category: literaturenote citekey: viegas10evolutioneudirective2005 title: 10. The evolution of the EU Directive on road charges and tolls for trucks authors: "Viegas, José M." year: 2005 date: 2005-00-00 2005 doi: 10.1016/s0739-8859(05)80013-7 publication: Research in Transportation Economics url: "https://doi.org/10.1016/s0739-8859(05)80013-7" zotero_key: TKEE4I23 zotero_storage: QHENAKGI collections: transpordi_planeerimine folder: 001_artiklid firstAuthor: "Viegas, José M."

status: converted

INTERURBAN ROAD CHARGING FOR TRUCKS IN EUROPE

RESEARCH IN TRANSPORTATION ECONOMICS

Series Editor: Martin Dresner

Recent Volumes:

Volumes 1–6: Research in Transportation Economics –

Edited by B. Starr McMullen

Volume 7: Railroad Bankruptcies and Mergers from

Chicago West 1975–2001: Financial Analysis and Regulatory Critique – Michael Conant

Volume 8: Economic Impacts of Intelligent Transportation

Systems: Innovations and Case Studies – Evangelos Bekiaris & Yuko Nakanishi

Volume 9: Road Pricing: Theory and Evidence – Georgina

Santos

Volume 10: Transportation Labor Issues and Regulatory

Reform – James Peoples & Wayne K. Talley

INTERURBAN ROAD CHARGING FOR TRUCKS IN EUROPE

EDITED BY

JOSE M. VIEGAS ´

CESUR – Instituto Superior T´ecnico, Portugal

2005

Amsterdam – Boston – Heidelberg – London – New York – Oxford Paris – San Diego – San Francisco – Singapore – Sydney – Tokyo

ELSEVIER B.V. ELSEVIER Inc. ELSEVIER Ltd ELSEVIER Ltd Radarweg 29 525 B Street, Suite 1900 The Boulevard, Langford 84 Theobalds Road P.O. Box 211 San Diego Lane, Kidlington London 1000 AE Amsterdam CA 92101-4495 Oxford OX5 1GB WC1X 8RR

The Netherlands USA UK UK

This work is protected under copyright by Elsevier Ltd, and the following terms and conditions apply to its use:

Photocopying

Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use.

Permissions may be sought directly from Elsevier's Rights Department in Oxford, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: permissions@elsevier.com. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions).

In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments.

Derivative Works

Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations.

Electronic Storage or Usage

Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter.

Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier's Rights Department, at the fax and e-mail addresses noted above.

Notice

No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.

First edition 2005

British Library Cataloguing in Publication Data

A catalogue record is available from the British Library.

ISBN: 0-7323-1142-8

ISSN: 0739-8859 (Series)

∞The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

LIST	OF CONTRIBUTORS	vii
	GLOSSARY	ix
	FOREWORD	xi
1.	INTRODUCTION: PAYING FOR ROAD USE José M. Viegas	1
2.	BASIC ROAD PRICING SOLUTIONS Heike Link and Louise Stewart-Ladewig	9
3.	TECHNICAL AND PROCEDURAL DESIGN ELEMENTS Andrea Felix	27
4.	LEGAL, INSTITUTIONAL AND CONTRACTUAL ISSUES Rosário Macário	43
5.	THE ACCEPTABILITY DIMENSION Max Herry	75
6.	ASSESSMENT OF ROAD PRICING SCHEMES BASED ON DIFFERENT NATIONAL SITUATIONS René Neuenschwander and Florian Gubler	89
7.	POSITION AND RECENT TRENDS IN EUROPEAN COUNTRIES Chiara Borgnolo, Louise Stewart-Ladewig and René Neuenschwander	109

8.	HARMONISATION AND INTEROPERABILITY OF NATIONAL TOLLING SCHEMES Jesper Mertner and Michael Knørr Skov	133
9.	ECONOMIC AND ENVIRONMENTAL IMPACTS OF
	ROAD TOLLS FOR HGVs IN EUROPE Lorenz Kleist and Claus Doll	153
10.	THE EVOLUTION OF THE EU DIRECTIVE ON ROAD CHARGES AND TOLLS FOR TRUCKS José M. Viegas	193
11.	CONCLUSIONS: PAYING FOR ROAD USE, THE WAY AHEAD José M. Viegas	201

LIST OF CONTRIBUTORS

Chiara Borgnolo TRT, Trasporti e Territorio, Milan, Italy

Claus Doll University of Karlsruhe, Germany Andrea Felix Rapp Trans AG, Basel, Switzerland

Florian Gubler Ecoplan, Bern, Switzerland

Max Herry HERRY Consult GmbH, Vienna, Austria

Lorenz Kleist University of Karlsruhe, Germany Michael Knørr Skov University of Karlsruhe, Germany

Heike Link German Institute for Economic Research,

Berlin, Germany

Ros ´ario Mac ´ario CESUR – Instituto Superior Tecnico; and ´

TIS.pt Consultores em Transportes, Inovac¸ao˜

e Sistemas, s.a., Lisbon, Portugal

Jesper Mertner COWI A/S, Kongens Lyngby, Denmark

Ren´e Neuenschwander Ecoplan, Bern, Switzerland

Louise Stewart-Ladewig German Institute for Economic Research,

Berlin, Germany

Jos´e M. Viegas CESUR – Instituto Superior Tecnico; and ´

TIS.pt Consultores em Transportes, Inovac¸ao˜

e Sistemas, s.a., Lisbon, Portugal

GLOSSARY

AVI Automatic Vehicle Identification

BOT Built Operate Transfer

CN Cellular Network

COR A charge, differentiated according to vehicle classes, raised

when entering a cordon around a defined area or city

DAREA Distance dependent area pricing, a regime where the use of

all roads within a specific country or other geographical area are subject to charge based on the kilometres driven

DSRC Dedicated Short Range Communication

DESIRE Designs for Interurban Road pricing schemes in Europe

EC European Commission

ECMT European Conference of Ministers of Transport

EFC Electronic Fee Collection

EU European Union

FP Framework Programme

GIS Geographical Information System

GPS Global Positioning System

GSM Global Standard for Mobile Communication

GPRS General Packet Radio Service

HGV Heavy Goods Vehicles

IRPS Inter-urban Road Pricing

LKW LastKraftWagen (HGV – Heavy Goods Vehicles)

LPR License plate reading

NET Distance dependent network pricing, a regime based on the

kilometres driven on a network of interrelated motorways of

the same hierarchy level

OBU On Board Unit

OCR Optical Character Recognition

PAS The levy of a toll for the use of a particular infrastructure

segment such as a motorway section, bridge or tunnel

PERM Driving permit or area licensing, a charge which is levied

for permission to drive within a certain perimeter and within a certain time period, and which is differentiated according

to vehicle classes

PPP Public-Private Partnership

NUTS Nomenclature of Territorial Units for Statistics (Europe)

SWOT A type of analysis often used in management to identify

Strengths, Weaknesses, Opportunities and Threats facing an

organisation

TEN-(T) Trans European Network (Transport)

TETRA TErrestrial Trunked RAdio

RSE Road Side Equipment

UMTS Universal Mobile Telecommunication System

WAP Wireless Application Protocol

FOREWORD

This book follows the production of a research project for the European Commission¹ on a theme that has been in the limelight of European Transport Policy since at least 1998: direct payment for the use of road infrastructure.

The consortium that produced that research included teams from 8 countries then in the EU, plus Hungary, Switzerland and Brazil. As the project evolved – from September 2000 to February 2003 – some of the countries present in the consortium (France, Italy, Portugal) had a long tradition of motorway tolls on all vehicles, another (Switzerland) started the implementation of an innovative tolling scheme for trucks on all roads, (Austria) took the final decision on a classical motorway tolling scheme, but only for trucks, (Holland) came to the final stages of preparation of a tolling scheme for all vehicles on all roads, but saw its Government fall precisely because of coalition disputes over this subject, and (Germany) announced its preference for a satellite based motorway tolling scheme for trucks, made the legislative decisions supporting it, and then went through a litigious tendering procedure, followed by a series of technological problems with the new solutions. On top of that, a new European Commission White Paper on Transport Policy, with a large role for infrastructure charging, was published in September 2001. Naturally, this turbulent environment on the real world outside has provided the consortium with a rich framework for its own work, with a great variety of points of view and arguments available from the various national discussions.

This wealth of external material helped set the priorities of the project: define a solid structure for the analysis of the multiple issues, and identify the circumstances where each type of solution is the preferred choice, recognizing from the outset that there will not be a "universally first best" solution. Solutions were analyzed in several dimensions related to the incidence and modulation of the tolls, the technology adopted, the institutional and contractual design. Besides these points aimed at providing the best solution for each case, two other research dimensions have been tackled, more connected to the very decision of having tolls or not: the acceptance of the measure in regions where access to good roads has been provided without direct payment, and the economic impacts of the introduction of tolls, on regional economies as well as on modal shares.

Shortly after the conclusion of the project, in July 2003, the European Commission presented its proposal for an amendment of Directive 1999/62, which had been the previous legal framework for making trucks pay for road use. Several of the recommendations of the DESIRE project were taken on board this proposal.

This amendment has gone through the normal process implied by rules of co-decision between the European Council and Parliament, and at the time of writing seems close to final form, having already had one version adopted by the Parliament on April 20, 2004, but failing to obtain approval in the Council of Transport Ministers on June 10th 2004.

The process has lived on, and strains are still present around the version prepared by the Irish Presidency for the June 10th Council meeting, so there seems to be good reason for higher visibility of the research done within DESIRE and following it. That is exactly the purpose of this book.

The editor is thankful for the technical contributions and strong engagement of all members of the consortium. The authors of the various chapters of this book are a part of them. All their contributions have been rewritten for this book, taking into account the developments occurred in the meantime. For them, and their availability to renew the effort, my sincere thanks.

Jose M. Viegas ´ Editor

NOTE

Project DESIRE – DESigns for Inter-urban Road pricing schemes in Europe, contract no. 2000 – CM.10501, in the fourth framework programme of the European Union (Competitive and Sustainable Growth Program), running from September 2000 to May 2003, of which the editor of this book was Project Coordinator.

1. INTRODUCTION: PAYING FOR ROAD USE

Jose M. Viegas ´

Making users of transport infrastructure pay for that use is not new, both in linear infrastructure modes and in nodal infrastructure modes, but with rather different styles according to the nature of the infrastructure, the level of integration between infrastructure and movement in transport production, and the level of atomisation of the set of uses (number of users and their distribution in time and space).

In the case of road transport, specific charges for special facilities providing improved speed and comfort has a long tradition, starting with turnpikes at the beginning of the 18th century, long before there was motorised transport. The opening of new turnpikes in Great Britain steadily increased until the 1850s, when the advent of the railways brought about a much better travelling alternative.

If we stick to the first decades of motorised transport by road, in the beginning of the 20th century, use of the infrastructure has generally been made available without direct payment at the point of use, except in connection with special pieces of infrastructure which had an especially high construction cost, but also provided an especially high degree of convenience, like bridges and tunnels. The rest of the road network was financed through public budgets.

Direct payment for every trip was not feasible because of the "open access" design of the network and the enormously high transaction costs it would have implied. On the other hand, indirect payment (through taxes) has allowed lasting and significant levels of subsidisation from public budgets as the social penetration of road transport was increasing, as well as cross-subsidisation among regions and among user groups, favouring the overall development of the road networks and

Interurban Road Charging for Trucks in Europe Research in Transportation Economics, Volume 11, 1–7 Copyright © 2005 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0739-8859/doi:10.1016/S0739-8859(04)11001-9

2 JOSE M. VIEGAS ´

the transformation of road based modes in the dominant providers of mobility at present.

In most countries, users of the road system paid indirectly on top of general taxation, through a dedicated package possibly, but not always, consisting of three types of taxes: on vehicle purchase, on annual circulation permits, and on fuel duties. The relative volumes of revenue from these three types of sources have varied from country to country as well as throughout time in each country.

Direct payment for the use of roads has been (re-)introduced in the 20th century in connection with *financing needs*for roads of improved standards, with examples from Germany (west of Berlin, construction started in 1909 and was open to service in 1921), and in Italy (the first real motorway, Milano–Varese, in 1924). The difficult economic circumstances of the 1930s, slowing the social penetration of the car and justifying massive State intervention for economic recovery, as well as the militarist intentions of those times in Germany and Italy led to publicly financed programs of motorway construction at a fast pace.

Elsewhere, development of motorways is basically a post-World War II process: the first stretch of motorway in France opened in 1946, with just 20 km, and the U.S. Federal Interstate freeway program was launched in 1956. An international definition for motorway was approved in the 1957 Road Congress and subsequently approved by the European Conference of Ministers of Transport.

In Western Europe, introduction of motorway tolls has been decided mostly in southern countries, as a way to accelerate their construction vis-a-vis what would ` have been possible under the single support by State budgets, although in many cases there was substantial engagement of public (sometimes local and regional) entities in the concessionaires. Fully private concessionaire societies only appear in the 1970s when the concept was introduced and generally accepted, and car ownership was much higher, thus substantially reducing the risk for investors. This allowed a strong acceleration of the pace of expansion of the motorway networks in those countries.

Drivers were persuaded that the toll, on top of the indirect payments that were always there, was the natural price supplement for access to a "superior good" and in general there was an implicit guarantee (but no legal provision for it) that a "parallel road of reasonable quality" would always be available. In the late 1990s in Portugal, after more than 20 years of tolled motorways and in the middle of a large expansion program, this proviso led to public rebellion in a region of difficult relief where the alternative was considered too bad, and ultimately to a 15 km toll-free stretch included in a tolled motorway concession of some 180 km.

More recently, some central European countries have adopted or tried to adopt similar schemes of tolled motorways, but the introduction of these new schemes, at a time of economic turbulence following their adoption of a market economy, Introduction 3

faced much greater difficulties than those earlier attempts which were made in periods of sustained growth and in more affluent societies.

In all of these cases, tolls were used as an instrument for financing construction and maintenance of motorways, but scientific interest on road tolls was raised many decades ago (Pigou, 1920) as an instrument to fight congestion, with which the marginal cost perceived by drivers when using the road could be brought to coincide with the social marginal cost of such use, thus allowing a more efficient use of the road.

Of course, congestion was a very localised problem then, but as it became more common on many roads, so more scientists began to work on ways to use pricing to manage it, and quite some time later, policy texts and even action began following the appointed path. Urban road pricing was introduced in Singapore in 1975 (as an Area Licensing Scheme, converted to full electronic road pricing in 1998), and as cordon tolls in the Norwegian cities of Bergen (1986), Trondheim (1988) and Oslo (1990).

In the Norwegian case, the main objective of the toll scheme was to finance infrastructure which would later help reduce congestion, not the direct reduction of congestion, so the toll level was relatively low and did not constitute a significant deterrent to driving into the city. London has introduced its congestion charging scheme in 2003, at a price level of £5 which has proved to be high enough to reduce traffic volumes in the tolled area by some 15% and congestion in those roads by 30% (Transport for London, 2004).

Since 1995 there have been several (and increasing) schemes of variable toll levels, generally called Value Pricing, managed to ensure a desired, sometimes contractually established, level of service. This now constitutes a Pilot Program of the U.S. Federal Highway Administration (Federal Highway Administration, 2004).

In parallel with some of these events, the European Commission developed a massive effort aiming at the introduction of pricing instruments towards sounder transport policies. The first official document in this stream is the 1995 Green Paper on Fair and Efficient Pricing (European Commission, 1995) where the discussion is focused especially on road transport and the principle of charging users for all types of costs – external as well as internal – is clearly put forward. Special attention is given to heavy goods vehicles, and the notion of a kilometre charge is repeatedly put forward. As for the setting of price levels, the objectives of cost recovery (full cost pricing) and of efficiency (social marginal cost pricing) are placed on a similar footing.

A White Paper on Fair Payment for Infrastructure Use followed (European Commission, 1998) in 1998, where the principle of internalisation of external costs is renewed, and an attempt is made to address all modes, albeit still with 4 JOSE M. VIEGAS ´

most specific examples being related to road transport. There is here a greater emphasis on ensuring fair competition between modes and between transport operators of different member countries, and a clear support for the adoption of prices, based on marginal costs is given. Shortly before the publication of this White Paper, the European Commission created a High Level Group on Transport Infrastructure charging, for advice on the development of a comprehensive set of charging principles. This group presented a series of reports in that direction, with a strong recommendation towards adoption of social marginal cost pricing, and specific recommendations for the systematic application of this principle.

In complement of these policy initiatives, the European Union fourth (1994–1998) and fifth (1998–2002) framework programs of Research and Technological Development included a significant number of projects directed at the various issues relevant for application of prices for the use of infrastructure. The European Conference of Ministers of Transport dedicated its Round Table 118, held in November 2000, to the subject of Interurban Road Tolls (ECMT, 2002).

But on the political front – namely in the European Council and Parliament – the principle of defining user prices based on social marginal cost was not adopted, as politicians know they have other objectives to follow besides maximum economic efficiency.

So, Directive 1999/62 "on the charging of heavy goods vehicles for the use of certain infrastructures" (European Union, 1999), allows the introduction of charges (time dependent) or tolls (distance dependent) to heavy goods vehicles of at least 12 ton gross weight travelling on motorways, stating that "The weighted average tolls shall be related to the costs of constructing, operating and developing the infrastructure network concerned," which shows preference for the (full) cost recovery principle.

However, the concerns of higher marginal costs in the areas of environmental impacts and congestion are also addressed through the possibility of the introduction of limited mark-ups: "Member States may vary the rates at which tolls are charged according to:

vehicle emission classes, provided that no toll is more than 50% above the toll charged for equivalent vehicles meeting the strictest emission standards;
time of day, provided that no toll is more than 100% above the toll charged during the cheapest period of the day.

There is no mention of the safety related external costs in the Directive.

It is worthwhile noting that this Directive does not impose that member countries apply road pricing schemes to Heavy Goods Vehicles on their motorways, but defines some principles and limits should they decide to do so.

Introduction 5

Shortly after, Switzerland, which is formally outside the European Union, but is enclosed in its territory and has very strong economic relations with many of its members, decided to apply from January 2001, a scheme of road tolls to heavy goods vehicles above 3.5 ton gross weight. This scheme went far beyond the EU directive, namely in charging for the use of all roads, lowering the weight limit for application of the tolls, and increasing the differentiation of the toll level according to emission levels.

In September 2001 the European Commission adopted the White Paper on "European Transport Policy for 2010: time to decide" (European Commission, 2001), where the principles of direct payment by the users and of internalisation of external costs are fully supported, but the expression "marginal cost" no longer appears, although the use of price to fight congestion is still clearly advocated.

This White Paper addresses all modes, and in it the European Commission promised to propose in 2002 a directive "to establish the principles of infrastructure charging and a pricing structure for all modes of transport." It was promised that such proposal "which will leave each Member State wide scope in terms of implementation, will include a common methodology for setting price levels which incorporate external costs, and will specify the conditions for fair competition between modes."

Manifestly this has not been possible, mainly because the independent practices of the various modes have led to such disparities that defining a converging path for a consistent policy of infrastructure charging across all modes has proved politically unmanageable in such a short term.

In July 2003, the European Commission presented to the European Council and Parliament a proposal for the amendment of Directive 1999/62, as a result of the requests presented by the European Council meetings of December 2002 and March 2003, as well as to permit a revised framework to several initiatives of member states who were revising their own road pricing systems in accordance with those policy documents mentioned above. This proposal has faced serious difficulties and, after the failure to get approval in the Council in June 2004, there is no clear indication of the likely outcome.

