and awarding protocols are public information, and
because PTAs are subject to the principle of public access, requests to access this material must
be administered in a timely manner (skyndsamt), and refusal to provide access to the information
must be motivated by reference to applicable paragraphs in national legislation (SFS 1949:105,
Chapter 2). In attempting to collect the contract data for this thesis, requests were made to the
PTAs in order to acquire enquiry documents and awarding protocols of previous tenders. Many
PTAs simply did not respond to the request(s), claimed the information was subject to corporate
secrecy (mostly with regard to placed bids), or claimed the information was lost. Despite this,
the tender documentation was eventually acquired. Figure 2 illustrates another obstacle, that
many of the 20 asked PTAs failed to administer the requests in a timely manner. The solid line
shows the accumulated number of PTAs relative to the number of days before they delivered
requested documentation, and the dotted line shows the 14-day mark. In most cases, 14 days
should be considered a “timely manner”, at least for these requests, although some leeway must,
of course, be given depending on the size of the request. Only eight PTAs returned the complete
documentation in two weeks’ time. What is even more worrying is that five PTAs took 100 days
or more to hand over the information, even after repeated requests, which should be in clear
violation of the principle of public access. This is a prime example of why data and publications
on contract design are relatively scarce; few researchers have the time or energy required to follow
up on the data requests.
The second data set compiled is on ticket prices for commercial passenger railway. Like the
previous studies on price changes following the onset of competition in Europe, a web-crawler
approach was used for this thesis, and a Python program was written for this purpose. This
program requested price information from the websites for departures of the Swedish state-owned
incumbent railway operator SJ and the entry operator MTR between July 2014 and June 2016.
More details on the price collection and the content of the data is available in Paper IV. The data
is unique in the passenger railway research field both in its richness (including price information on
most departures up to 31 days before departure) and that it include six separate lines, of which five
did not see entry during the sample period. However, perhaps most importantly, it is especially
2 More information on the principle of public access to official documents can be found at

http://www.government.se/how-sweden-is-governed/the-principle-of-public-access-to-official-documents/

8
rich because it spans both the pre- and post-entry periods of MTR on the Stockholm-Gothenburg
line. This allowed for a before-and-after analysis of the (short-run) effects of the entry on price.
The dataset holds a total of 2.15 million observations.
As already noted, market data on commercial transport operations in Europe are scarce.
There are, however, some sources that are worth mentioning and that can serve as role models
for improving the data situation. van de Velde (2015) notes good examples in the regional public
transport data collected by the Department for Transport in Great Britain and by the US
Department of Transportation. Studies mentioned earlier also suggest that data is available
on operators, but not contracts, in both Germany and Switzerland. For railway services, the
United Kingdom is again a good example in also providing (posted) fare data for the integrated
ticketing and fares system of franchised train operators (Rail Delivery Group, 2017). However,
the perhaps most extensive data source with both fare and passenger data is the Airline Origin
and Destination survey (DB1B) provided by the Bureau of Transportation Statistics at the US
Department of Transportation. The survey conducted quarterly since 1993 is a 10 percent sample
of tickets sold by all airlines in the US and contains itinerary details such as route flown, fare,
and fare class. Because of its richness both in variables and time, it has allowed for extensive
analyses of policies (Goetz and Vowles, 2009; Ciliberto and Williams, 2010; Aguirregabiria and
Ho, 2012), and competition in the airline market (Goolsbee and Syverson, 2008; Brueckner et al.,
2013; Luo, 2014), and has been important for methodology development. This is arguably an
example where the public accessibility of data can benefit both researchers and policy makers.
In Paper I, a third data source was used - the contract collection (“Avtalsinsamlingen”).
Transport Analysis sent out questionnaires to each of the 21 PTAs in Sweden, asking for
information and data on their running contracts. These data contain information on contract
payments the PTAs have made in their respective tendered contracts as well as some variables
related to contract design. The main difference compared to the enquiry document data used
in Papers II and III is that these are ex-post data. The data currently span 2010-2015, but the
quality is questionable for the years 2010-2012. This type of data is arguably more useful for
estimations of performance than the contract data with the bids. The reason for this is that bids
placed for tenders are merely estimates of what the traffic will cost and do not necessarily reveal,
for example, inefficiencies. At the time of writing this thesis, however, it seems that Transport

9
Analysis is no longer going to collect such data. In addition, PTAs are obliged to make available
similar data, in accordance with EU-regulation (EU 1370/2007), but few PTAs comply with
this (Transport Analysis, 2015).
A data source that was utilized in Papers II, III, and IV, and that was already available, was
Samtrafiken’s GTFS data, which essentially is the timetables for public transport lines in Swedish
public transport. All operators except airlines, both private and public, must report their supply
to this database, which is available for free through trafiklab.se and Samtrafiken. The GTFS
data also contain geographical information on all stops. This is what makes it possible to map
the contract data (containing information on which bus lines are included in each contract) to a
geographic dimension in Papers II and III. By doing this, more data sources could be used along
with the contract data, for example, the operator workplace data from Statistics Sweden used in
or population density data.

4 Contributions and summary of papers

The common denominator for all of the included papers is that they all deal, in one way or
another, with competition in the public transport sector. While Papers I, II, and III study the
tendering market of subsidized regional public bus services, the fourth covers on-track competition
in the passenger railway market. That is, the first three are about various aspects of competition
for the market, while the fourth is about competition in the market. The order of the papers
in this thesis has been made with respect to what would be the natural order of reading. In a
chronological order of writing, however, paper IV lies in between papers I and II.
The first contribution lies in having acquired and used new and relevant contract tender and
price data in public transport for economic analysis. As discussed in the previous sections, this
information is scarce both in Sweden and internationally. The collected data is rich in scope,
allowing for more detailed research than previous studies. Given the opportunity to work with
more detailed data, more knowledge has been gained on the issues of public transport tendering
and open access competition in railway markets.

10
Paper I. Cost Efficiency in Swedish Public Transport

Sweden has been procuring public transport services for almost 30 years, and both the PTAs and
operators have gained much experience in the process during this time. The costs for providing
the services have, however, increased substantially and more rapidly over the last decade than
the increase in supply and in the number of trips (Transport Analysis, 2017). There is also little
knowledge about what has caused this, or of what factors affect cost efficiency in general. Paper
I studies the latter and looks at the impact on cost efficiency of contractual and operational
environment factors for Swedish bus contracts. It also compares this cost efficiency across PTAs
(counties).
To do this, the paper employed a stochastic frontier methodology, which aimed at estimating
a cost frontier and attributed some of the deviations from the frontier as inefficiencies. The
method is well established both in the public transport (Farsi et al., 2006; Holmgren, 2013) and
other fields (Kumbhakar et al., 2015) as a way of estimating inefficiencies across units and of
identifying factors that affect the level of efficiency. The data used in the analysis were the
contract collection data (“Avtalsinsamlingen”) provided by Transport Analysis, as described
above. This dataset contains ex-post information mainly on production and costs in the PTAs’
respective tendered contracts, but also some information on the contract design and the operators.
As argued previously, ex-post data is assumed to be more appropriate for analyzing performance
than are ex-ante data. The dataset covers all 21 PTAs for the year 2013 and includes 280 gross
cost and incentives contracts. Also, data on operators’ labor and capital costs were mapped with
that data from the operators’ respective annual reports and from the Swedish vehicle registry.
The main results from the estimations show that a higher population density in the contract
area reduces cost efficiency. Similarly, traffic operated by an in-house operator (that is, the traffic
is not put out for tender) is also less efficient. Both results are in line with the previous literature
on the subject (Walter, 2011; Piacenza, 2006; Ottoz et al., 2009). The rest of the included contract
variables (contract term, the contract being operated by a large operator, and the use of incentive
payments) have no statistically significant effect on performance. The last result, that including
incentive payments has no statistically significant effect on cost efficiency compared to gross-cost
contracts, is positive for the use of the contract form, but the causal effect that incentives have
on attracting more passengers must be studied in more detail.

11
 The cost efficiency scores across PTAs do not differ more than marginally, and the difference
between the highest and 18th highest score is merely six percentage points. This implies fairly
similar efficiency across Sweden. The two counties of Skåne and Stockholm, however, see slightly
lower estimates than the others. These are also the densest counties. Finally, the county of
Västmanland has a substantially lower efficiency score than the country average, a potential
explanation being that they run all traffic with their in-house operator.

Paper II. How Many Want to Drive the Bus?

While paper I focuses on performance of the running public transport services, the tender design
and competition for the contracts is considered in Paper II. One relevant measure of competition
is the number of bidders a PTA sees in their tenders, which varies substantially both across and
within PTAs. This motivated the study in Paper II, which sought to determine how factors
relating to contract design and the operational environment affect the number of bids for tendered
contracts. Because of the lack of data in the Swedish public transport sectors, studies on
competition are scarce as well. Internationally, previous studies are also few, or use more limited
data than were used in this paper (Augustin and Walter, 2010; Beck, 2011; Amaral et al., 2013).
The analysis used the contract data described above that were collected from the individual
PTAs, and the number of bidders could be determined from the detailed data on how 268 contracts
tendered between 2007-2015 were constructed. Further, along with GTFS data and workplace
data from Statistics Sweden, the transport supply and local competitive environment could also
be inferred and be included as suitable variables in the analysis. Because the contracts state what
bus lines are included, the geographical line information from the GTFS data could be used to
construct the contract areas. Because the outcome (the number of bids) is an integer variable, a
Poisson regression model was appropriate. Furthermore, because none of the contracts see zero
bids and the dependent variable is underdispersed, a zero-truncated generalized Poisson model
was used for the analysis (Consul, 1989; Winkelmann, 2013).
The overall result is that most factors included in the analysis change the number of bidders
only by some 0.1-0.5 bidders. This means that the single PTA has few direct measures to use
when attempting to design their contracts to increase the number of bidders substantially. The
largest impact found was on how many other tenders are put out at the same time as the PTA’s