In parallel, Directive 2004/52/EC "on the interoperability of electronic road toll systems in the Community" has been approved in April 2004 (European Union, 2004), establishing a harmonised technological framework that should allow that foreign vehicles being charged in a given country can avoid stopping at toll gates if they have the proper equipment, and contract with a tolled motorway operator in their own country.

This brief survey on the evolution of directly paying for use of the road shows that, even if the general justice of that principle at least for some groups of users is no longer questioned, there are some pending issues preventing a EU-wide 6 JOSE M. VIEGAS ´

adoption of a new directive. Some of these difficulties are directly related with the international dimension of the measures and to the fact that foreign users have little power in influencing national governments towards resolution of the problems by means other than higher pricing.

Heavy Goods Vehicles have long been the preferred starting point for systematic application of road tolls in countries where they did not exist: they cause the most damage to roads and environment, and in many places there is a visible presence of foreign vehicles, which facilitates acceptance, as they also have to pay.

But there are in this debate several conflicting objectives, where naturally the interests and positions of the various countries or user groups may differ:

if road tolls are set very high this can constitute a barrier to free trade, but if they do not properly account for environmental costs there may be an excessive amount of traffic and environmental damage. This has a strong implication on the method for calculation of the economic value of environmental impacts and on the limit of the corresponding mark-ups;
if the motorways around a conurbation are frequently congested and international goods vehicles in transit through the region have no proper alternative, paying a severe congestion mark-up may be seen as constituting an unfair double payment, as the majority of vehicles on those motorways are local and regional private cars. These in turn feel that those lorries are consuming an important part of the available road space. This should have some implication on the possible application of road tolls to private cars, at least in the sections of the Trans European Road Network, where they are rivals in consumption with the trucks.

The study of this matter shows that this is a process dealing with complex problems which has not yet come to a final and stable solution, and where solutions have to be adopted in a step-by-step process, as social, political and technical bases for acceptance if successful transition to the next stage is to be reached.

This books aims to contribute to a better understanding of the difficulties in the multiple dimensions relevant to the problem and of the possible solutions developed to address them.

The rest of the book is organized as follows: Chapters 2–5 treat the various dimensions of interurban road charges and tolls for trucks in a rather general way, whereas Chapters 6–10 deal with the specifics of the European dimension. In Chapter 2 the possible multiple objectives of road tolls are introduced and the corresponding basic solutions presented; Chapter 3 treats the technological and procedure design elements that support the application of those different solutions; Chapter 4 deals with the legal, institutional and contractual issues in the application of tolls; and Chapter 5 is focused on the acceptability issues for introduction of those tolls. Chapter 6 assesses different tolling schemes vis-a-vis the various ` Introduction 7

situations encountered in Europe; Chapter 7 presents the positions and trends in several EU countries and at EU level, Chapter 8 looks at the interoperability issues, discussing their importance and the possibility of stepwise application; Chapter 9 presents an assessment of the regional economic and environmental impacts on the introduction of interurban road tolls for trucks in Europe; Chapter 10 makes a critical look at the evolution of the EU directive on road charges and tolls for trucks, trying to identify the main tensions and possible successful routes. Finally, Chapter 11 concludes with a synthesis of the whole book, presenting a view of the significance of road tolls for trucks in the framework of the process of creation of more efficient pricing schemes in transport systems.

REFERENCES

ECMT European Conference of Ministers of Transport (2002). Tolls on interurban road infrastructure. An economic evaluation. Report of Round Table 118, Paris, 30 November–1 December 2000, ISBN 92-821-1374-4, Paris.
European Commission (1995). White paper on towards fair and efficient pricing in transport: Policy options for internalising the external costs of transport in the European Union, COM (95) 691.
European Commission (1998). White paper on fair payment for infrastructure use: A phased approach to a common transport infrastructure charging framework in the EU, COM (1998) 466 final.
European Commission (2001). White paper on "European transport policy for 2010: Time to decide," ISBN 92-894-0341-1, Brussels.
European Union (1999). Directive 1999/62/EC of the European Parliament and of the Council, of 17 June 1999, on the charging of heavy goods vehicles for the use of certain infrastructures, Brussels.
European Union (2004). Directive 2004/52/EC of the European Parliament and of the Council of 29 April 2004 on the interoperability of electronic road toll systems in the Community, Brussels.
Federal Highway Administration (2004). http://www.fhwa.dot.gov/policy/otps/valuepricing.htm, consulted in 18th June 2004.
Pigou, A. C. (1920). The economics of welfare. London: Macmillan.
Transport for London (2004). http://www.transportforlondon.gov.uk/tfl/cclondon/cc monitoring-2ndreport.shtml, consulted in 22nd June 2004.

2. BASIC ROAD PRICING SOLUTIONS

Heike Link and Louise Stewart-Ladewig

Since Pigou suggested the idea of road pricing in 1920 (Pigou, 1920), the various, partly overlapping or even conflicting objectives of road pricing schemes and the basic charging solutions available have been debated among academics, politicians and the public (see for example: Arnott et al., 1988; Borins, 1988; Goodwin, 1989; Hau, 1992; Newbery, 1989, 1990; Vickrey, 1969; Viegas, 2002; Walters, 1961). Over the last two decades the road pricing debate has intensified because of the growing problems transport policy has had to tackle, for example: curbing congestion, optimising the provision of transport infrastructure capacity, reducing environmental and safety problems and financing transport infrastructure in times of public sector spending limits.

These problems together with technological breakthroughs have led to the introduction of electronic toll collection systems in Norwegian cities such as Trondheim, Bergen and Oslo, on motorways in Italy, France, Spain and Portugal, in the U.S., Australia, Hong Kong and Singapore. Across the member states, a patchwork of different charging schemes and charging levels for heavy goods vehicles (HGV) has evolved (see Fig. 1).

Clarifying the objectives of a road pricing strategy is arguably the most important prerequisite for determining the design of any charging system. For example, if curbing congestion (e.g. demand management) is the objective of introducing road pricing then all vehicles must be charged, and in order to influence the individual driver's behaviour, the system must be able to differentiate charges by time of day. Uncomplicated solutions, such as the sale of permits (for example the Eurovignette), can be sufficient if the charges raised are simply proportional to a predetermined period of time that a vehicle uses the system. Charges that are

Fig. 1. Current Motorway Charges for Heavy Goods Vehicles (as of February 2004). Note: Current road charges can be obtained from national motoring associations or from the website of the national information centres of the European Union "Info Points Europe" (IPE) for example: http://www.eiz-niedersachsen.de/.

intended to relate to the marginal costs generated by each individual trip require a more complex distance based solution using technology such as electronic or satellite based systems.

Given the importance of the objectives behind road pricing for the design of the technical solution, this chapter deals mainly with the objectives of road pricing, and describes basic road pricing charging strategies. This description is aimed at explaining to what extent technical road pricing solutions depend on the objectives of a specific road pricing scheme.

1. OBJECTIVES OF ROAD PRICING

Like pricing everywhere in the economy, road pricing is a method of resource allocation and revenue generation. Depending on the goals to be achieved and constraints to be considered, optimal pricing strategies play a greater role than setting the "right" price. Within the public discussion, objectives of road pricing such as financing infrastructure, internalising external transport costs, creating fair competition between the modes, achieving a modal shift, cost recovery and demand management are frequently mentioned. These diverse, and at times, overlapping objectives of road pricing can be grouped as achieving: (a) allocative efficiency in infrastructure use; (b) allocative efficiency in infrastructure provision and enlargement; and (c) financial viability.

1.1. Efficient Use of Infrastructure (Demand Management)

Socially optimal road use can be theoretically achieved by charging the difference between the marginal private costs of travel and the marginal social costs. This difference is called the Pigouvian charge (Pigou, 1920) and is equal to the marginal external congestion costs. In order for a road pricing solution to help achieve this allocative efficiency of infrastructure use, the charging technology is required to have technical features which enable it to differentiate prices by time of day and/or by roads with/without bottlenecks. A highly flexible technical system can vary the user charge to the actual level of congestion, however, for a real-time congestion charging scheme problems occur because road users do not know ex-ante the price of their trips, meaning the information signal of the charge is not ideal.

Congestion dependent charges can over time lead to a more efficient transport performance or bring about a change in the modal split. Current transport prognosis foresee an increase in the number of vehicles on the road. In order to deal with this increase, better speed management through the use of up to the minute transport data can help keep up a steady flow of traffic and reduce accidents and congestion.

2. INTERNALISING THE EXTERNAL COSTS OF TRANSPORT

The basic principle of the Pigouvian charge for pricing externalities has been further developed within environmental economics and provides the theory of how to charge for transport externalities such as environmental costs (air pollution, noise) and accident costs (the "Polluter Pays Principle, seeBaumol & Oates, 1988). Charges that relate to external costs can provide an incentive to make use of the safest, cleanest modes of transport or the least congested parts of the networks. Figure 2 shows the cost of infrastructure, environmental cost and accident costs per capita for the EU member states (Link et al., 2003).

A comprehensive description and classification of transport costs is given in Link et al. (2000) and can be summarised as follows:

Infrastructure related costs: Costs relating to infrastructure include the costs of planning, construction, inspection and maintenance of roads, the costs

Fig. 2. The Per Capita Costs of Infrastructure, Air Pollution, Global Warming, Noise and User External Accident Costs of Road Transport 1998. Source: Link et al. (2003).

of signalling, lighting, cleaning, snow removal etc. Also related to road infrastructure are the costs of transport police and the interurban road pricing scheme (IRPS) in use. Loss of property value and the splitting of townships and cities through the presence of transport infrastructure present further costs;

Environmental costs: The use of motor vehicles causes environmental costs, for example, health costs resulting from exposure to transport emissions such as air and noise pollution. Additionally, there are costs related to changes in the natural environment that are directly related to transport emissions such as soil and water pollution or damage to forests and other ecosystems. Buildings and monuments important to cultural heritage are damaged through transport related vibration, acid rain or particulate deposition;
Accident Costs: Accident costs include medical costs such as first aid and ambulance costs, emergency services, stationary and out-patient hospital care, athome care provided by local health staff and all the associated medicines, health aids and treatment arising from transport accidents. Employee replacement or retraining costs due to transport accidents, the cost of material damage to vehicles, public and private property such as buildings and transport infrastructure resulting from accidents, the costs of police, insurance companies and justice dealing with the consequences of transport accidents must also be considered. The cost of suffering and grief experienced by accident victims and their family and friends and other risk costs such as the cost of transport risk avoidance are further accident related costs;
Congestion costs: Costs related to congestion include the cost of lost work and leisure time, increased fuel consumption, increased vehicle wear and tear, additional air pollution and associated health related problems and unnecessary

stress and discomfort. Congestion slows the movement of goods and services, adding to the price of products and reducing the competitiveness of business.

As with congestion charging, in order to achieve allocative efficiency of infrastructure use, the economically most efficient first-best pricing principle for environmental and accident costs is marginal cost pricing. Charging the marginal external costs of environmental damage, noise and accidents puts further requirements to the design of road pricing schemes. The charging technology has to be designed in a way that makes it possible to distinguish between vehicle types and their emission levels, between road sections with high and low accident rates or between vehicles with few or many safety features.

There has been no extensive research in the area of quantifying marginal external transport costs to date. A summary of existing empirical estimates for specific countries and modes and guidelines on how to transfer them to other countries/contexts and on how to use them for pricing policy is given in Van der Bossche et al. (2003). There are some cases where marginal costs could be approximated by average variable costs, for example for some environmental cost categories and for infrastructure costs, and indeed infrastructure costs are the main element of many existing road charging schemes (for example in Germany, where however, a full cost pricing principle is applied). In particular the costs of road maintenance are considered to be variable; they are dependant on the number and type of vehicles using the infrastructure. The amount of damage a particular vehicle does to the road it is using depends not only on the road structure and surface but also on the vehicle weight and the number of axles (HRB, 1961). This means that road pricing systems have to be capable of differentiating charges by vehicle type and other vehicle characteristics such as the number of axles.

2.1. Financing Investment in Transport Infrastructure

The need for financing infrastructure construction and/or maintenance of the road network is frequently cited reason for the introduction of road pricing (TIS.pt, 2002). This objective has a theoretical foundation which states that where the revenues from an optimal road user charging scheme exceed the true resource costs, investments in transport infrastructure should be made. In our classification of road pricing objectives this refers to achieving allocative efficiency of infrastructure provision and enlargement. Furthermore, Newbery (1989) and Newbery (1990) show that the revenues from an efficient road pricing scheme would in principle be sufficient to cover the total costs of the road system. Another important practical point is that road pricing can provide a source of revenue, not tied to the annual Government budget, which can be earmarked for use (or partial use) within the transport sector.

Where regional equity is a national objective, road charges can be used to support transport infrastructure programmes in less developed regions. Revenues from road pricing can also be used as a source of funds particularly important for raising finance outside of traditional national budgets. Road charges can also be a source of finance for public budget relief, particularly where roads are supported by the Government budget and there is pressure to reduce overall taxes. Furthermore, in countries such as Italy, Spain, France and Portugal motorway development was only possible with tolling in order to generate the necessary re-financing. The goal of financing was the most important aim of road user charging. Finally, other objectives which are often cited in the political discussion are to create fair competition between the modes and to achieve a modal shift from road to rail and public transport. Both objectives are in principle contained in the overall economic goal of allocative efficiency in the transport sector.

3. CHARACTERISTICS OF ROAD PRICING STRATEGIES

In this section we discuss which characteristics of road pricing systems are required in order to meet the various objectives of road user charging. Before doing so, it should be borne in mind that the move towards more efficient road user charges is a process. Each of the EU Member States has a different set of policies towards the charging of transport, reflecting different national political priorities which have developed over time. Existing instruments include fuel duties, vehicle related fees and taxes, tolls for the use of specific infrastructure segments, distance related motorway charges, vignettes and congestion charges. At the EU-wide level, steps have been taken to harmonise fuel taxes but the current differences remain significant. Annual vehicle taxes also vary greatly between countries, not only in the level of taxation, but also in the criteria on which the tax is based (Kunert et al., 2003).

The introduction of a distance related user charge coupled with vehicle characteristics and in combination with the reduction of other transport taxes is a step towards the user pays principle and the harmonisation of transport taxes and charges in Europe. Keeping existing taxes alongside road pricing schemes might also be justified in some cases because additional goals can be achieved. An example is the annual circulation tax which, if the tax is appropriately differentiated by emission and safety standards, can be used as an incentive to steer the purchasing behaviour of consumers. Further examples are accident insurance systems and parking fees.

4. THE COSTS TO BE INCLUDED AND THE CHARGING PRINCIPLE

As indicated in Section 1.1, an efficient road user charge would be equal to the marginal external costs of each trip. In practice, however, existing charges in different countries are based on different charging principles (marginal cost-based schemes versus average cost schemes) and include different cost categories. For example, the German HGV charge is based on the distance travelled, the number of axles and the emission class of vehicles over 12 tonnes permissible weight, while in Austria the distance related HGV charge is raised on vehicles over 3.5 tonnes permissible weight and does not include an environmental cost element. At the European policy level different charging principles have been proposed in various documents. The White Paper from 1998 (EC, 1998) suggests short-run marginal cost pricing as the leading charging principle. The White Paper from 2001 (EC, 2001) formulates more generally that "the charge for using infrastructure must not only cover infrastructure costs, but also the external costs, that is, costs connected with accidents, air pollution, noise and congestion." The HGV charging directive (EC, 1999) was based on full infrastructure costs, and the amended HGV charging directive proposal (EC, 2003) suggests that road user charges be based on full infrastructure costs and external accident costs.

At present, the charging of vehicles for road infrastructure use is still regulated by the EU Directive on the charging of heavy goods vehicles over 12 tonnes for the use of certain infrastructures (EC, 1999). In accordance with this directive, charges can be raised on motorways and highways with four or more lanes. User charges should be equal to the allocated average costs of the construction, the maintenance and the operation of the road network. Exceptions can be made for bridges, tunnels and alpine passes. The directive has received several criticisms, such as the use of an average cost principle and the exclusion of external costs from the charge. These criticisms have partly been addressed in the proposed amendment of the directive (EC, 2003). The proposed amendment retains the average cost principle but requires additional charging for external accident costs (for a more detailed discussion of the directives see Chap. 10).

4.1. The Vehicle Types to be Charged

In a first-best, welfare optimal solution, all vehicles would be charged for the marginal costs generated by the trips they make. This requires advanced technological solutions where electronic toll collection systems are capable of differentiating charges by the distance travelled, the vehicle characteristics such as weight, number of axles, emission characteristics etc., the time of day, the actual level of congestion and so on. However, there might be good reasons to introduce road charging for HGVs as a first step. Firstly, there are practical grounds for this implementation path. Gaining experience with both the software and hardware components of advanced technological solutions, checking the reliability of the system and the dependability of enforcement measures are good arguments for introducing IRPS to a smaller collective of vehicles rather than the whole fleet.

The recent experience with the German IRPS, which has been postponed several times due to technical problems, reinforces this argument. Secondly, HGVs cause higher environmental costs per km driven than passenger cars (see Fig. 3) and the highest damage costs to the road infrastructure, justifying the introduction of road pricing for these types of vehicles. Thirdly, European transport policy pertaining to road user charges is focused on commercial vehicles in interurban transport, leaving the introduction of road pricing for passenger cars and for urban areas to the discretion of the member states. The main reason for this is because of the crossborder characteristics of road haulage, different national approaches regarding the charging principle and the technical solution have negative consequences on the competition situation between road hauliers from different countries and cause efficiency losses.

However, it should be emphasised that in the mid- or long-run, all vehicles should be charged. As Fig. 3 indicates, although passenger cars have lower average air pollution costs than HGVs they cause the highest total air pollution costs because of the size of the fleet and the high total mileage. Furthermore, congestion charging will only be successful if all vehicles are included within the scheme.

4.2. Distance Versus Time Related Charging Schemes

There are two basic charging principles central to the system solution chosen: the charge raised can be based either on the distance travelled or the time spent(driving or not driving) on a defined road network or within a defined geographical area. If the goal of road pricing is to recover the marginal costs of road use, then the system chosen must be capable of recognising vehicle characteristics and registering the number of kilometres travelled on the infrastructure or in the area where road charges are raised. The technology necessary to run a distance related charging system can be based on dedicated short range communication (DSRC) or a global navigation satellite system combined with mobile communication (GNSS/CN). If transport policy designates the introduction of vehicle charges for access to a system for a predetermined period of time (day, week, year) then a system solution based on the sale of permits – such as the Eurovignette system – can satisfy this

¹⁷ Fig. 3. Average and Total Air Pollution Costs of Passenger Cars and Heavy Goods Vehicles 1998. Source: Link et al. (2003).

requirement. Time-based systems have been used in the past because there was no viable alternative distance based charging technology.

4.3. Price Variation

The capability of the technical system to charge vehicles according to different criteria is one of the most important features of any road pricing scheme. Price variation can based on vehicle types, vehicle characteristics, sections of infrastructure, geographical areas and time of day. Existing HGV charging schemes distinguish between vehicles with different numbers and configurations of axles, by vehicle weight, by the use of trailers and/or by emission classes when determining vehicle charges. A price variation that relates directly to the vehicle type or class is imperative if charging for infrastructure damage and environmental cost is to be equitable.