12
specific tender. That is, the number of simultaneous tenders, which has a clearly negative effect
and can reduce the number of bidders by almost two bidders. This would mean that PTAs should
organize their tenders so that the asking time does not overlap for too many tenders, nor should
they tender too many large contracts simultaneously. The local competitive environment in terms
of the number of potential bidders on the contract was found to affect the outcome positively,
which means competition is not only a national-level matter.
Three variables indicating if the PTA added special requirements in how the operator conduct
the traffic was added to the model, and all estimates indicate that all three variables have a
negative effect: fuel requirements, forcing the operator to use a specific depot, or imposing
overtakes of existing vehicles on the winning operator. Although the effect of the individual
requirements was estimated, the result should be interpreted more generally; that including
special requirements and thereby restricting operator freedom lowers the number of bidders.
This is also something the industry has worked on for some time, constructing industry-wide
recommendations for contract design and the use of special requirements in the organization
“Partnersamverkan för Förbättrad Kollektivtrafik”.
Finally, the distinction between forcing a depot on an operator and only giving the operator
the option to use it was also found to matter for the result. While the former decreases the
number of bidders, the latter does not. This could be a consequence of small operators often
having their own depot in the local area and thus not participating in the tender if they would be
forced to use another depot for the contracted traffic and thus having to take on double costs for
such facilities.

Paper II received an honorable mention at the Thredbo 15 conference in Stockholm in August

2017.

Paper III. The Distance Factor in Bus Contracts

Building on a result from Paper II, that the local competitive environment has an impact on the
number of bids per contract, Paper III sought to investigate this finding in more detail. Previous
studies have found a negative relationship between a company’s distance from the contract area
and its propensity to participate in the bidding process in general (Jofre-Bonet and Pesendorfer,
2003; Bajari and Ye, 2003), and specifically in bus tenders (Cantillon and Pesendorfer, 2006).

13
The magnitude and importance of the distance factor has, however, not been investigated further.
The issue is that of dead running, which is the distance the operator has to run its buses between
the depot and line. They are not paid for this distance, and therefore want to reduce it as much
as possible (Preston, 2008; Hensher et al., 2013) The purpose of this paper was therefore to
investigate the impact of operators’ distance from contracts on their probability of participating
in tenders, consider differences in this probability across operator types, as well as what effect
there might be of including depots in the contract.
Using data from Statistics Sweden with information on all Swedish operators’ workplaces (for
example, a depot) for the years 2007-2015, each operator’s workplace was identified for each year,
and the corresponding distances from the contract were calculated using GIS methods. Thus,
the observational unit was at the operator-workplace level. The contract area was constructed
using GTFS supply data, which give coordinates for all stops along the lines of the contract.
Line information was available in the contract data used for Paper II. Because an operator can
have multiple workplaces, the distance variable was calculated as the median distance from an
operator’s closest workplace to each line in the tendered contract in question. This is intended to
represent the dead running for the operator were it to win the contract. Control variables for
factors relating to the contract and operating environment were also included using the contract
data. Using a probit regression model, the outcome variable takes the value 1 when an operator
places a bid in a tender, and 0 otherwise. Two variables are in focus of the paper; the operator’s
own (median) distance to the contract, and the second closest (to the contract) operator’s distance
(called the rival’s distance).
The results show a negative effect of distance on operators’ decisions to participate, which is
in line with the results of Cantillon and Pesendorfer (2007), and that an operator is more likely to
participate if the second closest operator is farther away from the contract area. Investigating the
marginal effects shows that being directly next to a contract area implies an almost 100 percent
chance of placing a bid, but that this probability decreases substantially with distance. With a
dead-running distance of, on average, 2 kilometers, the probability is down to 70 percent, while
being 10 kilometers away reduces the probability to 30 percent. This confirms what was partly
established in Paper II, that the local competitive environment matters. The PTAs must consider
what impacts their geographical decisions on contract design have, and not simply tender lines

14
in the same contract areas as they did 20 years ago, which is sometimes the case. The rival’s
distance to the contract has some effect, but the most important effect is the operator’s own
distance to the contract.
Large operators are less affected by their distance from the contract area when deciding whether
to participate, compared to other smaller operators and compared to cooperation companies
(co-ops). The large operators are also more inclined to place a bid if a depot is offered for use in
the contract, which is not the case for the rest of the operators.

Paper IV. Competition in Swedish Passenger Railway

In line with the deregulation and liberalization path of the railway sector in Sweden, and to some
extent the rest of Europe, Sweden was one of the first countries to impose open access competition
on the passenger railway market. Starting with night trains and charter traffic in 2007, the market
was gradually fully opened for commercial traffic in 2010 (Alexandersson, 2010). The market
opening led to new traffic only on a few minor routes or with niche services (Fröidh and Nelldal,
2015), However, a substantial entry occurred in March 2015 when the Hong Kong-based operator
MTR started running services on the Stockholm-Gothenburg line. This paper investigates whether
the entry on the line has lowered ticket prices faced by travelers.
Swedish railway market data is scarce, if not non-existent, as noted above. Therefore, for this
paper, a Python-scripted web crawler was created to collect price information from the Internet
on SJ and MTR trains. The program collected the lowest available price for almost all departures
on six Swedish routes from July 2014 to June 2016. The price request was also made for each
departure and day up to 31 days before the departure. This created a rich dataset, both in terms
of the number of lines and departures, but also over time. Because the dataset spans the pre- and
post-entry period of MTR, these factors arguably make the data unique not only in a Swedish,
but also international, context. Beria et al. (2016) use similar data for Italy, but for a shorter
pre-entry period and for only a single line.
Because price data is available for the (treated) Stockholm-Gothenburg line, as well as for
five other lines (controls), a difference-in-difference approach can be used to identify the price
effect following MTR’s entry. The identification strategy is that the price development on the
treated line should behave similarly as the control lines if the entry would not have occurred. The

15
difference between the treated and control groups is therefore the price change following MTR’s
entry. This is also known as the parallel-path assumption (Angrist and Pischke, 2008), and this
is a tested and validated assumption.
The results show that following the entry of MTR the average price for the incumbent SJ on
the Stockholm-Gothenburg line as of June 2016 was 12.6 percent lower compared to the average
price prior to MTR’s entry. The prices for MTR were shown to be systematically around 100
SEK lower on average than the average SJ prices. Therefore, the consumers are facing lower
prices than before the entry. The largest price reduction for SJ was found for 13 days before
departure (a decrease of 13.8 percent), while the smallest was 31 days before departure (a 9.4
percent decrease). Finally, the price change occurred gradually over the 15-month period after
the entry and was not an immediate downward change.
The mechanisms behind the incumbent’s price decrease are, however, not possible to infer
from the collected data. Two explanations are plausible - an active price adjustment by the
incumbent, or the price decrease being a consequence of a dynamic pricing system adjusting
for fewer passengers. It is argued in the paper that the latter is the more likely explanation.
Further, this paper analyzes the short-run effect of competition, and the current data cannot
answer whether the price decrease will remain stable or if it will decrease further or even increase
again in the future.

5 Conclusions

As noted by Gwilliam (2008), the regulatory cycle is seldom completed in the sense that a state
returns to its initial position in the cycle. Thus, the next step in the cycle for Sweden is not
necessarily that towards private sector area monopolies. Swedish public transport in general is
currently in a situation with competitive private supply, both in the subsidized regional public
transport services and in passenger railway services. The more liberalized markets have arguably
aimed at taking Sweden from the right-hand side of the regulatory curve in Figure 1 and more
towards the left. As for the market opening in railway transport, the results in paper IV where
prices were found to be lower after MTR’s entry suggest that this reform has improved the welfare
outcome, given that price and supply are the most important factors. Quality of service is also
an issue, but there is, as far as the author of this thesis is aware, no evidence that quality has