Road user charges should reflect the cost of repairing damage to transport infrastructure caused by a specific vehicle. The axle weight and number, and the suspension type of HGVs have a major influence on the extent and the period of time between necessary road repairs and maintenance. Generally, larger and heavier vehicles cause more road infrastructure damage than comparably smaller vehicles. However, a shift in the vehicle fleet towards smaller vehicles is not desirable and should not result from a charging scheme. Small or medium-sized lorries carry fewer tonnes of goods than HGVs (not only in absolute terms but also in relation to their deadweight), and contribute, more than in proportion to the number of tons carried, to the increase in congestion on roads, to the amount of transport emissions and to the number of road accidents.

To pay for the environmental damage caused by an individual vehicle, road user charges should reflect progress in clean vehicle technology and the charging structure should depend on the engine type, the energy source and emission standards (combined with other factors such as level of congestion, geographical area etc.). Relating the level of tolls to the vehicle emissions should have a positive impact on the total emissions of the fleet. For example, estimates made at the time of introduction of the new HGV road pricing scheme in Germany, where the kilometre charge is linked to the emission class of the vehicle, indicate a drastic reduction in EURO III vehicles and an increase of EURO V vehicles in excess of 50% by 2010 (Rothengatter & Doll, 2001).

If road pricing policy requires tariff modulation according to criteria which can change over time, specific sections of infrastructure or geographical area, the technical system must have the flexibility to enable this. Each extra pricing variation a system incorporates adds to the complexity and inevitably to the cost of the technical solution chosen.

4.4. Charging Technology

The technical IRPS systems available can be classified in three fundamental groups. Firstly, a system based on the sale of permits, such as the Eurovignette, which allows a specific vehicle to use a defined network of roads for a specific period of time. Secondly, a technical system based on dedicated short range communication (DSRC) where each vehicle has a transponder programmed with data relating to the charging structure. The transponder communicates with road side infrastructure allowing payment of tolls to be carried out either for a specific segment of infrastructure such as a bridge or motorway section or for the whole network where charges are raised, such as the HGV road pricing system in Austria. The third technical possibility is the use of a global navigation satellite system combined with mobile communication (GNSS/CN). These technical solutions will be described in Chap. 3: Technical and procedure design elements. Regardless of the charging principle to be exercised, each of the three fundamental IRPS systems can be used for a road pricing scheme where charges are adapted to the individual vehicle characteristics.

5. BASIC SOLUTIONS FOR INTERURBAN ROAD PRICING AND THEIR SCOPE

IRPS solutions can broadly be classified into either distance related or access and/or dwell time related schemes (TIS.pt, 2001). The basic types of IRPS to be discussed in this section are derived from these two principles combined with further variables such as the infrastructure subject to a charge or the vehicle differentiation.

5.1. Distance Dependent Network Pricing (NET)

NET (distance dependent network pricing) is defined as being a pricing regime based on the kilometres driven on a network of interrelated motorways of the same hierarchy level (TIS.pt, 2001). The vehicle charge is raised for the use of a specific segment of infrastructure and may be differentiated according to vehicle type and further parameters.

NET pricing is a classical pricing solution for funding investment in specific transport infrastructure projects and has been successfully used in many countries. The NET solution is also relevant if the aim of road pricing is to raise revenues outside of the traditional Government budget for the maintenance and upgrading of an existing road network (Viegas, 2002). NET pricing, like any other form of distance related pricing, is a pricing strategy that achieves an efficient use of infrastructure, provided that the charging principle is based on welfare optimisation principles. It gives an incentive to the vehicle owner to reduce the number of trips and increase the effectiveness of trips made on the network where charges are applied. If fewer trips are made, the reduction in the number of vehicles on the road leads to less congestion, fewer accidents and a decrease in overall transport emissions (ECMT, 2003). If road charges are structured so that less polluting vehicles pay less than those with a higher rate of emissions, this may, over time, move the structure of the vehicle fleet towards low emission vehicles and reduce total transport emissions. However, to reduce the overall external costs from road transport or to achieve the political goal of changing the modal split, a DAREA solution (distance dependent area pricing), where the use of all roads within a specific country or other geographical area entails a charge, has been identified as being more effective (TIS.pt, 2002).

The technical tolling system of NET can differ between manual tolling using toll houses, electronic tolling using dedicated short range communication systems (DSRC) and solutions based on satellite and mobile communication technology (GNSS/CN). NET solutions presently in use rely on manual toll collection at toll plazas or dedicated short-range communication systems (DSRC) for electronic payment of road tolls. DSRC is based on a standardised, well understood technology and has proven to be reliable for distance related charging in many countries world wide. The technical system uses road side beacons that communicate with a transponder inside the vehicle. Transponders can be programmed to transmit vehicle data such as vehicle weight or emission class and allow a price variation between vehicle classes (ASFINAG, 2004). Tariff modulation by time of day using a DSRC technical system is presently in use in Portugal, similar systems could be developed to include tariff variation relating to the actual level of congestion, for example through price computation based on relative speeds (Viegas, 2000).

The advantages of a DSRC technical system for NET and other pricing schemes must be weighed up with the limitations of the system. In particular, a DSRC system cannot be easily extended to include other roads especially as a temporary measure, for example, to discourage traffic diversion to secondary roads. If this level of pricing flexibility is required, then a satellite based system must be considered. These systems are based around the use of a technically sophisticated on board unit (OBU) contained within the vehicle. The OBU uses satellite positioning to locate the vehicle on the road network and mobile communication to send this information and other payment relevant data to a central charging centre, eliminating the need for road side equipment or toll plazas (Toll Collect, 2004). The German road pricing system in construction is an example of a satellite based NET pricing system.

The problems associated with manual payment at toll plazas such as congestion and increased accident prevalence are well documented. These problems can be somewhat avoided by a high level of vehicles using electronic payment, however, there will always be a considerable proportion of non-equipped users if a vehicle transponder is not mandatory. These vehicles must stop at the toll gate where the vehicle characteristics relating to the charge are verified and payment of the toll takes place. One clear advantage of the toll plaza system is the relative ease of compliance checks as every vehicle passing the toll area is automatically checked (TIS.pt, 2003).

5.2. Distance Dependent Area Pricing (DAREA)

DAREA (distance dependent area pricing) is defined as a pricing scheme based on the number of kilometres driven within an area limited by a certain perimeter (TIS.pt, 2001). The fee is proportional to the distance driven on all roads and differentiated according to vehicle classes and weights. Other parameters, such as the time of day or class of road etc. may be used in conjunction with DAREA pricing (TIS.pt, 2003). A current example of a DAREA pricing solution is road pricing for HGVs for the use of all roads in Switzerland.

DAREA pricing is a classical strategy for countries (or geographical areas) with a high density of road infrastructure. Motorway network charging of HGVs in these countries would presumably lead to a high rate of detour traffic onto secondary and urban roads, inevitably causing higher rather than lower transport related environmental and accident costs. DAREA is effective both for promoting an efficient use of roads and for raising revenues either for transport infrastructure construction or maintenance or simply flowing to the general budget. A DAREA system can also be used to monitor the flow of traffic into and out of specific areas. If charges are dependant on the rate of vehicle emissions, this can be seen as a step towards the internalisation of external environmental costs. Similar to NET pricing a DAREA solution, using a tariff system which takes the emission class of the vehicle into account, can be used over time to alter the structure of the vehicle fleet towards more fuel efficient, low emission vehicles. As in any kilometre based charging system, unnecessary trips will be reduced and the effectiveness of commercial transport increased. This inevitably leads to a reduction of congestion, transport emissions and transport related accidents (TIS.pt, 2003). A road pricing strategy that charges for the use of all roads within a country (such as in Switzerland) increases the cost of road transport considerably. These additional costs can be mitigated by a reduction of other transport related taxes and charges or they can be used intentionally to force an improvement in the present overall logistical efficiency or changes to the modal split – away from commercial road transport to other modes (ECMT, 2003).

The technical system chosen must always relate highly to the specific aims of the road pricing scheme. DAREA is, by definition, a distance related scheme requiring technical systems such as dedicated short-range communication systems (DSRC road side based system) or autonomous systems using satellite and mobile communication technology (GNSS/CN vehicle based system). The vehicle based system has the advantage of being more flexible, the user charge can theoretically be modulated via the OBU to reflect actual road and environmental conditions or variation in the perimeter of the geographical area where the charge is raised.

Non-equipped users are defined as vehicles without a transponder, where a road based system is in place, or vehicles without an OBU using a vehicle based system. In both cases they pose a problem for the operation the technical system. As the vehicle charge is distance related, tachograph readings must be taken at the system entry and exit points and the appropriate user charge calculated and paid before leaving the system. Another possible option for transit vehicles is to give the route to be taken, and based on this information, pay the appropriate charge in advance of the trip. As with all distance related payment systems, the problems related to non-equipped users increase with the general complexity of the tariff scheme.

5.3. Driving Permit or Area Licensing (PERM)

PERM (driving permit or area licensing) is defined as being a charge which is levied for permission to drive within a certain perimeter and within a certain time period, and which is differentiated according to vehicle classes (TIS.pt, 2001). A typical example of a PERM pricing scheme is the Eurovignette, a time and vehicle related permit (not distance related) which must be purchased by HGVs using motorways in Belgium, Denmark, Luxembourg, the Netherlands or Sweden. Effectively a PERM solution involves the vehicle operator purchasing a permit and placing this on display in the vehicle before entering the charge area.

The great advantage of a PERM solution is that it can be launched with no road side or vehicle based technical system. System costs are low compared to all other solutions as existing infrastructure and organisations (such as petrol stations) can be used for the sale of the permit. The system is simple, understandable and easy for the user to handle. However, PERM has very little impact on traffic flow, as long as the user charge is not excessively high. Because the price of permits is not distance related, the average cost per kilometre will decrease with a higher transport intensity. Since there is no relationship to the actual amount of use of the infrastructure PERM is not suitable to promote allocative efficiency in using

Solution	Description	Examples of Countries where Solution is Used
NET: Distance dependent network pricing	Traditional form of motorway tolling systems using manual and electronic road side based toll collection systems Motorway tolling using electronic vehicle based tolling systems	France, Italy, Spain, Portugal Germany (in construction)
DAREA: Distance dependent area pricing	Tolling of all roads within a specific geographical area	Switzerland
PERM: Driving permit or area licensing	Purchase of time related permit for system use	Denmark, Benelux, Netherlands, Sweden (Eurovignette) Bulgaria, Czech Republic, Slovakia (national vignettes)

Table 1. Summary of the Use of Inter-Urban Road Pricing Solutions.

roads. Even when the level of the PERM charge is structured by vehicle emission categories it has proven to be insufficient to induce changes to the composition of the vehicle fleet and thus significantly reduce emissions. Also, PERM has a lower revenue generation potential than a DAREA or NET solution. Furthermore, the PERM solution is inflexible and cannot be modulated to reflect the actual road conditions (TIS.pt, 2003) (Table 1).

5.4. Road Pricing Schemes with Different Incidence

5.4.1. Passage tolling (PAS)

PAS is defined as the levy of a toll for the use of a particular infrastructure segment such as a motorway section, bridge or tunnel. Tolls are differentiated according to vehicle classes. Passage tolling is mostly charged for (re-)financing construction, operation and maintenance of the infrastructure subject to the fee. A recent example of PAS tolling is the charge raised to use the Great Belt Link in Denmark. A PAS solution is not suitable if the charging aim is to reduce congestion or vehicular emissions. No change to the modal split or the structure of the fleet can be expected with PAS. The technical requirements are similar to those for NET pricing: toll plazas with either manual or electronic payment. Passage tolling is generally easily understood and well accepted by transport users especially when related to a new infrastructure segment with a non-tolled alternative route available.

5.4.2. Cordon pricing (COR)

Cordon pricing or access related area tolling is defined as being a charge, differentiated according to vehicle classes, raised when entering a cordon around a defined area or city (TIS.pt, 2001). The charge is paid when entering the defined area and is independent of the type of infrastructure used or kilometres travelled within the cordon. An example of COR is London City Congestion Charging and the charging schemes in Norwegian cities. A variation of this solution is COR dwell time related area tolling, based on the time spent within the defined area (which also falls within the PERM definition of tolling). To calculate the time spent in the tolling area the exact time of entry into and departure from the area must be registered/recorded. This type of COR pricing is often compared to car parking fees and is planned to be used, for example, within the Stockholm environmental charging scheme.

The basic COR solution has been introduced in a number of cities worldwide and is considered to be a successful method to alleviate congestion in city centres and promote the use of public transport, for example London Congestion Pricing (LCP, 2004). Raising of revenues can also be a goal of introducing COR pricing, as was the case with the introduction of COR in Norwegian cities. Simple vehicle identification technology is adequate to successfully run a COR solution.

REFERENCES

Arnott, R., de Palma, A., & Lindsey, R. (1988). Schedule delay and departure time decisions with heterogeneous commuters. Transportation Research Record, No. 1197, 56–67.
ASFINAG (2004). Toll system for trucks and buses. Internet document http://www.asfinag.at/ accessed 12.05.2004.
Baumol, W. J., & Oates, W. E. (1988). The theory of environmental policy (2nd ed.). Cambridge: Cambridge University Press.
Borins, S. F. (1988). Electronic road pricing: An idea whose time may never come. Transportation Research, 22A(1), 37–44.
TIS.pt (2001). Possible designs of inter-urban road pricing schemes for heavy goods vehicles: DESIRE (DESigns for Inter-urban Road pricing schemes in Europe) Deliverable 2. Project funded by the European Community under the 'Competitive and Sustainable Growth' Programme (1998–2002), Lisbon.
TIS.pt (2002). Case study results, analysis and reference scenario: DESIRE (DESigns for Inter-urban Road pricing schemes in Europe) Deliverable 3. Project funded by the European Community under the 'Competitive and Sustainable Growth' Programme (1998–2002), Lisbon.
TIS.pt (2003). Identification of best functional schemes and implementation issues: (DESigns for Inter-urban Road pricing schemes in Europe) Deliverable 5. Project funded by the European Community under the 'Competitive and Sustainable Growth' Programme (1998–2002), Lisbon.
EC (1998). Fair payment for infrastructure use: A phased approach. White Paper, COM (1998) 466/4, Brussels.
EC (1999). The charging of heavy goods vehicles over 12 tonnes for the use of certain infrastructures. Council Directive 99/62/EC, European Commission, Brussels.
EC (2001). European transport policy for 2010: Time to decide. White Paper, COM (2001) 370, from 12.9.2001. European Commission, EUR-OP, Luxembourg.
EC (2003). Proposal for a Directive of the European Parliament and of the Council amending Directive 1999/62/EC on the charging of heavy goods vehicles for the use of certain infrastructures. COM(2003) final, Brussels.
ECMT (2003). Reforming transport taxes. European Conference of Ministers of Transport. Paris: OECD publication service.
Goodwin, P. B. (1989, October). The rule of three: A possible solution to the political problem of competing objectives for road pricing. Traffic Engineering and Control, 29(10), 495–497.
Hau, T. D. (1992). Economic fundamentals of road pricing: A diagrammatic analysis (pp. 1–99). World Bank Policy Research Working Paper Series WPS 1071, December, Washington, DC: World Bank.
HRB (1961). The AASHO-road test History and description of project. Special Report 61A. Highway Research Board, Washington, DC.
Kunert, U., Kuhfeld, H., Bach S., & Keser, A. (2003). Die Abgaben auf Kraftfahrzeuge in Europa. Berlin: Sonderheft 174 des DIW Berlin.
LCP (2004). Transport for London. London congestion charging: Six months on Internet document http://www.tfl.gov.uk/tfl/downloads/pdf/congestion-charging/cc-6monthson.pdf accessed 13.05.2004.
Link, H., Stewart, L., Maibach, M., Sansom, T., & Nellthorp, J. (2000). The accounts approach. UNITE (UNIfication of accounts and marginal costs for Transport Efficiency) Deliverable 2. Project funded by 5th Framework RTD Programme. ITS, University of Leeds, Leeds.
Link, H., Stewart-Ladewig, L., Nellthorp, J., Bickel, P., Schmidt, S., Loog, T., Lass, K., Villemi, M., Himanen, V., Idstrom, T., Certan, C., Van der Bossche, M., & Gouw, N. (2003). Future approaches to accounts. UNITE (UNIfication of accounts and marginal costs for Transport Efficiency) Deliverable 14. Project funded by 5th Framework RTD Programme. ITS, University of Leeds, Leeds.
Newbery, D. M. G. (1989, May). Cost recovery from optimally designed roads. Economica, 56, 165–185.
Newbery, D. M. G. (1990, Summer). Pricing and congestion: Economic principles relevant to pricing roads. Oxford Review of Economic Policy, 6(2/Special Issue on Transport), 22–38.
Pigou, A. C. (1920). Wealth and welfare. London: Macmillan.
Rothengatter, W., & Doll, C. (2001). *Anforderungen an eine umweltorientierte Schwerverkehrsabgabe f ¨ur den Stra*ßeng ¨uterverkehr. Forschungsbericht 200 96 130 (UBA-FB 000217). Texte des Umweltbundesamtes 57/01, Berlin.
Toll Collect (2004). Truck toll system. Internet document http://www.toll-collect.de accessed 12.05.2004.
Van der Bossche, M., Certan, C., Veldman, S., Nash, C., Johnson, D., Ricci, A., & Enei, R. (2003). Guidance on adapting marginal cost estimates. UNITE (UNIfication of accounts and marginal costs for Transport Efficiency Deliverable) D15. Project funded by 5th Framework RTD Programme, NEI, Rotterdam.
Vickrey, W. S. (1969). Congestion theory and transport investment. American Economic Review, 59 (Papers and Proceedings), 251–260.
Viegas, J. M. (2000). A portagem como instrumento da gest ˜ao rodovi ´aria: Princ´ıpios base e desafios da convers ˜ao a esta nova miss ˜ao (Road tolls as in instrument for road traffic management:
Principles and challenges of conversion to this new mission), Proceedings of the 1st Portuguese Road Congress, CRP, Lisbon.
Viegas, J. M. (2002). Tolling heavy good vehicles on European roads: From a diverse set of solutions to interoperability? Second Seminar of the IMPRINT-EUROPE Thematic Network "Implementing Reform on Transport Pricing: Identifying Mode-Specific Issues" Brussels 14/15 May 2002.
Walters, A. A. (1961). The theory and measurement of private and social cost of highway congestion. Econometrica, 29(4), 676–697.

3. TECHNICAL AND PROCEDURAL DESIGN ELEMENTS

Andrea Felix

This chapter presents an overview of technical and procedural design elements for tolling systems.

Technical and procedural design elements have a strong correlation. The main design elements for tolling systems can be split in following areas:

General design elements.
Basic general design elements.
Tariff class determination and tariff modulation.
Technical options for Equipped users.
Technical options for Non-Equipped users.
Enforcement concept and technical approach.

Each of the mentioned areas is displayed with all relevant options in an over-view table and described with additional comments and details in sub-sections.

1. GENERAL DESIGN ELEMENTS

General design elements are triggered by legal, political and institutional regulations and have a basic impact on design and technical approach for any kind of charging scheme.

In the following table the general design elements are summarised, showing the possible complexity of Electronic Fee Collection Schemes (EFC Schemes) (Table 1).

Interurban Road Charging for Trucks in Europe Research in Transportation Economics, Volume 11, 27–42 Copyright © 2005 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0739-8859/doi:10.1016/S0739-8859(04)11003-2

28 ANDREA FELIX

Table 1. General Design Elements.

With respect to the technical and procedural design of an EFC Scheme, in particular the characteristics "Type of road" and "Vehicle type charged" have a very strong impact, which influences the design of the scheme from the scratch. The "User group charged" is of secondary importance.

2. BASIC TOLLING DESIGN ELEMENTS

The two most important basic tolling design elements are:

charging principle; and
traffic/road situation.

The Charging principle defines the basic approach of the charging scheme. The two most common options are "Distance related" and "Access and/or dwell time related" schemes. Both of these self-explaining characteristics can be split in different alternatives, reflecting the objectives of the charging scheme. "Distance related schemes" basically point to the efficient use of infrastructure and the internalisation of costs. "Access and/or dwell time related" are used for the reduction of traffic flow and/or the financing of infrastructure.

The Traffic/road situation defines the impact to traffic-flow and/or the necessity of space and acceptance for toll-plazas. In Table 2 the most common alternatives for the two characteristics are summarised:

Depending on charging principle and road/traffic situation, specific technologies can be favoured as they are better suited to cover the requirements.

Table 2. Basic Design Elements.

2.1. Charging Principle

Distance related charging schemes apply to the registration of the driven kilometres in a defined area or network of roads of the same hierarchy level. The following different schemes are known:

Distance related area tolling is based on the amount of kilometres driven within the area. The fee is levied according to a kilometre charging concept, where (for each vehicle) the fee is proportional to the distance driven on all roads within the defined area.
Distance related network charging is based on the kilometres driven on a network of interrelated motorways of the same hierarchy-level. The systemlayout of network tolling can differ between open, closed and half-open systems.

Access and/or dwell time related schemes apply to the registration of the access and/or to the dwell time in a defined area. Following different schemes are known:

Access related area tolling/cordon pricing is charged for entering a cordon around a defined area or city. But no fee is collected for trips inside the area.
Dwell time related area tolling is based on the time spent within the defined area. To calculate the time spent in the tolling area, the exact time of entry in and departure from the area must be registered/recorded.
Passage tolling is levied for the use of a particular object like a motorway, bridge or tunnel.
Area licensing is levied for the permit to drive within a certain perimeter and within a certain time period (day, week, month, year). For all those who drive inwards from the outside, it can be seen as similar to cordon pricing, but with the

30 ANDREA FELIX

additional equity dimension that also those that live within the perimeter have to pay (possibly less) to drive in that area.

2.2. Traffic/Road Situation

Traffic and road situation is a very important design element because it has a strong impact on the technology and procedures.

2.2.1. No Impact to Existing/Nominal Geometrical Profile of Road and Traffic Flow is Accepted

It is a very basic decision if the charging scheme has to be operated without any impact to traffic flow. A multi-lane free flow system has the advantage of a higher security for the traffic flow because there is no need to reduce speed, to choose a particular lane or even to stop at a tolling point. On the other hand, a free flow system causes higher requirements towards road-side equipment (RSE), handling of non-equipped users and leads to different enforcement concepts.

In particular on existing motorways it is quite difficult to build new toll plazas or even to make any changes on the existing geometrical profile of the road. The availability of space is of importance in the context of new road projects. Lack of space can be caused by topographical reasons, dense built-up area or high cost for land purchase.

2.2.2. Impact to Existing/Nominal Geometrical Profile of Road and Traffic Flow is Accepted

Conventional toll plazas or similar tolling points require extensive space and, depending on their share, handling of non-equipped users also requires a much higher number of employees.

Equipped users are driving in electronic charging lanes, enforced by barriers or video systems. Although this is a non-stop charging procedure, sometimes these lanes still require a speed reduction and therefore influence the throughput of the road section.
For Non-equipped users the implementation of automatic and manual charging lanes or alternative log-in facilities must be foreseen.
Automatic charging lanes normally are equipped with toll machines accepting and processing practically all forms of payment such as coins, tokens, banknotes, magnetic and chipcards. Manual charging lanes have to be manned.

Alternative log-in facilities can consist of:

service point with log-in terminals or automates;
call centre service;
log-in via internet.

3. TARIFF CLASS DETERMINATIONS AND TARIFF MODULATION ELEMENTS

The second very important impact to technology is the determination on tariff classes and the option of modulation of the tariffs.

Comparable with the general design elements, both characteristics are driven by and based on political and policy decisions (Table 3).

3.1. Tariff Class Determination and Tariff Modulation

3.1.1. Vehicle Characteristics

The determination of the tariff class normally is based on the vehicle characteristics or combination of several characteristics (see Table 4):

3.1.2. Tariff Modulation

It is quite important to decide whether a modulation of tariffs within the same vehicle class is foreseen or not. Tariff modulation will place higher requirements on the technical performance of equipments both inside vehicles and on the road side. Possible parameters could cover (see Table 5):

Table 3. Tariff Class Determination and Tariff Modulation Elements.

32 ANDREA FELIX

Table 4. Vehicle Characteristics.

Vehicle Characteristics	Comment
Vehicle type	The most common gradation bases on the vehicle type.
Emission class	The emission class can only be determined by verification of the vehicle registration papers or the emission class certificate of the vehicle. Nevertheless the consideration of the emission class will become more and more important.
Weight of vehicle	Maximum permissible weight can only be determined from the vehicle registration papers. Determination of actual weight is technically feasible, even in motion, but is quite complex.
Number of axles and/or tires	Easily determinable even if the vehicle is in motion.
Vehicle dimensions	Easily determinable even if the vehicle is in motion.
Trailer status	Regarding trailers, it is important to distinguish whether a trailer itself is subject to fee or there is a change of tariff class for the vehicle hauling it. Easily determinable even if the vehicle is in motion.

Table 5. Tariff Modulation Elements.

Possible Parameters	Comment
Road infrastructure	Hierarchy level/Quality of road.
	The tariff level for a certain class might differ according to the hierarchy
	level/quality of the road used.
	Construction, operation and maintenance costs of road.
	Costly roads with lots of bridges and tunnels might be charged at a higher
	tariff level than normal roads.
Traffic	Depending on the time of use e.g. during peak hours or congestion the
situation/Demand	tariff might be higher than the standard tariff.
Consumption quantity	On the basis of consumption quantity by the same vehicle it might be
	appropriate to offer discounts or levy an extra-charge for the multiple-use
	of a service. This can be based on the number of passages or on the amount
	of kilometres driven.

4. TECHNICAL OPTIONS FOR EQUIPPED USERS

Equipped users are considered to be vehicles fitted with an electronic recording device, the On Board Unit (OBU). Technical performance, design, operation and handling of the OBU depends on the performance of the system operator and technological approach for the electronic fee collection (Table 6).

Table 6. Technical Options for Equipped Users.

4.1. Localisation of Vehicle and Initiation of Charging Process of Equipped Users

For equipped users the initiation of the charging process can be initiated either by road side infrastructure or by an on board based system with additional equipment for positioning and data transfer/data exchange.

4.1.1. Road side Infrastructure/Infrastructure External of Vehicle

The term of road side infrastructure covers a large range of concepts from dedicated Road Side Equipment (RSE) for electronic fee collection purpose as far as global infrastructure for positioning (GPS) and communication (CN, Cellular network).

Dedicated Road Side Equipment (RSE) systems uses the so-called Dedicated Short Range Communication (DSRC) to detect the position of the vehicle or more precisely the position of the OBU and for the data transfer. The DSRC air-link can be based on Radio Frequency or Infrared and perform one way (read only) or two way (read and write) communication. For the drafting of the scheme the detailed technical performance is not of relevance, it's just important to know that construction at the road side is necessary.

Electronic fee collection systems using GPS and CN such as GSM or TETRA and in the near future GPRS and UMTS are often labelled as autonomous systems and use already existing global infrastructure. The implementation of this infrastructure is not part of the EFC project but its utilization could be charged in the future.

34 ANDREA FELIX

4.1.2. On Board Based System

Future OBUs may be much more sophisticated and use moving sensor technology to detect the distance driven and additional equipment for positioning and data transfer. Vehicle positioning and data transfer are supposed to use the same technology as mentioned above for the autonomous systems.

4.2. Classification of Equipped Users

The main issue of the classification of the users is the allocation of a vehicle to the correct tariff class. For Equipped users all three common methods of registration of the vehicle characteristics are feasible (see Table 7).

Table 7. Options for Classification of Equipped Users.

Option	Remarks
Centrally stored vehicle characteristics	The relevant vehicle characteristics can be claimed via application.
Vehicle/Tariff class is	If original vehicle documents are required (e.g. for
determined from	determination of relevant weight or emission value) the
information stored in the	application has to be done at manned service centre/point of
central database	sale of system operator.
	Otherwise the application can also take place
	Via on-line internet
	By mail/fax
	by contacting call-centre
Claimed vehicle characteristics	Common ways to determine tariff class in OBU are:
Vehicle/Tariff class is determined from	OBU related to tariff class
information stored in the	Each tariff class has a specific OBU
OBU	Switch on OBU to define tariff class, e.g. number of axles; operated by driver
	Software parameter
Measured vehicle characteristics	Several technologies exist to measure dimensions and weight of vehicles:
Vehicle/Tariff class is	inductive/traffic loop
determined from physical	weight in motion devices
characteristics measured at	treadles
the moment of passing	scanning and sensor devices
charging point	radar systems

Table 8. Options for Charging Equipped Users Via Central Account.

Options	Example
Direct debiting by the operator, e.g. monthly invoices.	Heavy Goods Vehicles Fee in Switzerland (LSVA)
A registration and verification of the address of the user is necessary.
Charging the fee/toll to petrol-, credit- or debit-card of the user.	Heavy Vehicles Toll Austria (GO-Maut)
A registration and authorisation of the chosen means of payment is necessary. Invoicing is done by the card-issuer.
All equipped users must provide a bank account number, where the value of the monthly invoice is automatically deducted	Via Verde in Portugal
from. Initiation of the OBU is made by the user himself, in an ATM machine, using
the bank card related to that bank account.

4.3. Charging Equipped Users

For debiting to and payment by equipped users either a central account or an on board account can be considered.

4.3.1. Central Account

The most convenient way of charging is the periodic invoice of the amount due to the operator. For post-payment all kind of central accounts can be used. For debiting, basically three options are available (see Table 8).

Table 9. Options for Charging Equipped Users Via On-Board Account.

Options	Example
The use of a chipcard with a stored value of units. The chip card has to be fitted into the OBU. To reload units, the chip card has to be removed and new credits are loaded to the chip card again.	City Access Tolling System Singapore
The credits are stored straight directly on the OBU. To reload units, the OBU has to be removed from the vehicle and new credits are loaded again on the OBU.	Heavy Vehicles Toll Austria (GO-Maut)

36 ANDREA FELIX

4.3.2. On-Board Account

An on board account presumes pre-payment. The user has to "load" the account on his OBU before using the service. Accepted means of payment are defined by the operator of the system. Basically two options of on-board account exist (see Table 9).

5. TECHNICAL OPTIONS FOR NON-EQUIPPED USERS

In Table 10, technical options for detection, classification and charging of Nonequipped users are summarised.

5.1. Initiation of Charging Process of Non-Equipped Users

The charging process for Non-equipped users can be initiated either by road-side infrastructure with/without direct intervention of the user or by the user himself using check-in/check-out facilities.

5.1.1. By Road Side Infrastructure

Also Non-equipped users can be automatic registered from road side equipment and within this process the charging is initiated. The most common technology for this kind of non-stop charging is Automatic Vehicle Identification (AVI) with License Plate Reading.

Table 10. Technical Options for Non-Equipped Users.

The registered use of the scheme is charged to an existing account that the user must have previously established with the operator. No account or no active checkin (see below) of the user will be followed by a corresponding enforcement process for the violation.

The second, more common initiation of charging process by RSE for nonequipped users is related with a stop of the vehicle at the charging point.

This is the classic way conventional toll-plazas are designed. The user initiates the charging on the spot using additional/dedicated Road Side Equipment (RSE).

Non-manned registration at automate or self-service terminal.
Manned registration at tollbooth/counter.

In specific cases the user has to leave his vehicle to initiate the charging process. For example, in the Swiss LSVA scheme non-equipped users have to declare actual kilometre-reading, trailer-status and wished means of payment each time they enter Switzerland. The dedicated automates are located at easily accessible location at the border-stations, but can not be operated without leaving the vehicle.

5.1.2. By User of the Charging System

The initiation of charging by the user differs from the road-side infrastructure based process in particular from the option to use additional facilities than those offered on the spot/on the road-side (e.g. automates). These are mainly internet and call-centre services, which can also be used by other personnel than the driver of the vehicle and independently from the exact time of use of the system.

Normally the registration/check-in has to be done before using the system. In a few cases a registration after the system has been used is permitted as well. But this "after-registration" has to be done within an exactly defined time-span, otherwise an enforcement process is initiated.

Besides the indication of vehicle data and payment-mean the check-in process can include also exact time and exact stretch that will be used.

Very modern check-in facilities will be adopted as part of the heavy vehicle charging scheme in Germany.

5.2. Classification of Non-Equipped Users

The allocation of Non-equipped users to tariff classes can be covered only by centrally stored vehicle characteristics or by measured vehicle characteristics. The procedural approach is basically analogue to equipped users.

38 ANDREA FELIX

Centrally stored vehicle characteristics. The Vehicle/Tariff class is determined from information stored in the central database. The user has to pre-register using offered registration methods. The registration has to be done before using the system for the first time.

Measured vehicle characteristics. The Vehicle/Tariff class is determined from physical characteristics measured at the moment of passing the charging point. This method is only feasible for conventional toll-plazas.

5.3. Charging Non-Equipped Users

Basically two options are available for charging Non-equipped users.

5.3.1. Using Non-Stop Lanes

AVI and LPR/Central account (post-payment with central account).
Payment when checking in/out of system (internet and call-center) using means of payment that can be authorised on-line (e.g. petrol- and creditcards).

5.3.2. Using Stop Lanes/Service Area

Payment at (non-manned) automates, coin- and vending-machines etc.
Payment at manned tollbooth/counter using cash, credit cards etc.

6. ENFORCEMENT CONCEPT AND TECHNICAL APPROACH

Enforcement is a rather important part of any charging scheme. Table 11 and the following section give an overview of conceptual design elements and technological approach for detection and registration of violators.

6.1. Enforcement Concept

The enforcement concept mainly covers the extension, process and impact to traffic flow, location and equipment used. The extension of enforcement can cover:

All vehicles.
All vehicles using the tolled road are systematically enforced, e.g. all conventional toll-plazas using barriers.
Only sample of vehicles.

Table 11. Technical Options for Non-Equipped Users.

Spot-check of a sample of vehicles only, e.g. non-equipped users in Switzerland, check of kilometre-reading when entering/leaving Switzerland.

The impact to traffic flow can be distinguished between:

Automatic and non-stop; no impact on traffic flow at all, dedicated automatic enforcement stations on the motorway like the ones used in Switzerland and Austria: vehicles subject to the fee are automatically detected and checked for correct payment of the fee. Also used on electronic toll lanes at conventional toll plaza (e.g. Via Verde in Portugal). The vehicles drive at normal speed in dedicated lanes equipped with EFC-equipment, with no barrier. Pictures of all non-equipped or suspicious vehicles are stored and the data is manually verified.
Automatic and non-stop; reduced speed e.g. as used on electronic toll lanes at conventional toll plaza (e.g. Telepass in Italy). The vehicles drive at reduced speed in dedicated lanes equipped with EFC-equipment. As soon as the automatic tolling process is completed a barrier opens. If the vehicle or the means of payment is black-listed the barrier remain closed and the vehicle has to stop.
Manual by stopping vehicle Mobile enforcement teams identify and stop violators. e.g. "Sicherheits- und Kontrolldienst" of ASFINAG in Austria. Several mobile enforcement teams are either cruising on the motorway or screen systematically the HGV at dedicated check-points (Verkehrskontrollplatze). ¨

40 ANDREA FELIX

The location of enforcement can vary:

Enforcement procedure at charging point (integrated) The enforcement procedure is combined with the charging process. Vehicles which have not paid the fee correct are stopped or resisted by taking pictures and storing time and place of offence. e.g. tolling gates in Portugal, fix and portable enforcement stations in Austria.

Enforcement procedure initiated at different time and location (separate) Basically enforcement procedure executed by mobile teams and dedicated special enforcement stations on the tolled network without tolling functionality. e.g. "Sicherheits- und Kontrolldienst" of ASFINAG in Austria or network of enforcement stations in Switzerland.

6.2. Enforcement of Equipped Users

The enforcement process of Equipped users can be split into two main processes:

Check of user/OBU against white-list and black-list.
Compare declared/claimed characteristics in the OBU/central-system vs. measured characteristics on the enforcement site.

A white-list check includes mainly the check if the user/OBU has a valid contract.A black-list check can cover different points. Reasons for entries in operators blacklist can be:

black-listed means of payment;
(already) register fraud of the system by the user;
lost/stolen OBU.

The comparison of declared/claimed characteristics vs. measured characteristics can cover:

Number of axles: In schemes based on the number of axles, the fee varies according to trailer presence. The driver must manually adjust and declare the number of axles directly on the OBU.
Trailer presence: In schemes based on the maximum permissible weight, the fee varies according to trailer presence. The driver must manually declare the correct trailer status (presence or absence) on the OBU.
OBU in correct vehicle: In schemes based on dedicated OBU for different vehicles categories and/or with special OBUs for exempt vehicles it is very important to verify that the vehicle is using the correct OBU. This can basically be done by a licence-plate check.

Some characteristics like emission classes or maximum permissible weight of vehicle (-combination) cannot be detected or measured automatically. These characteristics can only be enforced by mobile teams checking vehicles documents.

6.3. Enforcement of Non-Equipped Users

The enforcement process of Non-equipped users mainly covers a check whether the user/vehicle has a valid and correct registration to use the system. This most basic enforcement check is the simple verification if there is a valid "log-in" of the vehicle to use the system and if the declared characteristics (e.g. number of axles) correspond with the measured facts. In more complex schemes exact time and exact section of the road used has to be declared as well and must be checked by the enforcement process.

In tolling schemes with the compulsory use of an OBU detected users without OBU have to be checked against a white-list which includes vehicles permanently or temporarily exempt from the fee and from installation of the OBU.

6.4. Equipment Used for Enforcement

The enforcement process has to cover mainly four steps:

detection of the vehicle;
registration of violation and identification of the vehicle;
transmission of data to central-system;
manual verification of the proof of fraud.

At automatic enforcement stations the most common detection method is the use of laser scanners. According to the profile of the vehicle the system can detect vehicles subject to the fee and by using complex algorithms even an allocation to vehicle classes is possible.

For the registration of violation and identification of the vehicles pictures are taken and the license-plate is automatically analysed. LPR/OCR (License-Plate Reading/Optical Character Recognition) algorithms achieve nowadays a very good reliability, but nevertheless a manual check of the automatically collected and analysed data remains necessary.

The OBU of equipped users is read out by using the communication link (DSRC). The data can be used to verify the correct payment of the fee (e.g. correct declaration of number of axles) and/or to support the LPR/OCR process to identify the vehicle. 42 ANDREA FELIX

The digital pictures and additional enforcement data is sent to the central-system for manual verification. Main technology used for transmission are terrestrialbased data networks. Transmission using Cellular Network (e.g. GPRS, UMTS) is not very common yet. By matching the enforcement data with the stored data of the central-system a fraud of the system can be identified and followed up.