16
decreased on the line. Tomeš et al. (2014) noted quality improvements in the Czech Republic
following open access competition, but the long-term effects of the open access regime in Sweden
are yet to be determined.
One of the overall conclusions in Paper II is that there are few single measures that PTAs can
use to substantially increase the number of bidders. However, one thing they can do is to lower
the number of simultaneous tenders. Still, it is not irrelevant how the PTAs design their contracts.
Some contract design factors do in fact have a statistically significant effect, and the PTAs should
therefore still consider what impact their contract specifications can have on outcome variables
such as costs, patronage, social welfare, and the competitive environment, and combining many
contract design variables simultaneously might achieve better bidding outcomes. The combined
results of Papers II and III show that the contract area can have an impact participation and
the number of bids seen, which in turn affects the outcome variables. The way in which special
requirements are imposed in the contracts also affects the behavior of operators, especially in
terms of the costs to meet the requirements (Camén and Lidestam, 2016), and this must be
weighed against the intended benefits of the same requirements.
The current competitive tendering regime in Sweden is, arguably, rather uncontested and is
supported by the EU regulations on public transport (EU 1370/2007). EU and Swedish law does,
however, allow traffic to be run in-house by a publicly owned operator. The in-house alternative
has received more attention in recent years, most notably in the counties of Örebro and Jämtland
who both have investigated the issue because they are experiencing low quality and cost increases,
but they have limited possibilities to adjust the traffic with the tendering regime and have little
competition in their respective markets (Region Jämtland Härjedalen, 2016; Region Örebro län,
2016). The last concern is discussed to some extent in Papers II and III, but is also raised
in Mathisen and Solvoll (2008) and Mathisen (2016) where structural changes and ownership links
in the operator market following tendering are discussed. The discussion on co-ops in Paper III is
also relevant here because these could increase competition, as well as lower it. While the region
of Jämtland Härjedalen decided not to implement in-house operations, the region of Örebro will
start insourcing public transport operations in 2019 when some of the region’s current contracts
expire. This is a step towards the regulated public monopoly on Gwilliam’s (2008) regulatory
cycle, which is in the opposite direction of the joint work through the “Partnersamverkan för

17
Förbättrad Kollektivtrafik” to enhance tenders and the direction of legislative developments, which
in 2012 liberalized the public transport market for commercial actors to provide traffic within
counties3 . The arguments for going in the opposite direction and the arguments for continuing to
use competitive tendering are also similar - quality while curbing cost increases. The contrasts are
perhaps best illustrated in the regulatory curve where both regimes aim at reaching the optimal
level of prescription, but are approaching it from different sides of the optimum. The results
from Paper I, and other studies for Europe, suggest that the in-house regime is less cost efficient,
and will, all else being equal, not yield improved welfare. The difference in quality between the
tendering and in-house regimes is, however, not well explored in the transport literature. The
argument for competitive tendering is that quality control can be imposed in the contracts, while
the counter-argument would be that the PTA then does not have the ability to make changes to
the service during the contract period and that such adjustments can be made more easily with
an in-house regime. This issue calls for more empirical research, for which the Örebro case would
be interesting to follow.
The in-house regime is one direction the future of public transport regulation can take.
Procurement through competitive tendering is not likely to be abandoned soon, but the research
literature has been concerned with its potential deficiencies. Yvrande-Billon (2006) argues that the
success of competitive tendering relies on transparency. Among others, Hensher and Stanley (2008)
argue for increased use of negotiated performance-based contracts (NPBCs) to accommodate
situations with high transaction costs and few bidders, and this is a contract form where the
operators receive a subsidy per kilometer driven and receive incentive payments relating to key
performance indicators (KPIs). This would tend to steer the operator’s behavior towards higher
quality and better service from the societal point of view (Hensher and Stanley, 2003). A crucial
part of the NPBC, argue Hensher and Stanley, is also that the contract is negotiated between
the relevant parties (most importantly the PTA and the operator(s)) to “create” the service,
rather than “determining” it as in the case of competitive tendering (Hensher, 2015). For this,
trust between the actors and open books are necessary. Ultimately, if agreements are hard to
reach, the PTA always has the threat of tendering the service (Hensher, 2015). This is not very
different from some performance based-like contracts (for example, incentives contracts or gross
3 The new law has yet not had any substantial impact on the establishment of new lines by commercial

operators (Transport Analysis, 2013).

18
cost contracts with incentives) that are already used in Sweden. The main difference is the
negotiation part. A transparent process is, however, necessary, and all relevant parties must be
included in order to fulfill the important principle of equal treatment in public procurements.
Finally, what has been much of the focus in this thesis is the public availability of data. This
situation must improve. The data collection undertaken for this thesis has proven that it is
possible to acquire this information, and it has shown how such information can be put to use.
Regarding public transport tenders, the situation illustrated in Figure 2 is not acceptable with
close to one year until full enquiry documentation and awarding protocols were sent out by some
of the PTAs. References to corporate secrecy (företagshemligheter) were often claimed by the
PTAs, which is also worrying. In many cases, the bid sums and evaluations, which decide the
contract winner, were claimed to potentially hurt the bidder or reduce the chances of future bids.
However, not being able to evaluate the tendering and bidding process must be in clear violation
of the principle of public access to official documents, which in turn limits the possibility for
the public to claim responsibility over the public sector. Evaluation and research on the public
transport area is an important contribution to enhancing the welfare outcome in the Swedish
transport sector, irrespective of the level of prescription, but such research is only possible with
transparency and good data availability. Finally, van de Velde (2015) makes the important point
that “Researchers need to avoid playing down the importance of this issue [data availability and
quality, authors note] in this sector in order to guarantee the quality of quantitative research”.