Only very simple charging schemes, e.g. flat rate for all vehicles, can renounce on the additional manual verification by accepting a higher rate of unjustified fines, which have to be refunded/cancelled afterwards.

Depending on legal regulations and system design the identification of the vehicle can be supported and verified using data from the official road traffic licensing department. The fine can directly be sent to the owner of the vehicle. At the moment this is often only possible for domestic vehicles; cross-border enforcement is not yet current practice. The prosecution of foreign vehicles basically has to be covered by mobile enforcement teams.

At *manual operated enforcement*stations the enforcement process can basically be reduced to manual registration of the license plate. Equipment used is dedicated hand-held with an on-line mobile communication interface to the central-system. Within a very short time an actual or earlier fraud on the system can be verified. The violator can be charged on the spot.

4. LEGAL, INSTITUTIONAL AND CONTRACTUAL ISSUES

Rosario Mac ´ ario ´

1. INTRODUCTION

The institutional setting adopted for the application of road tolls plays a fundamental role in the success of its implementation. It involves the identification of which institution will manage revenues and other financial flows associated with the road pricing schemes, i.e. which institutions will have the power of financial means for the implementation and maintenance of the schemes.

Institutional design relates with the concept of "goodness of fit": a well designed institutional setting should be the one that is internally (in this case, within the transport system) consistent but also harmonized with the external order of the overall legal and institutional setting.

It is often acknowledged that optimal design of institutions and policies lies in a normative-empirical sort of complementarities (Goodin, 1996). However, the normative approach usually does not indicate the way optimal arrangements are obtained or why a certain design fails to reach an optimal configuration. This information is largely obtained from field observation.

In addition, as Goodin points out, there is no reason to suppose that agents in the real world will pursue optimal designs, or even that they will recognize one when they see it. It is the goodness of their fit that makes us elect them as best in class and at the same time makes them well suited to survive in their broad environment. This construct justifies cases of long living inefficient institutions – lack of internal consistency but high external fitness.

But institutional design does not refer exclusively to shaping and reshaping formal structures. It also covers other arrangements that intervene on the societal behaviour and interaction of individuals. So three issues can be considered as part of this field of concern: behaviour of people in society (as individuals and as agents); the opportunities and incentives associated with the socio-economic set-up; the evaluation criteria which determine that some behaviours are more acceptable than others.

Following this rationale the institutional design, seen as an instrument to reduce uncertainty in human interaction (North, 1990), is an important factor for the competency of authorities to integrate a road pricing measure with other policies fostering coherent achievement of strategic objectives.

Different government levels and structures have different potential, for example to integrate a road pricing measure with other measures or policies in order to reduce congestion, accidents and pollution as well as to co-ordinate between the different agents involved in the process of implementation of any measure that requires a more holistic approach.

European research found evidence that allocation of financial power to local authorities contributes to increase people willingness to pay, as they understand the potential of a direct advantage from the application of these revenues. Besides, trustworthiness of governments seems to be higher as decision-makers are closer to the field and so perceived by people as being more politically accountable and, consequently, more controllable and of less risk (TIS, 1999).

The implementation stage, in turn, requires a thorough assessment to understand the need to undertake a phased approach. Some factors influencing the success of implementation are: the socio-economic environment where the measure is to be implemented; the government level that is supposed to monitor the progress of the implementation process, and possibly develop adjustments based on feedback information from the field and need to set-up any sort of contingency measure for risk mitigation. Indispensable steps in the practical feasibility assessment are:

legal, technical, managerial feasibility;
legal, technical, managerial interoperability, whenever a cross-border situation is involved between different governing structures, e.g regional or national;
necessary accompanying measures (e.g. land-use, fiscal, environment, legal and regulatory adjustments, enforcement, etc);
distribution of costs of change against the benefits of implementation of the new scheme;
assessment of the main risk factors against a successful implementation and any potential spill-over effects.

2. STRUCTURING THE FUNCTIONS OF AN IRPS

In order to follow the normative-empirical compromise in assessing the adequacy of institutional design elements to accommodate the implementation of an IRPS with internal consistency, we must have a clear picture of which functions are to be fulfilled and whether there are any legal, technical and managerial constraints.

For the definition of the system functions three different decision levels must be considered:

Strategic level, where the main objectives of the system are defined and constraints identified;
Tactical level, where the functions that will enable the accomplishment of the strategic goals are designed, and the monitoring tools conceived;
Operational level, where the system is put to work and monitoring and control is exercised on whether the IRPS is actually performing according to the objectives and design defined at the higher levels.

Depending on the legal and administrative organization of the country and rules set for the provision and exploitation of the different associated services, the functions corresponding to these levels may be allocated to different institutions.

Despite these various options on the allocation of functions, it is consensual that functions associated to the strategic level have by their character a strong political component and so should be performed by bodies composed of elected politicians (Parliaments and partly Governments), while the ones related with the tactical level can be either developed by the state or by a private body under contract. Finally, recent experiences show that, when objectives and rules are clear, the operational level can be performed by private companies, possibly with cost efficiency and accountability advantages and better quality of service. However, in the European reality the operational level is still largely in hands of public agencies.

The following Tables 1, 2 and 3 provide a list of functions identified for the different decision levels.

In addition, a fundamental criterion to define institutional roles and responsibilities in the implementation and management of a road tolling scheme is the potential to mitigate risks. In principle, roles and responsibilities should be better allocated if risk is put under the management control of the entities empowered to best develop mitigation measures.

Tables 4 and 5 indicates the major types of risk associated with road tolling and possible mitigation measures. Depending on the institutional and political environment the management of these risks will be allocated to the different government levels and operational agents who are in best position to control and

Table 1. Functions at Strategic Level.

Source: Adapted from TIS.pt (2001).

mitigate. It is thus a tailor-made configuration limited by legal, regulatory and institutional frameworks.

Besides the levels of government existing within the legal framework of each country we still have to consider the allocation of functions at the European level which is a very particular governance level that results directly from the existence of the European policies in the Transport domain or having some interaction with it. Its influence is extended only up to the moment of production of European legislation, which has a strategic character. From that point onwards, even if an open co-ordination method is thought as a consequence of the proposal made in the Lisbon Council of 2000, it is the duty of national authorities to ensure its application, leaving to the European Commission and Courts of Justice a monitoring and enforcement role regarding the application of that legislation.

Within this European governance level, the discussions about the assignment of pricing tasks to levels of governments, largely supported by research projects (such as TRANS-TALK, PATS and CAPRI, in the 4th and 5th RTD Framework – Transport Program) lead to the following conclusions:

Table 2. Functions at the Tactical Level.

Source: Adapted from TIS.pt (2001).

Table 3. Functions at the Operational Level.

Table 3. (Continued )

Source: Adapted from TIS.pt (2001).

The definition of pricing principles must be done at the European level, in order to ensure fair competition between countries (and between modes to the extent currently possible), and to avoid discrimination of operators or operations based on nationality;

Table 4. Type of Risks and Possible Mitigation Measures (Part 1).

According to the subsidiarity and territoriality principles, Member States play a determinant role in defining and implementing transport pricing schemes at national level, which entail both the setting of taxation and pricing levels and the design and implementation of the corresponding schemes. Compatibility with EU definitions must be achieved in pricing principles, tariff levels, and (when applicable) technological interfaces between the user and the entity applying the price;
The paradigm for the definition of minimum quality levels, as a citizens' right, is also to be defined at the European level in order to ensure that a minimum basis of fair an equitable treatment of citizens is achieved across countries, although the

Table 5. Type of Risks and Possible Mitigation Measures (Part 2).

parameters for operational definition should be left for the national and regional levels of governance.

Finally, regarding the implementation of incentive-based instruments, it is important to identify:

The various actors being involved in or being affected by the political decisionmaking process;
The interests of the various participants in the political-economic process;
The influence of the various (groups of) actors given the specific institutional setting;
The relevant aspects determining the acceptance of the various instruments being used in the implementation of any pricing policy, which is extremely relevant to identify who are the winners and losers and their capacity to raise barriers to the implementation of a given IRPS. For this a number of issues have to be discussed and considered since the conception of the scheme, such as:
- what are the incentives of various groups of actors to organise themselves according to their specific interests; and
- what are the possibilities of the actors both on the demand side and on the supply side of the political market to influence the outcome of the political

decision-making process. According to the economic approach to explain human behaviour the outcome of this process is determined not only by the preferences and interests of the various actors but, even more importantly, by the restrictions, i.e. the institutional interactions and the institutional constraints.

3. PERFORMANCE ASSESSMENT OF DIFFERENT INSTITUTIONAL SETTINGS

Based on the systematisation presented in the previous chapter on objectives, decision levels and functions, two main issues have to be discussed for the choice of institutional settings that define the public-private framework within which the different actors have to perform their roles:

Degree of retention of ownership and control functions to be left in public hands;
Dimension of the private investment role in the infrastructure, with corresponding construction and revenue risks;
Allocation of responsibilities with technical procedures for transactions, i.e. identification, classification and collection.

The existence of a dimension of private investment is normally associated with the necessity to engage private funds for infrastructure investment (thus obtaining a relief of the State budget for other purposes), whereas the allocation of responsibilities of transactions to private entities usually results from a political orientation in the sense of private management of public services, connected to search for higher operational efficiency (and sometimes of accounting transparency). From these two main dimensions the following three basic organizational models can be derived (TIS.pt, 2001):

All public: public investment and public responsibility for transactions;
Mixed: Public investment and private responsibility for transactions;
All private (meaning mainly private as some sovereign functions will always be kept in the hands of the Government): Private investment and private responsibility for transactions.

The "all public" model (Fig. 1) assumes the following:

The central or local government finances and operates the IRPS-scheme. Permission to operate a IRPS-scheme is given by the representatives of the people. If the scheme is distance-based the public operator measures the kilometres driven and makes out the debiting according to the vehicle type and

Fig. 1. All Public Model: Public Investment and Public Responsibility for Transactions. Source: TIS.pt (2001).

tariff class regulations as determined by the government. The communication network is also publicly owned;

The vehicle owner is responsible for installing the OB-equipment needed and to control that the OB-equipment is working. He has to pay, either (periodically) by invoice or by automatic debiting, to the public operator responsible for the collection of the revenues;

The national government determines the specifications for the OB-equipment. To stimulate interoperability a European framework has been adopted with specifications for the OB-equipment and that national authorities can only decide on details. National authorities are responsible for the inspection of OBequipment;
Private manufacturers produce the OB-equipment that is installed in the HGV by private enterprises (authorized retailers or garages or by the transport companies themselves);
Surveillance of HGVs is the duty of the traffic police. In cases of fraud they impose sanctions (fines);
The public operator and central/local authorities do the collection of the revenues. The central/local authority takes care of the distribution of the revenues.

A "Mixed" institutional model (Fig. 2) assumes the following:

The IRPS-scheme is financed by the authorities but operated by private firms. These private firms get a concession from the state to operate the IRPS. The private IRPS operator is paid by the state (e.g. according to a specific rate, by time period or by vehicle served). By using the IRPS scheme the vehicle owner has an automatic agreement with the system operator. If the scheme is distance based the operator measures the kilometres driven and sends an invoice or uses automatic debiting, according to the vehicle type and tariff class regulations of the state. The operator therefore has to pay for the use of an existing public communication net;
The vehicle owner is responsible for installing the OB-equipment needed and to control that the OB-equipment is working. He has to pay, either (periodically) by invoice or by automatic debiting, to the private operator responsible for the collection of the revenues;
Specification, production and instalment of OB-equipment, as well as surveillance and enforcement of correct payment, are identical to the all-public case.

Depending on the payment system adopted, there may be (or not) a function to be carried out by the private operator in the collection and the use of revenues. In this model of public (infrastructure) investment, this is the duty of a state organisation, possibly in co-operation with a financial institute, but the private operator may act as an intermediary for at least some forms of payment (for instance cash payments).

The so called "All private" institutional model (Fig. 3) assumes the following:

Fig. 2. Mixed Model: Public Investment and Private Responsibility for Transactions. Source: TIS.pt (2001).

The IRPS-scheme is financed and operated by a private enterprise. This enterprise has received permission from the authorities to build and operate the IRPSscheme. The revenues from the IRPS-scheme belong to the private system operator, possibly with some contractual arrangement providing for downside and upside revenue risks. By using the IRPS system the vehicle owner has an

Fig. 3. All Private Model: Private Investment and Private Responsibility for Transactions. Source: TIS.pt (2001).

automatic agreement with the private system operator. If the scheme is distance based, the operator measures the kilometres driven and sends an invoice or uses automatic debiting, according to the vehicle type and tariff class regulations of the state. The operator has also built a private communication net to make the communication possible between the OB-equipment on board of the vehicle and the detection-installation along the side of the road;

The vehicle owner is responsible for installing the OB-equipment needed and to control that the OB-equipment is working. He has to pay, either (periodically) by invoice or by automatic debiting, to the private system operator;
Specification, production and instalment of OB-equipment is identical to the all-public case;
Surveillance of HGVs is the duty of the local traffic police. In some cases it may be possible to outsource surveillance functions to specialised private enterprises. In cases of fraud, sanctions (fines) will have to be imposed by public authorities;
The private system operator does the collection and distribution of the revenues. A clearing house will play a role in the case of multiple system operators.

A distinction should be highlighted between the institutional models and the basic forms of IRPS. The latter represent the basis for computation of the toll values while the former relate to the allocation of roles and functions in the system. Between the two there is a weak level of interdependence.

For example in the mixed model, where operators do not have an infrastructure investment role, it is possible to totally detach the tolling functions from any function of road network management, what makes the previous function a pure payment service that can be performed by another service provider (e.g. bank or telecom company). In these cases, although interoperability is made much easier, there are some risks that the toll is perceived as a sort of tax without a close association to the road service provided.

Despite the weak relation between soft and hard features of the system, there are a relatively wide number of requisites to assess the institutional-technical fitness of the solutions being addressed. These are:

(a) Capacity to address, and gain the confidence of governments and citizens, the performance difficulties raised by the different tolling operations, such as: identification, classification, fraud detection, billing, collecting and enforcing.
(b) Ensure that tariff levels and allocation of revenues are in line with policy goals and public expectations regarding the existence of tolls.
(c) Ensuring high levels of productive efficiency and transparency of accounts and procedures.
(d) Show capacity to impose sanctions and other enforcement related measures.
(e) Ensure adequate treatment of clients' complaints as well as of possible guarantees of quality of service.
(f) Minimize risks of conflict of competencies.
(g) Minimize risks of abuse of position towards State agencies, private companies (namely when more than one toll operator exists).
(h) Achieve adequate compromise between the required stability and long-term commitments and flexibility to adapt to possible evolving policy goals and associated movements of change.
(i) Appropriate auditing processes to assure regular verification that the above goals are being properly met.

The work developed for assessment of the three institutional settings previously identified (TIS.pt, 2001) revealed that there is no model which can claim the overall best performing statute. At the opposite the merit is in the detailed design and so in the advantages and disadvantages that each model offers confronted with the requirements expressed above.

The methodology adopted for the assessment, which results are presented in the tables below, consists of a systematic screening of all the functions identified above, across all three institutional models. The resulting matrix presents the argument of the element critical for good performance followed by the letter of the requirement in the list previously presented.

The assessment presented in the referred tables takes into account that each model has a considerable variety on the institutional configuration effectively implemented in what concerns the following dimensions:

Policy;
Legal and contractual;
Financial;
Technical/Transactional;
Maintenance;
Interoperability related.

These tables identified for each function the respective critical relations for good performance of each institutional model. A full assessment still requires the identification of the critical level of performance for each function. This degree of criticality should be established on the basis of the number of incidences that each requirement reveals in all models (Tables 6 and 7).

The importance of the requirement, in turn, is given by its incidence. Table 8 provides a qualitative assessment, based on degrees of sensitivity, marked by the levels "High" "Medium" and "Low," resulting from the counting of incidences in the previous functional assessment. The analysis done led to the understanding that no assertion can be made on the best performing institutional model. On the contrary all models revealed more or less sensitive to the same critical issues, only varying in the degree of emphasis enabling to assess the sensitivity of each institutional model against a particular technical acceptability requirement.

Table 6. (Continued )

Table 6. (Continued )

| TIS.pt (2001).
Source: | | | |---------------------------|--|--|

| So
T
I
S.
(
2
0
0
1
).
t
ur
ce
:
p | | | |------------------------------------------------------------------------------|--|--|

Table 7. (Continued )

| S.
(
2
0
0
1
).
So
T
I
t
ur
ce
:
p | | | | |------------------------------------------------------------------------------|--|--|--|

Table 8. Critical Requirements of Institutional Models.

	All Public	Mixed	All Private
(A) showing technical capability to properly address the challenges of the tolling operations (identification, classification, fraud detection, billing, collecting and enforcing) and earning the trust of government and citizens in its performance	Low	Low	Medium
(B) ensuring that tariff levels and allocation of revenues is in line with policy goals and public expectations regarding the existence of tolls	Medium	High	High
(C) ensuring high levels of productive efficiency and transparency of accounts and procedures	Low	Medium	Medium
(D) showing capability to impose sanctions on abusers		Low	Low
(E) ensuring proper treatment of clients' complaints as well as possible guarantees of quality of service	Medium	Medium	High
(F) minimising risks of conflict of competencies	Low	Low	Low
(G) minimising risks of abuse of position towards State agencies, private companies (namely other toll road operators) or persons	Low	Low	Medium
(H) showing an adequate compromise between the stability required for long-term commitments and flexibility to adapt to possibly evolving policy goals			Low
(I) including the auditing components for regular verification that the above goals are being properly met	Medium	Medium	Medium

Source: TIS.pt (2001).

Besides the concerns on technical acceptability there are other policy determinants which have to be addressed in the decision process of which is the best suitable institutional model for a given situation. These are:

Investment needs;
Organizational efficiency;
Competitive pressure;
Long term flexibility (that is capacity to change for better adaptation to new circumstances).

From the policy point of view, the main factor of choice is the level of investment needed beyond State capability. Whenever major investments are at stake a clear financial advantage is seen in the all-private model that has to be balanced with the level of development and management control the State is willing to retain in the project.

In cases where the infrastructure is already set and only partial investments are needed, then the advantage goes to the mixed model. An extreme situation is when the operator does not have to take any investment role, in which case it is possible to detach the tolling functions from the other functions related with road network management. The model reduces then to a payment service facilitating integration and reducing interoperability problems.

In what concerns organizational and productive efficiency, the concentration of functions within the same entity may offer potential synergies, in particular for the cost of co-ordination if the IRPS is operated through a central unit, which seems to favour the public model. However, when the productive efficiency is the prime concern, selection of private operators enhances the chance of adopting more innovative management processes, although this brings additional co-ordination costs and additional procedures at the interface between public and private institutions. For example the grant of a new concession will impose new efforts on management and control functions, which in turn will cause an increase in the number of potential incidents and with it the sensitiveness of the requirements referred above.

The competitive pressure introduced in the system is often indicated as a driver for efficient management and delivery of good quality service. Tendering the technology-based services is seen as one of the ways to introduce competitive pressure, in which case only the private and mixed model will be suitable. The more the market is open to several competing operators the more cost-cutting dynamics will reveal, such as:

Operators will exert pressure on the producers of on-board and roadside equipments to lower costs;
Vehicle owners will tend to choose the IRPS operator offering the best possible mix of different services, whenever technology allows more than one IRPS operator in the same road;
Privatization of the function "inspection of on-board and road-side equipment" and "surveillance" will have additional cost reduction effects, although assessment is required for economical feasibility and profit potential.

Another determining issue is the balance between stability of service and flexibility to adjust to changes. Flexibility issues are usually raised regarding two main aspects: technological equipment and level and structure of charges. In what concerns the former the public and mixed models seem to have a tendency towards the choice of only one type of on-board equipment (e.g Swiss vehicle fee), which reduces the flexibility to adapt to future technological developments.

At the opposite, if public authorities restrict themselves to the function of formulation of standards that the equipment has to meet, then the institutional model will be more open to adapt to future changes. In what concerns flexibility for setting the level and structure of charges this is usually inhibited by the typical contractual arrangements between public and private entities. Due to the fact that contracts involve long-term tightly drafted commitments re-negotiation usually entails very high financial compensations. A possible solution is to reduce the duration of the concession contracts. The following section addresses the issue of completeness of contracts and the consequences that usually lead to financial compensation between the parties.

4. CONTRACTUAL DIFFICULTIES AND CONTRACT EVOLUTION

Today there is a strong awareness that contractual completeness is virtually impossible in almost any type of partnership, but especially in sectors where there is still a strong State intervention as it is the case for the road sector in Europe.

In Guasch (2001) an analysis is made of 1000 utilities and transport contracts signed during the 1990s, indicating that renegotiation happened in around 50% of the cases, and suggesting that the likelihood of renegotiation is highest when the auction criteria is driven by objectives of minimization of average prices paid by the users of the services submitted to bidding procedures. One of the major pitfalls identified in that study was the low effort put in the assessment of the factors influencing fluctuations of demand.

Many authors from different sectors have dedicated research efforts to constructing a framework of analysis of incomplete contracts and so far this unique framework seems to be non-existent. One of those authors was Tirole (1999) who summarized a set of reasons that can justify contract incompleteness:

Contingencies that may arise during the execution of the contract and that cannot be foreseen;
Even if there is prospective capacity to anticipate all possible contingencies the costs of doing it would be prohibitive;
Contracts can only be contingent upon variables that can be verified by a third party, otherwise the contract is non-enforceable and consequently non-effective.

In such cases, an incomplete contract may be preferable to a complete one that has a high probability of having to be renegotiated in the future.

We can easily deduce that renegotiation entails a certain number of risks, given the loss of bargaining flexibility by all partners and increasing the collusion/corruption potential between the firm and the public administration body in charge of the renegotiations. So, promised efficiency is far from being the only variable to consider when selecting a firm, and this attribute can even be jeopardized at exchange for the reduction of the probability of renegotiation.

Road development and operation is no exception to the previous evidence. Being a sector of great importance to the overall performance of the economy, strong government intervention has a long tradition and polices may evolve in response to difficulties, for instance of congestion or environmental aggression. From the moment that transport policy is oriented in favour of variable tolling schemes as an instrument to fight congestion, contracts for motorway construction and operation, or simply for operation of a tolling system (NET scheme) will have a very different object from those currently in place.

Existing concession contracts would have to be deeply changed or cancelled, because of the changes in the object of concession and on the financial flows. An addition degree of complexity exists if variations of tolls are made in accordance with real time observed traffic conditions.

Under these circumstances the concept of traffic management on the motorway will be redefined and the corresponding function will require re-assignment, specially if the institutional setting involves separate firms dealing with road maintenance, traffic management (mostly incident management) and toll collection. This would imply modification of the nature of existing contracts. If tolls are used as a traffic management instrument, concession contracts have to be changed as the nature of the underlying business process is strongly changed.

Several solutions are possible to overcome this problem. At one extreme of the spectrum of solutions we can consider that the concessionaire evolves into a mobility manager, which might bring along a transition from a NET scheme to a DAREA scheme, or at the opposite extreme the withdrawal of the concession back to the public hands, which normally would represent a very expensive solution since traffic management is only necessary on roads with high traffic volumes. In between, several intermediate contractual forms in public-private partnerships are possible.

The concept of a "regional mobility manager," developed in a public-private partnership has already been called forth as a possibly useful evolution when analysing the problems facing urban mobility and evolution of public transport systems (Viegas & Macario, 1999 ´ ) in Europe, and it is also in line with some trends in the United States to set up operations agencies which are responsible for both public transport and road development within a certain geographical limit from the city centre. This development should be seen as no surprise since in the large conurbations there is a strong overlap of urban and interurban mobility concepts and of the infrastructure elements which are conceived to serve both.

Similar changes occurring in a case where a DAREA scheme is implemented would have a much wider and deeper impact, which would be extended not only to motorways but also to national, regional and local roads within a certain perimeter. Beyond the contractual relations between the concessionaire and the conceding authority, this implies involvement of other levels of governance in the discussion and concertation of decisions regarding the justification for tolling (competing with other policy alternatives), exceptions, allocation of responsibilities for the various tactical and operational functions and of course the impacts of distribution of revenues and definition of the principles underlying that distribution.

There is a wealth of detail that require definition, such as the scope of functions of the entity responsible for managing the tolling scheme. Is it only tolling, or is there some element of road maintenance, signalling and traffic management? And regarding the issuing of permissions for new accesses to the road network, how is this power shared between the traditional authority (with separation according to the status of the road in question) and the new tolling entity?

The concept of private concession is always associated with some transfer of risk to the private side. Whenever road tolling is seen as an instrument for financing road construction the concessionaire takes both the production risk (associated with construction costs) and the commercial risk (associated with varying demand). Recognizing the different life cycle of the two types of risks it is common practice that the State allows renegotiation of the financing contracts (between a concessionaire and the bank syndicate) after the end of construction that is, when production risk ceases. In addition, interoperability requirements will very likely imply some degree of harmonization of vehicle classes across countries and this will have financial implications.

A clause is normally included in contracts defining how toll levels can vary during the life of the concession, and the definition of classes is supposed to be fixed during this period.

Quality of service clauses, when they exist, are still defined only by the number of hours of lane closure for repair, although more recent contracts already have some clauses related to safety levels. This practice contributes to increase the commercial risk and also the premium to be paid to the banks. As soon as quality of service starts to be defined in the user perspective, considering travel speed, safety, etc., with implications on the toll levels and on indemnities in case of poor service, severe implications will accrue to the contractual and financial domains, but this should be an effective instrument of pressure on the concessionaires to strongly reduce the incidence of situations of poor service.

In such cases, financial surpluses will be generated in parallel with fines for under-performing traffic demand management. This will imply changes in contracts resulting from a significant change in the risk profile of the concessions associated with traffic management, and consequently a new negotiation of the financing conditions with the banks.

The above examples of evolving requirements for the good execution of the missions of a motorway show that very complete and definite contract of long duration, even if they are the standard practice in the PPP sector, may create serious barriers to the adaptation of transport policy to those evolving requirements. Alternate solutions are not in practice yet, and should be sought either in the field of incomplete contracts or in the adoption of complete contracts of shorter duration (Viegas & Colac¸o, 1998). In the latter case, the State would have to pay the residual value of the infrastructure at the end of the concession, but could immediately launch a new tender for a new period of concession, with the full right to redefine the new terms of concession.

Enforcement is a special contractual function given the impact caused in the effectiveness of the contractual arrangement. Formally focused on the general obligation to detect fraudulent use of the services and apply the corresponding penalties, it has however a more complex contractual definition and application given its potential to overlap with the normal scope of road police mission, like speeding and other forms of dangerous driving. Given this controversial aspect there is room to discuss on what should be the most efficient and legally acceptable form of enforcing payment of tolls, considering that the different European countries do not have legislative harmonization regarding the application of penalties for the several possible injuries, as was recently observed in a study developed for the European Commission (TIS.pt, 2004).

In several countries motorway concessionaires also have the right and obligation to perform some parts of the enforcement function. They detect fraudulent drivers, and are given access to their identification through the vehicle number plate (by the Police or administrative authorities) and are entitled to contact the user requesting payment, in some cases with a surcharge. In case of denial to payment their action is then limited to send the case to court.

Also here, enforceability varies according to the scheme used. Detection of fraudulent use and enforcement of payment is much easier with a NET scheme using road side (DSRC) technology, by which the user is detected and identified as in a fraud condition in a limited number of places. The vehicle can be photographed and the enforcement process is based on both the identification present within the OBU (if there was one) and on the number plate.

In a NET scheme with GNSS/GSM technology, as well as in a DAREA scheme without vehicle detection and identification at entry and exit points, things become more complicated and costly in this dimension. Several ways are possible to achieve this, but transaction costs are much higher, as well as the risk of undetected fraud (and thus of the incentive to free-ride on short trips). The same concept of fraud detection and payment enforcement by the concessionaire could be applied for the case of DAREA schemes, even with the higher levels of real time traffic management described above.

But there might be here a case of privacy intrusion in the sphere of the contractual relation between the user and the concessionaire, especially if this is done in a sophisticated way, implying for example measurement of speed levels. The following discussion questions have been already raised in the context of NET schemes:

To what extent can (or even must) a concessionaire supply to the police evidence of speeding or excessive driving hours by certain drivers?
To what extent can/should information collected with a certain purpose (under a specific contract with one public agency) be of compulsory delivery to the police if it contains evidence of practice of crimes or public offences?

Answers to these questions are not harmonized from country to country, and can even vary from one type of offence to another, but the possible implications should not be ignored when conceiving the IRPS scheme and the various contractual relations underlying it.

Enforcement of fee collection from foreign drivers can also be a delicate subject, and will interfere with the question of roaming charges. When one concessionaire in a given country requests identification of the owner of a vehicle with a certain national number plate, an official national agency replies (with no difficulty in most cases). But when the concessionaire is in one country and the vehicle is registered in another, this would be very different, at least in current conditions.

Only very recently are moves for cross-border cooperation in the case of traffic offences being enacted, and it should be expected that extension of this pan-European transparency to the case of fee collection follows, but it will not be quick or easy. But also here an equity discussion is raised regarding the different rights given to the national drivers in comparison with the foreign driver in what concern dimension and application of penalties.

5. CONCLUSIONS

From what was said in this and other chapters of this book, the choice of a particular model is mainly influenced by four factors: investment needs; organisational and productive efficiency; competitive pressure; and long-term flexibility.

The all private model is favourable where there is a substantial need for investment to develop infrastructure and telecommunications. Besides, the private sector can still constitute a leverage for more efficient and innovative management practices. Short concession terms will introduce greater flexibility in contractual arrangements, facilitate competitive pressure and prevention of technical obsolescence. In this model risks of low efficiency are likely to be reduced, but sustainability of the model is possibly dependent on the development of contracts where investment and exploitation are at least partly decoupled.

The mixed public/private model is favourable in countries where the infrastructure is already well developed and the need for investment is mainly related with communication systems. In these cases the operators can be detached from road network management functions and remuneration of the private operator should not be based on the actual tolls collected, since this could lead to segmentation to maximise revenue. Instead, remuneration should be based on the number of vehicles served, using other instruments to induce a desired traffic distribution.

For the all public model to be adopted it requires the public organisation to function with great efficiency in order to enable the preference for this model. Depending on the given situation there might be a potential of synergy from having all IRPS functions concentrated in one agent. An additional advantage is that the absence of private revenue eliminates possible conflicts regarding enforcement of the system. However, there are additional risks accruing from the capacity of the public agency to handle complex operations with efficiency, which may lead to higher costs.

Research in DESIRE (TIS.pt, 2001) concluded that the various institutional models are in general sensitive to the same critical issues for good performance although some features are felt more sharply in particular models. Regardless of the institutional model it is important to ensure that the following requirements are met:

Tariff levels and allocation of revenues must be in line with policy goals and public expectations;
Client's complaints must be treated properly and service must be guaranteed;
Auditing components are necessary to verify that goals are being met;
High levels of productive efficiency and transparency of accounts and procedures must be ensured;
Risk of abuse towards State agencies, private companies or persons must be minimised;
The user friendliness of the system must be supported by the principle of "one single onboard unit and one single contract for one vehicle."

Two distinct situations have to be considered: the fresh start and the change of an existing system. In the former the main concern is the goodness of fit between the objectives to be achieved by the IRPS and the various models discussed above. The latter situation brings additional constraints since the institutions already engaged in the operation and management and control of the system have to be submitted to a change process to ensure practical feasibility of the new scheme.

Whenever a given situation implies fine tuning of prices with respect to traffic management, innovative concession contracts are needed if an all private model is to be adopted. In particular, it would be necessary to ensure that search of revenue maximization would not lead to market segmentation and price differentiation practices.

Depending on the starting situation and the desired end state different transition stages might have to be considered, possibly calling for short term adjustment contracts. When moving from one model to another, a major concern is choosing the appropriate institutional design and defining the role of the public and private sector respectively.

From what was previously said we conclude that in order to be successful any project for implementation of road pricing measures should assure from the outset the following components within the institutional setting:

An overall management structure capable of meeting public policy objectives and to satisfy acceptability criteria from stakeholders, including objectives of area and network development, demand management and equitable cost recovery;
Adequate configuration and empowerment of public entities, which should effectively own the project so that, even when passing it over to private entities through concession, residual powers would still enable to carry out essential governance functions whatever circumstances may arise along the legal lifetime of the project;
Adequate rules providing an environment of trust to foster participation of private partners in the project, allowing them to bring more efficient management methods and also more innovative solutions, while keeping the control and enforcement functions in the hands of public entities;
A sound and well controlled productive process, assured through one or more contractual arrangements that should be thought of as interactive components of the project, entailing:
- Delimitation and allocation of roles, responsibilities, risks and rewards;
- Protection of public interest (normally assumed as reflected in political objectives);
- Capitalisation of entrepreneurial, technical and financial fields resources from private sector;

Creation of adequate funding sources and mechanisms either tax-based or direct project revenues or a mix of both.

REFERENCES

Goodin, R. E. (1996). The theory of institutional design. Cambridge: Cambridge University Press.
Guasch, J. L. (2001). A decade of concession contracts in infrastructure in the 1990s. In: A. Estache & L. Quesada (Eds), Concession Contract Renegotiations: Some Efficiency Versus Equity Dilemmas. World Bank, //www.worldbank.org/.
North, D. C. (1990). Institutions, institutional change and economic performance. Cambridge: Cambridge University Press.
Tirole, J. (1999). Incomplete contracts: Where do we stand? Econometrica, 67, 741–781.
TIS (1999). PATS Acceptability in transport systems, 1999, 4th RTD, EC, D1.
TIS.pt (2001). DESIRE DESigns for Inter-urban Road pricing schemes in Europe, 4th RTD EC, D2.
TIS.pt (2004). Road traffic regulation study, 2004, for DGTREN, available at http:// europa.eu.int/comm/transport/road/publications/trafficrules/index en.htm.
Viegas, J. M., & Colac¸o, V. (1998). Le "project finance" pour des autoroutes: Recensement des risques principaux et encadrement des etudes de trafic, actes du Symposium International sur le Financement de la Route, Paris, Nov. 1998.
Viegas, J., & Macario, R. (1999). ´ Financing urban mobility. Porto Alegre: Panamerican Congress of Public Transport.

5. THE ACCEPTABILITY DIMENSION

Max Herry

1. BACKGROUND

The main issues connected with acceptability of changes in pricing of transport services – in general – are shown in Fig. 1.

The economic aspects are related to economic efficiency which is a major requirement to the transport sector. The social and political aspects are important for acceptability since measures identified to be economically efficient might be perceived to conflict with fairness and equity concerns. Since economic efficiency in transport pricing may require the use of sophisticated technological solutions the issue of technological and functional requirements must be addressed. However, this is not the only reason for addressing these requirements: user friendliness or transparency of pricing schemes, fall under the category of functional requirements, too. The legal and regulatory framework is particularly of concern if efficient pricing schemes are to be implemented. Furthermore, the question of who operates pricing schemes and who has the revenue competence might affect acceptability of pricing schemes.

Evaluation of acceptability in relation to inter-urban road pricing (for Heavy Goods Vehicles) requires understanding how these questions are formulated and understood in different countries/regions and by different types of people, because the reasons for possible low or high acceptance will vary. Comprehensive identification of the relevant issues is necessary for the various social groups and regions. The aim is to clarify influence factors on acceptability.

76 MAX HERRY

Fig. 1. Structure of Acceptability Issues (Schlag & Teubel, 1997). Source: Pats consortium.

1.1. Theoretical Approach

The objective of this section is to provide a synthesis of the main theoretical approaches from the most relevant literature on acceptability. This will ultimately serve to pinpoint the various dimensions upon which acceptability has been looked at and the theoretical tools used to that effect, so that the most suitable framework for survey analysis can be constructed.

The AFFORD research project (AFFORD, 2001) on the acceptability of fiscal and financial measures and organisational requirements for demand management defines the term "acceptability" as an affirmative attitude towards a specific topic prior to its introduction, whereas "acceptance" refers to a certain behaviour as a reaction towards an already introduced measure. In the case of the IRP/HGV, the focus is on attitudes towards (elaborated – but not yet introduced) IRP/HGV designs and on established IRP/HGV.

The main issues relevant to acceptance are to be found in Fig. 2.

Generally speaking, one can define three kinds of priorities that should be addressed in order to contribute to increasing acceptability:

Pricing for Financing, where drivers pay part or all of infrastructure costs (normally, but not necessarily, the costs of the piece that is being used).
Pricing for Regulation, which enables demand management in order to avoid peak period congestion, or to dissuade some road uses.

Fig. 2. Elements of Users Acceptance.

Pricing for Redistribution, which explicitly sets up transfers between the different kinds of users or even between different modes.

Another critical issue in terms of acceptability is the use of the revenue collected through the charges, which concerns fairness and equity considerations. A number of unresolved arguments between economists and practitioners about how revenue should best be spent, still looms large in the pricing debate. The importance of such arguments has to do with the fact that the outcome of such decisions will ultimately determine whether a road pricing scheme is progressive or regressive in income distribution terms.

The Theory of Planned Behaviour (TOPB) is a social-psychology theory dealing with the relation between attitudes and behaviour (Ajzen, 1991). The TOPB states that an individual's behaviour can be best predicted by his intention to exhibit this behaviour. The intention in turn is influenced by three variables: the attitude towards the behaviour, the subjective norm and the perceived behavioural control. The latter can also directly influence behaviour. The extent to which the respective variables have an effect on the intention or behaviour varies between different situations and between different persons. Also attitude, subjective norm and perceived behavioural control influence one another reciprocally. In terms of the TOPB the three variables are defined as follows:

"Attitude" towards a specific behaviour in a specific situation is defined as the added products between expected behavioural consequences and the subjective valuation of these consequences.
"Subjective norm" is the perceived social pressure, that is: perceived opinions of others (about various possible behaviours in the context) multiplied with the importance of others' opinions for the individual.

78 MAX HERRY

Fig. 3. Structure of Acceptability Issues (Schlag & Teubel, 1997).

"Perceived behavioural control" is considered as the ease or difficulty for the person to exhibit a specific behaviour in a certain situation.

The theory of planned behaviour differentiates the theory of reasoned action as one expectancy-value theory (e.g. Ajzen, 1991), taking up and extending the economic rational choice approach.

Taking on the TOPB a model of road pricing acceptability (Schlag, 1997, 1998; Schlag & Teubel, 1997) was developed in AFFORD. This model defines the following essential issues determining acceptability (see Fig. 3):

Problem perception: the perception of traffic related problems (especially regarding the HGV component) is a necessary precondition for regarding problem-solving measures as important.
Mobility related social norms as the perceived social pressure to accept measures like road pricing.
Important aims (of the society) to reach by the measures: these aims can compete with certain mobility related aims of various interest groups. The potential

conflict that may arise between these possibly different aims is crucial for the question of acceptability.