19
 References

Aguirregabiria, V. and Ho, C.-Y. (2012). A dynamic oligopoly game of the US airline industry:
Estimation and policy experiments. Journal of Econometrics, 168(1):156–173.

Alexandersson, G. (2010). The accidental deregulation: Essays on reforms in the Swedish bus
and railway industries 1979-2009. PhD thesis, Stockholm School of Economics.

Amaral, M., Saussier, S., and Yvrande-Billon, A. (2013). Expected number of bidders and winning
bids: Evidence from the London bus tendering model. Journal of Transport Economics and
Policy, 47(1):17–34.

Angrist, J. D. and Pischke, J.-S. (2008). Mostly harmless econometrics: An empiricist’s companion.
Princeton university press.

Athey, S., Coey, D., and Levin, J. (2013). Set-asides and subsidies in auctions. American Economic
Journal: Microeconomics, 5(1):1–27.

Augustin, K. and Walter, M. (2010). Operator changes through competitive tendering: Empirical
evidence from German local bus transport. Research in Transportation Economics, 29(1):36–44.

Avenali, A., Boitani, A., Catalano, G., D’Alfonso, T., and Matteucci, G. (2016). Assessing
standard costs in local public bus transport: Evidence from Italy. Transport Policy, 52:164–174.

Avenali, A., Boitani, A., Catalano, G., D’Alfonso, T., and Matteucci, G. (In Press, 2017).
Assessing standard costs in local public bus transport: A hybrid cost model. Transport Policy.

Bajari, P., McMillan, R., and Tadelis, S. (2008). Auctions versus negotiations in procurement:
An empirical analysis. The Journal of Law, Economics, & Organization, 25(2):372–399.

20
Bajari, P. and Ye, L. (2003). Deciding between competition and collusion. Review of Economics
and statistics, 85(4):971–989.

Beck, A. (2011). Experiences with competitive tendering of bus services in Germany. Transport
Reviews, 31(3):313–339.

Beck, A. and Walter, M. (2013). Factors affecting tender prices in local bus transport evidence
from Germany. Journal of Transport Economics and Policy, 47(2):265–278.

Bergantino, A. S., Capozza, C., and Capurso, M. (2015). The impact of open access on intra-and
inter-modal rail competition. a national level analysis in Italy. Transport Policy, 39:77–86.

Beria, P., Redondi, R., and Malighetti, P. (2016). The effect of open access competition on average
rail prices. the case of Milan–Ancona. Journal of Rail Transport Planning & Management,
6(3):271–283.

Brueckner, J. K., Lee, D., and Singer, E. S. (2013). Airline competition and domestic US airfares:
A comprehensive reappraisal. Economics of Transportation, 2(1):1–17.

Camén, C. and Lidestam, H. (2016). Dominating factors contributing to the high(er) costs for
public bus transports in Sweden. Research in Transportation Economics, 59:292–296.

Cantillon, E. and Pesendorfer, M. (2006). Auctioning bus routes: The London experience. In
Cramton, P. C., Shoham, Y., and Steinberg, R., editors, Combinatorial Auctions, chapter 22.
MIT Press, Cambridge.

Cantillon, E. and Pesendorfer, M. (2007). Combination bidding in multi-unit auctions. CEPR

discussion paper 6083. Retrieved from http://ecantill.ulb.be/ecantillon/uploads/
CEPR-DP6083.pdf.

Ciliberto, F. and Williams, J. W. (2010). Limited access to airport facilities and market power in
the airline industry. The Journal of Law and Economics, 53(3):467–495.

Consul, P. C. (1989). Generalized Poisson Distributions. Dekker New York.

Dalen, D. M. and Gómez-Lobo, A. (2003). Yardsticks on the road: Regulatory contracts and cost
efficiency in the Norwegian bus industry. Transportation, 30(4):371–386.

21
Diandas, J. and Roth, G. (1995). Alternative approaches to improving route bus services in
Sri Lanka. In Proc. 4th International Conference on Competition and Ownership in Land
Passenger Transport (Thredbo 4), pages 463–482, Rotorua, New Zealand.

EU 1370/2007 (2007-12-03). REGULATION (EC) No 1370/2007 OF THE EUROPEAN PARLIA-

MENT AND OF THE COUNCIL of 23 October 2007 on public passenger transport services
brail and by road and repealing Council Regulations (EEC) Nos 1191/69 and 1107/70. Official
Journal of the European Union, 50(L 315):1–13.