Information and awareness of options: people have to know and understand planned measures. They have to be aware of the background, of the aims as well as of the specific ways in which the measures are to be implemented in practice, and of the expected impacts.
Perceived effectiveness and efficiency: the proposed measure have to be perceived as an effective and efficient mean to control traffic problems.
Equity: first of all in the sense of a distribution of costs and benefits as being fair. Here, equity is tentatively operationalised as personal outcome expectations. An important distinction is made between intra-personal, interpersonal and alternatives on which the evaluation of outcome expectations may be based.
Revenue allocation: public acceptability strongly depends on how the revenues are used. Hypothecating revenues to purposes of benefit to those being subject to the new price increases public support considerably.
Attribution of responsibility for the solution of perceived traffic problems. If people consider themselves as at least partly responsible for solving the problems (internal as compared to external attribution), this should lead to increased agreement with measures that raise the price of or restrict their use of the road space.

1.2. How to Measure Acceptability

Rienstra, Rietveld and Verhoef (1999) have described the various types of analyses, which can be applied to investigate the acceptability of transport policies. The first big divide has on the one hand theoretical models, which assume rational behaviour of individuals while on the other hand we find theories extending the economic rational choice approach and drawing on social and psychology theories. The classical rationality approach (such as cost-benefit calculation in a delegated political decision system) has come under attack in the study of acceptability, with many critics pointing out that it cannot provide enough decision support since the complexity (or "wickedness") of road pricing problems calls for a widening of the rational approach. But, the major weakness of the classical rationality approach is that it only considers "social collective" interest, i.e. the sum of costs and benefits, without looking at its distributive implications, which are explicitly considered in the other approaches mentioned in this chapter.

As such, this is not (only) due to a particular attribute (wickedness) of road pricing measures, we find it in any problem where there is change of entitlements. Thus, research has tended to stay clear from such rational models. This is where we find empirical studies, e.g. studies making use of questionnaires and interviews, 80 MAX HERRY

and ex post studies, e.g. investigation of behavioural changes of individuals due to implementation of the measure(s).

2. PRICING ACCEPTABILITY IN THE TRANSPORT SECTOR

The PATS project takes on a political economy theoretical framework to explain acceptability of implementation of pricing instruments. Success in the implementation of policy instruments in the transport sector is seen as the outcome of a (rational) political market. And, the specific interests of the various actors in such a political economic process determine this outcome.

According to the political economy explanations, questions such as:

what are the incentives of various groups of actors to organise themselves according to their specific interests;
what are the possibilities of both the actors on the demand side and the actors on the supply side of the political market, respectively, to influence the outcome of the political decision-making process; and
help explain the success of particular policy options. According to this economic approach to explain human behaviour, the outcome of this process is determined not only by the preferences and interests of the various actors but, even more importantly, by the restrictions, i.e. the institutional interactions and the institutional constraints.

For the analysis of chances for acceptance of implementation of incentive-based instruments in the transport system, it thus becomes important to identify:

the various actors involved or affected by the decision-making process;
the interests of the various participants;
the influence of the various (groups of) actors, given the specific institutional setting; and finally
the relevant aspects determining the acceptance of various means of transport policy, and particularly to the case at hand in PATS, the pricing instruments.

The major conclusion from AFFORD with impact to the issues discussed here is the importance of the issue of perception.

It follows from AFFORD that the perception of traffic-related problems is a necessary precondition for dealing with acceptability of problem-solving measures. Perception of problems influences the support for or rejection of policy measures (i.e. acceptability). And this holds for two different aspects of the perception.

On the one hand, people who feel personally affected by a (severe) problem or threat will support measures to overcome this personal problem. On the other hand, empirical evidence suggests that the evaluation of a perceived state as a "societal problem" (the so-called general problem perception), i.e. without people being personally affected by the problem, leads to an increase of support for measures to overcome the problem.

3. PRICING MEASURES ACCEPTANCE

The PRIMA project has been concerned about how to construct the "optimally" functioning decision process. One of the main objectives of this project was to search for and suggest general characteristics of decision processes that can make it possible to get public acceptance for the introduction of an urban region roadpricing scheme.

The political process can be seen as a struggle for publicly produced goods and services.

In order to understand the barriers within the decision process PRIMA draws upon three theoretical approaches to the decision process (PRIMA, 2000):

Decision support theory.
Public choice theory.
Policy games.

The decision support provided by experts, either independent experts, or experts with affiliation to an agency (or both) generally provide analysis of some decision alternatives. The politicians decide whether to follow the advice or not. This is the classical rational approach that is generally followed in decision making (such as cost-benefit calculation in a delegated political decision system) being applied to the field of transport planning. Obviously, a shortcoming is that a gap between planners and politicians can lead to that decision support not being sufficient for political approval.

The public choice theory explains some of the reasons why politicians choose to follow or not the recommendations of the analysis conducted by experts. The public choice model suggests that politicians compete for public support. The issues which politicians choose to promote are those that match the political vision of the party, and which are expected to generate sufficient support among voters.

In the case where strong conflicting interests are present, the policy process may develop to the stage of policy games. The policy games analysis assumes that some kind of strategic behaviour is present, and such strategic behaviour implies that 82 MAX HERRY

the impacts of a road pricing scheme are played out by interest groups. Since road pricing gives rise to distinct groups of losers there is a risk that the decision process will be blocked by strategic behaviour as different interest groups overstate some of the impacts.

4. RETRANSFERENCE

A main and often underestimated problem by introducing or changing IRP/HGV is the acceptance of the charge, the retransference and the conflict potential, to which this section is mainly dedicated (Herry, 1995, 2001).

By introducing or changing IRP/HGV users are in a large scale prisoner's dilemma, insofar as they do not know how the others react. The first result of this process is the initial acceptance that means the acceptance without knowing the reaction of the other users (Fig. 4).

For instance, if the IRP/HGV on a section is rather high and the initial acceptance is then low, a lot of users will first choose – for instance – an alternative parallel route without charge. Therefore, the conditions of this "alternative" will get worse and some of these "first refusers" will switch back to the tolled road. This last

Fig. 4. Acceptance Scheme at Short-Term View (Herry, 2000).

Fig. 5. Acceptance Scheme at Long-Term View (Herry, 2000).

mentioned process is called retransference, which represents a difference between short and long-term acceptance. The difference of the two levels can be rather high (Herry, 2000, Fig. 5).

4.1. Main Results on Retransference

A retransference takes place especially in form of behaviour changes of change of route, modal change, travel frequency reduction as well as substitution of destination.

For the general retransference the following can be said: The higher the IRP/HGV charge the lower is the percentage of retransference of the initial refuser. That means the absolute number of them who switch back will raise (because there were many refusers and the level of service on the free access road got very bad), but it will decrease in relative terms (because the monetary penalty for retransfering is still high).

The highest retransference takes place in case of the route change, which arise from the biggest part of behaviour alternatives in the beginning.

84 MAX HERRY

The retransference based on modal and destination change stays, relatively considered, very constant.

4.2. Conclusions Concerning Retransference

The phenomenon of the retransference is a widely underestimated process.

The retransference is not only important concerning the financial results of the IRP but it is very important for the policy handling of the issue.

The size of the retransference and with that of the conflict potential can be rather big as was shown by concrete results of case studies: 50% and more of the final users!

Studying the impacts of (new) IRP/HGV the consideration of the retransference and the conflict potential should become a standard procedure.

Studying the retransference and the conflict potential, one needs a quite well developed survey, as used for instance in the DESIRE project.

5. BARRIERS

There are different kinds of barriers regarded:

Barriers from the transport policy, legal, institutional, technical barriers, barriers concerning information and communication, behavioural, organisational, financial and other barriers.

The dependencies among these barriers are exhibited in Fig. 6.

The main constraints on implementation of optimal strategies are likely to concern legislation and acceptability. An important theme in road pricing is the identification of the institutional and political barriers to the implementation and acceptance of such (pricing) measures in Europe, and how they can be overcome. Political economy theory, public choice theory, and theory of planned behaviour are examples of different theoretical approaches that have been used to try to explain the phenomenon of public acceptance.

Although in general there does not seem to be any opposition with regard to the policy objectives associated with road pricing, in practical terms a number of schemes have faced strong opposition at the implementation stage. That result probably stems from the fact that the nature of public acceptability issues is marked by a great degree of "wickedness" (Rittel & Webber, 1973), which is to say that the degree of complexity of such problems is high and marked by

Fig. 6. Barriers of Transport Pricing and their Interdependencies (Herry, 2001).

value conflicts. Hence, a comprehensive identification of the main priorities is necessary.

6. FAIRNESS

The way a problem is identified and defined is a very real preoccupation of participants in every policy process, because the way a problem is conceptualised influences the choice set of solutions considered and, more important in this context, the acceptance of charging as one of them.

Economic theory is fond of a pricing approach to solve congestion or financing problems. But this solution has not been wholly politically acceptable since economic theory fails to go beyond efficiency objectives in the introduction of pricing. The prominence of the efficiency argument in the pricing literature narrows down the whole discussion, precluding consideration of the wider set of arguments (and objectives) in support of introduction of pricing. For road pricing to be successful (and effective) there is the need to consider a richer set of interests than just economic efficiency (Viegas, 2001). In that respect, the main consideration 86 MAX HERRY

should be effectiveness, which is to say, to proceed in a piecemeal and acceptable way towards a wider set of objectives.

7. CONCLUSIONS

Mainly three kinds of priorities should be addressed in order to contribute to increasing acceptability: pricing for financing, pricing for regulation, and pricing for redistribution. Furthermore, the use of the revenue collected through the charges is an important issue which concerns fairness and equity considerations.

There are various types of analyses, which can be applied to investigate the acceptability, on the one side theoretical models, which assume rational behaviour of individuals while on the other side are theories extending the economic rational choice approach and drawing on social and psychology theories. The classical rationality approach has come under attack with many critics pointing out that it cannot provide enough decision support since the complexity of road pricing problems calls for a widening of the rational approach. But, the major weakness of the classical rationality approach is that it only considers "social collective" interest, i.e. the sum of costs and benefits, without looking at its distributive implications. As such, this is not (only) due to a particular attribute (wickedness) of road pricing measures, we find it in any problem where there is change of entitlements. Thus, research has tended to stay clear from such rational models. This is where we find empirical studies, e.g. studies making use of questionnaires and interviews, and ex post studies.

Altogether, you can draw the following conclusions, primarily based on the PATS study (PATS, 2000):

The objectives behind transport pricing in general and the purpose of each single measure must be clear, understandable and reasonable for those affected by the measures.
In order to enhance acceptability it is important to choose the suitable/sensible pricing principle for each pricing purpose, separately.
The price has to reflect the real costs of transport.
Intermodal fairness of pricing is important although lower charges for "green" modes are accepted.1
For making pricing measures acceptable it is important that those affected perceive them as effective and suitable to solve transport-related problems.

It is clear that new measures are under this aspect more difficult to accept than well-known ones since the effectiveness of new measures is – at least for citizens – not proven. This is also one reason why attitudes "before" acceptance and "after" acceptance usually differ. The retransference reflect this issue.

The phenomenon of the retransference is a widely underestimated process. The retransference is not only important concerning the financial results of the IRP but it is very important for the policy handling of the issue.

Information on the successful implementation of new measures elsewhere may help to convince those affected by the measures that they are suitable to solve transport related problems, improving in this way the information of the effectiveness of the proposed measure.

Since obviously citizens fear negative impacts on their daily mobility and are not convinced on the positive impacts such as traffic reduction, less congestion, better living conditions and environmental improvement, governments have clearly to explain the consequences of intended pricing measures.

The appropriate and transparent use of revenues raised by pricing measures is essential in terms of acceptability. Use of the revenues in the transport sector for the benefit of the paying users is helpful. Cross-subsidisation for example of other modes can enhance acceptability but has to be explored on a case to case basis.

In terms of perceived fairness it has to be made clear that pricing measures do not lead to tax duplications.

Compensation measures for disadvantaged groups have to be considered in policy packaging.

Privacy protection is a necessary precondition for an acceptable pricing scheme applied to private vehicles. Provided that the conditions are given to properly guarantee privacy this argument does not seem to be a major obstacle for introducing pricing measures. In the case of HGVs this concern for privacy protection is of minor significance.

The charging scheme has to be transparent and its use should be easy to handle and understandable.

Pricing measures should be introduced in a stepwise way, avoiding price shocks. EU-wide harmonisation of principles applied for IRP/HGV is important in the design and implementation of pricing schemes at regional, national and European level.

Main barriers are coming from the transport policy, and are legal, institutional, technical barriers, barriers concerning information and communication, behavioural, organisational, financial. Some of the main constraints on implementation of optimal strategies are likely to concern legislation and acceptability.

An important theme in road pricing is the identification of the institutional and political barriers to the implementation and acceptance of such (pricing) measures in Europe, and how they can be overcome.

88 MAX HERRY

NOTE

However, caution should be paid with generalised decisions on lower taxes and charges for more environmentally friendly modes. Since this touches upon cross-subsidisation it has to be explored on a case-to-case basis whether and to what extent so-called "green" modes should be supported.

REFERENCES

AFFORD (2001). Economic and equity effects of marginal cost pricing in transport. European Commission Transport Research Fourth Framework Programme.
Ajzen, I. (1991). The theory of planned behaviour. Organizational Behaviour and Human Decision Processes, 50, 179–211.
Herry, M. (1995). Road pricing Evaluation of acceptance including Information feed-backing. In: Proceedings of the 23rd PTRC European Transport Forum.
Herry, M. (2000). Retransference and conflict potential by introducing tolls on intercity road infrastructure. Paper to the ECMT Round Table 118.
Herry, M. (2001, November 21–22). Barriers to interurban transport pricing. Paper to the EU-IMPRINT Europe Seminar in Brussels.
PATS (2000). Pricing acceptability in the transport sector. European Commission Transport Research Fourth Framework Programme.
PRIMA (2000). Pricing measures acceptance. European Commission Transport Research Fourth Framework Programme.
Rienstra, S. A., Rietveld, P., & Verhoef, E. T. (1999). The social support for policy measures in passenger transport. A statistical analysis for the Netherlands. Transportation Research Part D, 4, 181–200.
Rittel, H. W. J., & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy Sciences, 4, 155–169.
Schlag, B. (1997). Road pricing-Maßnahmen und ihre Akzeptanz. In: Schlag (Hrsg.), Fortschritte der Verkehrspsychologie 1996 (pp. 217–224). Bonn: Deutscher Psychologen Verlag.
Schlag, B. (1998). Zur Akzeptanz von Straßenbenutzungsgeb ¨ uhren. ¨ Internationales Verkehrswesen, 50, 308–312.
Schlag, B., & Teubel, U. (1997). Public acceptability of transport pricing.IATSS Research, 21, 134–142.
Viegas, J. M. (2001). Making urban road pricing acceptable and effective: Searching for quality and equity in urban mobility. Transportation Policy (October), 289–294.

6. ASSESSMENT OF ROAD PRICING SCHEMES BASED ON DIFFERENT NATIONAL SITUATIONS

Rene Neuenschwander and Florian Gubler ´

In the previous chapters, issues related to IRPS have been described in a general way. In this chapter, we take into consideration the different national situations that may influence the feasibility and choice of IRPS. In Section 1 we first present criteria to describe different initial situations with regard to IRPS on a national level, then sensible solutions for the different types of country situations are derived. Section 2 then shows opportunities and threats of different implementation paths.

1. SENSIBLE SOLUTIONS FOR DIFFERENT TYPES OF NATIONAL SITUATIONS

1.1. Different National Situations

In the preceding chapters technical and institutional solutions for basic forms of IRPS have been presented. This was presented in a general way, not taking account of circumstances European countries are living in, for example, their geographical location and history, their transport system and culture or other variables.

In this section a first step towards reality is made. In an overview, criteria are presented with which different types of initial country situations can be

Interurban Road Charging for Trucks in Europe Research in Transportation Economics, Volume 11, 89–107 Copyright © 2005 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 0739-8859/doi:10.1016/S0739-8859(04)11006-8

Table 1. Criteria to Describe Different Types of Initial Situations and their Relevance for Technical and Institutional Assessment.

| | Adequacy | | |-----------|---------------|--| | Technical | Institutional | |

Objectives of IRPS: These criteria describe the major (political) objectives that may be the reason to introduce an IRPS for HGV. The adequacy of the different IRPS depends on these objectives as not every IRPS is suited equally well to satisfy one or several of them.

1. Financial needs	X	X
2. Environmental concern	X	X
3. Management of traffic flows	X	X
4. Determination to introduce fair competition among modes		X

Legal and institutional framework: Especially in the short run, existing legal restrictions will restrict the feasibility of some of the IRPS.

5. Legal constraints for HGV road pricing: EU level (EC 62/1999) X
6. Legal constraints for HGV road pricing: National level X

Transport infrastructure: The quality and density of the transport infrastructure as well as the density of land use around the motorways are criteria having an important influence on the feasibility of some of the technical and institutional solutions discussed.

7. Availability of space for toll plazas/possibility for land take	X
8. Density of existing road network	X	X
9. Regional differences of density of road network	X	X

Culture: Last but not least, the cultural dimensions play an important role when determining the feasibility of different IRPS – whether it is existing experiences with motorway tolling or the quality of the public administration or the sensibility towards data protection, to name only a few.

10. Motorway tolling culture	X	X
11. Flexibility towards institutional change		X
12. Experience with private operators	X	X
13. Inclination towards fraud/degree of acceptance of regulations	X
14. Need for privacy/sensibility towards data protection	X

described. Table 1 consists of a list of 14 criteria which are divided into four groups. They describe situations with respect to technical and institutional adequacy that may differ between countries. For every criterion there are different possible initial situations on which the degree of adequacy of each of the IRPS and institutional models depends. Some of the criteria are only important to assess the institutional adequacy, others will be mainly taken to assess technical feasibility. Table 1 shows which criterion is relevant for which type of adequacy.

Since technical questions are dealt with in Chap. 3, the following discussion is restricted to criteria of institutional adequacy.

1.1.1. Objectives of IRPS

The introduction of an IRPS may have different political goals and the priorities between these goals will vary from country to country.¹ The political position vis-a-vis IRPS is strongly influenced by geographical and economic aspects that ` will be discussed in the following chapters. But already it has to be emphasized that criteria like the regional position of a country within the EU, the importance of transit traffic, the structure of a country's transport sector, the relevance of transport intensive sectors in a country as well as the volume of trade and the quality of railways and waterways will heavily influence a country's judgement of the possible objectives of IRPS discussed below.

Financial needs (Criteria 1 in Table 1): If a country has difficulties in financing the construction and/or maintenance of its motorway infrastructure, NET is the "classic" IRPS to deal with this problem. DAREA is suitable for raising revenues too. The revenue generation potential of PERM is much lower.
Environmental concern (2): If the protection of environment from transport related sources is one of the country's most important objectives, it should introduce DAREA since this IRPS prices transport on any road within the area. Network pricing is suitable as long as the problem of detour traffic does not occur.
Management of traffic flows (3): To fight congestion, DAREA is generally suitable, although not very effective on a local scale. NET may be more suitable, since it allows for pricing within a defined network. However, if transport can still switch to other roads, congestion might occur there instead. In order to fight congestion only with PERM, the cost of permits would have to be prohibitively high. But it should be noted that, in order to fight congestion, all road vehicles should be charged, not only HGVs.
Determination to introduce fair competition among modes (4): Permit pricing is least feasible, because the charge is not mileage-dependent and may be not differentiated enough with regard to vehicle types in order to contribute to fairness and efficiency. DAREA fulfils this policy goal if all vehicles are charged, but it is difficult to determine the right differentiation of fares. NET is suitable as long as detour traffic will not occur.