Farsi, M., Filippini, M., and Greene, W. (2005). Efficiency measurement in network industries:
Application to the Swiss railway companies. Journal of Regulatory Economics, 28(1):69–90.

Farsi, M., Filippini, M., and Kuenzle, M. (2006). Cost efficiency in regional bus companies: An
application of alternative stochastic frontier models. Journal of Transport Economics and
Policy, 40(1):95–118.

Filippini, M. and Prioni, P. (1994). Is scale and cost inefficiency in the Swiss bus industry a
regulatory problem? evidence from a frontier cost approach. Journal of the Economics of
Business, 1(2):219–232.

Fiorio, C. V., Florio, M., and Perucca, G. (2013). User satisfaction and the organization of local
public transport: Evidence from European cities. Transport Policy, 29:209–218.

Fröidh, O. and Nelldal, B.-L. (2015). The impact of market opening on the supply of interregional
train services. Journal of Transport Geography, 46:189–200.

Goetz, A. R. and Vowles, T. M. (2009). The good, the bad, and the ugly: 30 years of US airline
deregulation. Journal of Transport Geography, 17(4):251–263.

Goolsbee, A. and Syverson, C. (2008). How do incumbents respond to the threat of entry?
evidence from the major airlines. The Quarterly journal of economics, 123(4):1611–1633.

Graells, A. S. (2015). Public Procurement and the EU Competition Rules. Bloomsbury Publishing.

Gwilliam, K. (2008). Bus transport: Is there a regulatory cycle? Transportation Research Part
A: Policy and Practice, 42(9):1183–1194.

22
Hart, O. and Moore, J. (1988). Incomplete contracts and renegotiation. Econometrica, 56(4):755–
785.

Hart, O., Shleifer, A., and Vishny, R. W. (1997). The proper scope of government: Theory and
an application to prisons. The Quarterly Journal of Economics, 112(4):1127–1161.

Hensher, D. (2015). Cost efficiency under negotiated performance-based contracts and bench-
marking. Journal of Transport Economics and Policy, 49(1):133–148.

Hensher, D., Mulley, C., and Smith, N. (2013). Towards a simplified performance-linked value
for money model as a reference point for bus contract payments. Research in Transportation
Economics, 39(1):232–238.

Hensher, D. and Stanley, J. (2003). Performance-based quality contracts in bus service provision.
Transportation Research Part A: Policy and Practice, 37(6):519–538.

Hensher, D. and Stanley, J. (2008). Transacting under a performance-based contract: The role of
negotiation and competitive tendering. Transportation Research Part A: Policy and Practice,
42(9):1143–1151.

Holmgren, J. (2013). The efficiency of public transport operations - an evaluation using stochastic
frontier analysis. Research in Transportation Economics, 39(1):50–57.

Hultén, S. (2015). Kontrakt och konkurrens i den regionala kollektiva busstrafiken. Uppdrags-
forskningsrapport 2015:7, Swedish Competition Authority.

Iossa, E. and Waterson, M. (In Press, 2017). Maintaining competition in recurrent procurement
contracts: A case study on the London bus market. Transport Policy.

Jansson, J. O. (1979). Marginal cost pricing of scheduled transport services: A development

and generalisation of Turvey and Mohring’s theory of optimal bus fares. Journal of Transport
Economics and Policy, 13(3):268–294.

Jarboui, S., Forget, P., and Boujelbene, Y. (2014). Transport firms’ inefficiency and managerial
optimism: A stochastic frontier analysis. Journal of Behavioral and Experimental Finance,
3:41–51.

23
Jarboui, S., Pascal, F., and Younes, B. (2013). Public road transport efficiency: A stochastic
frontier analysis. Journal of Transportation Systems Engineering and Information Technology,
13(5):64–71.

Jofre-Bonet, M. and Pesendorfer, M. (2003). Estimation of a dynamic auction game. Econometrica,

71(5):1443–1489.

Johnson, D. and Nash, C. (2012). Competition and the provision of rail passenger services.
Journal of Rail Transport Planning & Management, 4:104–113.

Kumbhakar, S. C., Wang, H., and Horncastle, A. P. (2015). A practitioner’s guide to stochastic
frontier analysis using Stata. Cambridge University Press.

Lundberg, S., Marklund, P.-O., Strömbäck, E., and Sundström, D. (2015). Using public procure-
ment to implement environmental policy: An empirical analysis. Environmental Economics
and Policy Studies, 17(4):487–520.

Luo, D. (2014). The price effects of the Delta/Northwest airline merger. Review of Industrial
Organization, 44(1):27–48.

Mathisen, T. A. (2016). Competitive tendering and cross-shareholding in public passenger

transport. Transport Policy, 48:45–48.

Mathisen, T. A. and Solvoll, G. (2008). Competitive tendering and structural changes: An

example from the bus industry. Transport Policy, 15(1):1–11.

Mohring, H. (1972). Optimization and scale economies in urban bus transportation. The American
Economic Review, 62(4):591–604.

Nash, C., Nilsson, J.-E., and Link, H. (2013). Comparing three models for introduction of
competition into railways. Journal of Transport Economics and Policy, 47(2):191–206.

Nilsson, J.-E. (2011). Kollektivtrafik utan styrning. ESO Report 2011:6.

Obeng, K. (2013). Bus transit technical efficiency using latent class stochastic indirect production
frontier. Applied Economics, 45(28):3933–3942.

24
Ottoz, E., Fornengo, G., and Di Giacomo, M. (2009). The impact of ownership on the cost of bus
service provision: an example from Italy. Applied Economics, 41(3):337–349.

Piacenza, M. (2006). Regulatory contracts and cost efficiency: Stochastic frontier evidence from
the Italian local public transport. Journal of Productivity Analysis, 25(3):257–277.

Preston, J. (2001). An overview of public transport reforms in Great Britain and forecasts for
the future. International Journal of Transport Economics, 28(1):23–48.

Preston, J. (2008). Competition in transit markets. Research in Transportation Economics,

23(1):75–84.

Preston, J. and Robins, D. (2013). Evaluating the long term impacts of transport policy: The
case of passenger rail privatisation. Research in Transportation Economics, 39(1):14–20.

Preston, J., Whelan, G., and Wardman, M. (1999). An analysis of the potential for on-track
competition in the British passenger rail industry. Journal of Transport, Economics and Policy,
33(1):77–94.

Rail Delivery Group (2017). Fares data. http://data.atoc.org/fares-data. Data set.

Region Jämtland Härjedalen (2016). Mötesprotokoll 2016-02-29. §10 Kollektivtrafik i egen regi
eller upphandlad trafik (RUN/1053/2015). Utskottet för infrastruktur, pages 1–4.

Region Örebro län (2016). Mötesprotokoll 2016-03-16. §18 Alternativa driftsformer av kollektiv-
trafik (16RS1093). Nämnden för samhällsbyggnad, pages 5–6.

Roy, W. and Yvrande-Billon, A. (2007). Ownership, contractual practices and technical efficiency:
The case of urban public transport in France. Journal of Transport Economics and Policy,
41(2):257–282.

Sakano, R. and Obeng, K. (2012). Examining the inefficiency of transit systems using latent class
stochastic frontier models. In Journal of the Transportation Research Forum, volume 50.

SFS 1949:105 (1949). Tryckfrihetsförordningen.

Tomeš, Z., Kvizda, M., Jandová, M., and Rederer, V. (2016). Open access passenger rail
competition in the Czech Republic. Transport Policy, 47:203–211.

25
Tomeš, Z., Kvizda, M., Nigrin, T., and Seidenglanz, D. (2014). Competition in the railway
passenger market in the Czech Republic. Research in Transportation Economics, 48:270–276.

Transport Analysis (2013). Utvärdering av marknadsöppningar i kollektivtrafiken. Report 2013:13.

Transport Analysis (2015). Avtalen för den upphandlade kollektivtrafiken 2013. Report 2015:13.

Transport Analysis (2017). Local and regional public transport 2016. Retrieved from tables at
http://www.trafa.se/kollektivtrafik/kollektivtrafik/.

van de Velde, D. (2015). Local and regional public transport. In Nash, C., editor, Handbook of
Research Methods and Applications in Transport Economics and Policy, pages 345–358. Edward
Elgar Publishing, Cheltenham.

van de Velde, D. and Preston, J. (2013). Workshop 3B: Governance, ownership and competition
issues in deregulated (free market) public transport: Lessons that can be learnt from developed
and developing economies. Research in Transportation Economics, 39(1):202–207.

Veeneman, W., Wilschut, J., Urlings, T., Blank, J., and van de Velde, D. (2014). Efficient frontier
analysis of Dutch public transport tendering: A first analysis. Research in Transportation
Economics, 48:101–108.

Vickerman, R. (2015). Economic impacts of transport policy. In Nash, C., editor, Handbook of
Research Methods and Applications in Transport Economics and Policy, pages 345–358. Edward
Elgar Publishing, Cheltenham.

Walter, M. (2011). Some determinants of cost efficiency in German public transport. Journal of
Transport Economics and Policy, 45(1):1–20.

Winkelmann, R. (2013). Econometric analysis of count data. Springer Science & Business Media.

Yvrande-Billon, A. (2006). The attribution process of delegation contracts in the French urban
public transport sector: why competitive tendering is a myth. Annals of public and cooperative
economics, 77(4):453–478.

26