1.1.2. Legal and Institutional Framework

Legal constraints for HGV road pricing on EU level (5): For EU member states, current EU-legislation makes it impossible to introduce any form of distancedependent area pricing for HGV in the short term. Network pricing can only be applied above the 12 tonnes weight limit, with the exception of a motorway concession where all vehicles are charged. The possibility to differentiate

between different vehicles is restricted. PERM, however, corresponds to the Eurovignette in use and is therefore less restricted by EU-legislation.
Legal constraints for HGV road pricing on national level (6): Sophisticated forms of IRPS like DAREA and NET will not be feasible without a national legislation allowing to introduce them. Rather simple schemes from the point of view of procedures like PERM will probably need less changes in national legislation.

1.1.3. Transport Infrastructure

Density of existing road network (8): DAREA is particularly suited for countries or regions with a dense non-motorway network of good quality adjacent to the motorways. On the other hand, network pricing is problematic since HGVs can easily dodge road pricing by switching to freely accessible trunk roads and cause additional congestion there. The feasibility of PERM is similar to NET, since a permit is needed only on motorways.
Regional differences of density of road network (9): DAREA is suitable for countries with a heterogeneous road network density, as HGV are charged independently of the road type. With NET, detour traffic could become a problem in regions with high density of road network of good quality. NET pricing could lead to regional injustice since transport in regions with low density would become more expensive than in more developed parts of the country.

1.1.4. Culture

Motorway tolling culture (10): A country that has been tolling its motorway network for many years will not easily change to DAREA pricing. Having no motorway tolling culture will lower the feasibility of NET compared to countries with an existing motorway tolling culture. Since PERM is the mildest form of road pricing, the absence of a tolling culture will not be too large an obstacle for the introduction of this IRPS.
Flexibility towards institutional change (11): In countries with a tendency for quick institutional changes, the introduction of any IRPS is, at least, not hampered by this factor. As for the other case, we assume PERM to require generally less institutional change and therefore to be more feasible in such countries.
Experience with private operators (12): The feasibility of public models is higher in countries with little or bad experience with private operators. On the other hand, it might be more difficult to introduce a public model IRPS if solutions with private operators already exist and is satisfactory. If the experience with private operators is high and predominantly good, private models could be preferred for future IRPS. Institutional models with mixed responsibilities of public and private bodies will be less feasible in countries with little or bad experience with private operators.

1.2. Deriving Sensible Solutions for Different European Contexts2

Sensible solutions include three main aspects: The choice of the basic form of IRPS and of the technical and institutional model.

1.2.1. Deriving the Preferred Form of IRPS

It is important to differentiate between what is possible in the short term and what would be desirable in a long-term perspective.

In the short-term perspective two criteria describing initial situations were identified as being most relevant: Legal constraints on European and national level and the existence of a motorway tolling culture. Taking these criteria into consideration, NET tolling is the most sensible starting point for all EU countries.

From a long-term perspective other criteria were identified to be more important to derive the preferred IRPS, especially the need to finance road infrastructure and the density of the road network. Figure 1 shows the general influence of these two criteria on the choice of the basic form of IRPS. In this context, "financial needs" are

Fig. 1. Decision Tree to Derive the Basic Form of IRPS in a Long-Term Perspective.

understood always in connection with the expansion/maintenance of the motorway network. Therefore a "yes" to this criteria is a first, but not final indication for NET (distance dependent network pricing). A "no" leaves the choice of the IRPS rather open. Secondly, the density of the road network of good quality determines if network pricing could cause undesired detour traffic. Especially, if tolls are applied in a congested region, and this congestion occurs not only on the motorway but also on the road network around it (for instance in a densely occupied corridor or in a large peri-urban agglomeration), a NET scheme involves significant risks of inducing high levels of detouring traffic to free access roads, which normally run closer to urban areas, thus strongly increasing external costs (local environmental aggressions and accidents). In such cases a distance dependent area pricing scheme – DAREA – is the appropriate option, although not the only solution, as preventing detour traffic can be made through other types of intervention, namely traffic engineering measures that make crossing of urban areas very inconvenient (e.g. itinerary restrictions and/or time restrictions).

1.2.2. Deriving Feasible Institutional Models

A similar figure can be drawn with regard to the institutional model (see Fig. 2). Here, we present the rationale behind opting for a particular institutional model. Even more than in the previous discussion on the choice of the IRPS, the following

Fig. 2. Decision Tree for Deriving the Preferred Institutional Model. (See Viegas et al., 2002, p. 19).

decision tree has to be understood as showing the tendency towards a certain model. Thereby, "Perceived risks of inefficiency" means the potentially lower efficiency of Public Administration agencies in controlling and operating a tolling operation (organisational and productive efficiency) in comparison with a private agency. "Risks of capture" denotes the possible restrictions of the policy-maker created by a road charging concession to a private organisation: As the concession model implies some assumption of traffic risks by the concessionaire, for a certain number of years (depending on the length of the concession) the State gives away important instruments of transport policy, such as fixing toll prices or taking investment decisions, which should be weighted against the advantages of engagement of the private party in a concession contract. But there may be engagement of a private party in a more traditional service contract, through which this party simply runs the operation of collection of the charges, delivering them to the State, and keeping transparent accounts of all the transactions, and does it for a fee, which may be fixed per time period or dependent on the number of charging transactions. In such cases, there is no significant capture of the State by the contract.

One of the main historical reasons justifying application of tolls to interurban roads is the existence of financing needs beyond current State budget capability. The financing objective traditionally favours the choice of an All private institutional model.

If there is no significant investment need beyond State capability for specific road infrastructure, the choice of the institutional model is basically open. It will, in most cases, be made between the Mixed and the All public models. The choice then depends on the balance between the perceived risks of inefficiency of operation by public agencies versus the public capacity to adequately control the concession operations by private parties.

If perceived risks of lower efficiency of Public Administration agencies in controlling and operating a tolling operation (organisational and productive efficiency) are dominant, the option will be for a Mixed model, or as a second choice, for the All private model. However, the risks of capture of Public Administration by a private concessionaire should also not be forgotten. If such risks of capture are considered more significant than the lower efficiency of Public Administration, then the All public model, or as a second choice, the Mixed model should be chosen. Another significant problem similar to the "perceived risks of inefficiency" is enforcement. This problem occurs in situations without a direct private revenue interest, particularly in "open" IRP systems (i.e. without toll barriers). The two most obvious cases are prosecution of foreign vehicles and backoffice prosecution costs in relation to "free riders." A solution for such problems

Fig. 3. Interdependencies Between Institutional Model and IRPS Option.

is based on writing contract incentives that approach the interests of the Principal (the State) and of the Agent (the concessionaire), in order that it is their common interest that prosecution of violations is enforced. Difficulties in this area may also help the decision fall into the All public model.

We conclude that there are certain interdependencies between institutional model and basic form (see Fig. 3, where the size of the dot corresponds to the estimated interdependency).

The All private institutional model is naturally linked to the NET IRPS as this institutional approach is best suited to finance privately motorway network expansion projects. However, the NET scheme lends itself well to all three institutional models.

In the case of DAREA, fine-tuning of prices with respect to traffic management goals will probably be such a recurrent need, that the All private model would be virtually impossible or require very innovative (much more open) concession contracts. Adoption of a Mixed model for the DAREA scheme is possible and without inconvenience, as long as the remuneration of the private operator is not based on the actual tolls collected (which would lead to segmentation to maximise revenue) but on the success achieved towards the desired traffic distribution. An All public model for DAREA is possible, but the complexity of the scheme increases the risks of higher costs due to inefficiency.

The PERM scheme, with its simplicity and much lower connection with the actual extension of road access exercised, corresponds to a flat charge, almost like a tax, and is thus naturally associated in the popular mind with an All public model. Moreover, its comparatively small administrative burden and very little investment requirements do not give strong arguments in favour of the involvement of the private sector.

2. EVOLUTIONARY PATHS

2.1. Definition of Three Basic IRPS

Three basic IRPS are defined, thereby differentiating between the initial situation and possible evolutionary paths (Table 2). Among the countries analysed within the DESIRE case studies, five countries are tolling their roads for HGV, whereby only Switzerland according to a DAREA scheme; Austria (since 1.1.2004), France, Hungary and Portugal have a NET pricing solution. Other countries do have just PERM pricing for HGV at the moment (Germany, Netherlands and Nordic countries). Germany and the Netherlands are recognized as "mature" countries

Table 2. Three Basic IRPS and Adequate Evolutionary Paths.^a

| Element of IRPS | Present | Future | | | | |--------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------|--|--|--| | Countries with a motorway tolling culture (Austria, France, Hungary, Portugal) | | | | | | | Basic form (F) | NET | NET, DAREA | | | | | Technical solution (T) | Roadside based | Extend to vehicle based
autonomous system,
Fulfil/achieve international
standards, roaming | | | | | Institutional solution (I) | Mixed | Mixed, All private | | | | | Pricing principles (P) | Flat rate | Differentiation (time,
emissions), higher tariffs in
sensitive areas | | | | | Countries with DAREA already introduced (Switzerland) | | | | | | | Basic form (F) | DAREA | DAREA | | | | | Technical solution (T) | Roadside based | Extend to vehicles equipped
with autonomous OBU,
extend to vehicle based
autonomous system, roaming | | | | | Institutional solution (I) | All public | All public, mixed | | | | | Pricing principles (P) | Flat rate, emission differentiated | Higher tariffs in sensitive areas | | | | | | "Mature" countries shortly before introduction of IRPS (Germany, the Netherlands, U.K.) | | | | | | Basic form (F)
Technical solution (T) | PERM
Vehicle based (Vignette) | DAREA, NET
Extend to vehicle based
autonomous system,
fulfil/achieve international
standards, roaming | | | | | Institutional solution (I)
Pricing principles (P) | Public, mixed
Flat rate, emission differentiated | Mixed, all private
Higher tariffs in sensitive areas | | | |

aSee TIS.pt (2003, p. 49).

shortly before the introduction of an IRPS within the next few years. The U.K. could also be classified as a mature country, as there are plans to introduce a DAREA-scheme within the next years.

To assess the possible evolutionary paths, a SWOT analysis was carried out, answering the following question: What are the opportunities and threats of a change in the IRPS from the starting situation towards the possible evolutions? Criteria used for this SWOT analysis include political objectives, economic and mobility effects, costs, reliability and user friendliness of an IRPS. In the following sections the results of the SWOT analysis are summarised in a table for each of the three initial situations. Afterwards the pros and cons of the possible evolutionary paths are discussed.

2.2. "Mature" Countries Shortly Before Introduction of IRPS

These countries now have no pricing scheme for HGV at all, or they have a driving permit system (PERM) where a charge is levied for the permit to drive within a certain perimeter. The pricing solution is a flat rate by time period, differentiated only by emissions. The opportunities and threats of changing the PERM scheme towards a distance-dependent area pricing scheme are presented in Table 3.

In the following section, we answer several practical questions on the basic form and the pricing system of a change to a DAREA solution (a). Since countries can also move to a NET solution first, the same issues are treated with regard to this solution (b).

2.2.1. Why Choose DAREA as the Basic Form of Pricing?

2.2.1.1. Does DAREA help finance the road network?. Generally yes, but this is not the main reason for a DAREA scheme. Although the potential revenue of a DAREA road pricing scheme is notably higher than that of a NET or of a PERM solution (or of no road pricing, of course), in a traditional NET scheme tolls are clearly aimed at financing the construction of the motorways where they are charged. Especially countries with a dense non-motorway network could increase their revenue significantly because all roads are subject to charges and detour traffic is not possible. Of course, a great deal of the financial success of the introduction of DAREA depends on the level of the fee and on how much the government can withstand political demands to lower other taxes and fees correspondingly.

2.2.1.2. Does DAREA help reduce congestion?. Only under certain circumstances. With flat tariffs, a DAREA scheme is generally able to reduce the burden on the whole network but not to reduce congestion on specific spots or stretches

Table 3. Summary of SWOT Analyses for Countries Shortly Before Introduction of IRPS.

| SWOT Analyses: "Mature" Countries Shortly Before Introduction of IRPS | | | | |-----------------------------------------------------------------------|---------------------------------------|--|--| | Strengths | Opportunities | | | | Neutral to safety | Revenue potential | | | | No influence on traffic flow | Traffic management | | | | Low costs | Add-on services | | | | Simple and easy to understand | Addressing environmental issues | | | | Anonymity for users | Improvement of road safety | | | | Low-tech | Low operation costs | | | | | Motivation to become an equipped user | | | | | OBU can be paid by user | | | | | Mature technology (DSRC) | | | | | Secure against fraud | | | | | Several enforcement possibilities | | | | | Interoperability | | | | Weaknesses | Threats | | | | Low revenue potential | Substantial initial investment | | | | Unsuitable for fighting congestion | Complex pricing scheme necessary | | | | Low effectiveness for environmental improvement | Undesired detour traffic | | | | Inflexible for traffic management | Legal issues on mandatory equipment | | | | Dependence on PoS (Point of Sale) | Handling non-equipped users | | | | Vignette can be manipulated | Extra costs from authorised staff | | |

of the network. However, DAREA can be a very appropriate instrument to fight congestion if it is combined with a flexible technical solution (GPS autonomous solution) that allows differentiated charging according to the congestion situation. But as yet such a technical solution is not yet in operation on a large scale.

2.2.1.3. Does DAREA help reduce the environmental burden?. Compared to a PERM solution or no pricing, DAREA is much more suitable to achieve the objective of environmental care since the whole network is being priced. Under a PERM pricing scheme, the average cost per vehicle-km will even decrease with a higher traffic volume, which makes PERM a non-effective instrument of environmental policy, despite the possibility of price differentiation with respect to emission categories. On the contrary, DAREA sets a clear incentive to reduce the number and length of trips and/or to use the fleet more efficiently. Additionally, the differentiation of the charge with regard to the emission category of the vehicle will encourage the use of cleaner vehicles. Advanced forms of DAREA could even allow charging more in environmentally sensitive areas (e.g. the Alps).

2.2.1.4. Does DAREA improve road safety?. Probably yes, in a small measure. Whereas PERM pricing is more or less neutral towards road safety, the introduction of DAREA pricing will probably reduce traffic volumes of HGV or at least slow its growth. This is a generally favourable development with regard to road safety. On the other hand, there is an incentive to make trips shorter, by using straighter connections on trunk roads, thus avoiding longer trips on motorways. This is negative to road safety, since traffic on motorways is generally safer than on mixed roads in settled areas. This effect could be limited if the DAREA scheme would have lower tariffs on motorways, but this is contrary to logic, and thus have a difficult political acceptance, as prices would be lower where the service received is better.

2.2.2. What are the Main Aspects of a Pricing Structure for DAREA?

The EU directive 1999/62/EC does not allow a general pricing of the entire secondary road network. Exceptions are allowed in case pure NET pricing systems cause considerable problems with decreasing traffic safety and congestion. Consequently, EU legislation does not provide a framework for calculating DAREA charges.

A direct transition from no road pricing or PERM systems to DAREA will impose a drastic increase of costs for hauliers, which is even higher than the introduction of NET as: (1) average costs per vehicle-km in DAREA are higher than in NET; and (2) the possibility of avoiding tolls by using the secondary road network is removed. This strong increase of operation costs for the hauliers would in most cases have to be compensated by a reduction in their fixed costs, to improve the level of acceptance. Of course, changing the fixed costs in one country (where DAREA tolling is introduced) and not in others (where that is not the case), raises the question of cost aggravation especially for the hauliers of those other countries, which can lead to resistance not only at domestic level but also at the European level, with accusations that such tolls effectively constitute barriers to trade.

On inter-urban roads the lower traffic volume and lower construction standards on secondary roads leads to much higher average infrastructure costs than is calculated for the motorways. This relationship is supported by results of the UNITE project for Germany 2005, where average HGV infrastructure costs are twice as high on the entire network (0.21 Euro/km) than on motorways (0.09 Euro/km). This ratio might be somewhat different in or near agglomeration areas, where high traffic volumes decrease average costs. Average and marginal costs of air pollution and noise disturbance are also considerably higher on secondary roads than on motorways as they are closer to built-up areas. For the costs of traffic accidents, global warming and congestion no clear statements on the difference of costs between the two network categories can be made.

Also in case the basic motive of setting DAREA charges is environmental protection or the financing of particular projects, the resistance of toll levels against legal claims and their acceptability will be drastically improved if they are based on a cost calculation.

If tariff levels were to be based on actual costs imposed by users, the tolls to be raised on the secondary road network would considerably exceed pure motorway charges. While for reasons of cost efficiency this structure might be appreciated, it contradicts the principle of prices to reflect the quality of supply. As the latter would mean to have higher tolls on motorways, the compromise of introducing equal prices on all road categories is intuitively fair and minimises market distortions.

A differentiation of charge levels by regions is more justified under DAREA conditions than under NET. In particular when introducing DAREA pricing from PERM, the high charge levels and the missing possibility of avoiding payment by detouring the motorways might cause considerable distortions in economically weak regions. Accordingly, a regional differentiation, possibly in the secondary network only, should be considered.

The differentiation of prices by environmental standards of the vehicles and by the sensitivity of regions will enhance the sustainable development of the transport sector and is thus recommended in case the motivation of introducing DAREA charges leads into the same direction. However, a high level of tariff differentiation according to the environmental sensitiveness of the region may be interpreted as barrier to movement through that region and be subject to strong opposition given the uncertainties of the cost estimation process.

2.2.3. Why Choose NET as the Basic Form of Pricing?

2.2.3.1. Does NET help finance the road network?. Generally yes. The potential revenue of a NET pricing scheme is significantly higher than of a PERM solution (or of no road pricing at all, of course). In fact, NET is the "classic" IRPS chosen by countries with a primary need to finance their transport network in most cases charging all vehicles. This is because the NET scheme is based in most cases on a relatively cheap and "off-the-shelf" technical solution that is in operation in almost 30 countries all over the world. (DSRC/road based system). Of course, the potential increase in revenue also depends on the level of the tariffs. The financing objective is endangered if detour traffic on the non-priced road network is a valuable alternative for the drivers. As a consequence, NET is probably not an optimal solution for countries with a dense secondary road network of good quality. DAREA would be a more sensible solution in this case.

2.2.3.2. Does NET help reduce congestion?. Partially. NET is strong in reducing congestion on specific stretches of the priced motorway network, especially if tariffs are higher at peak hours. If congestion is shifted to other stretches of the motorway network, they could be integrated in the existing network of priced road infrastructure. On the other hand, NET is weak on managing traffic flows within conurbations since the possibility of detour traffic exists.
2.2.3.3. Does NET help reduce the environmental burden of transport?. Only to a limited extent. Of course, NET pricing is a way of internalising at least some of the external costs of transport and is therefore likely to reduce environmental aggression from HGV transport to some extent. Any distant-dependent pricing scheme – even with flat tariffs – sets incentives for fewer trips, especially less deadheads. However, if detour traffic on the non-priced road network is a valuable alternative for the hauliers, this will probably result in longer distances per trip and consequently in increased consumption of fuel. Especially in the likely affected densely populated areas, increasing noise and pollutant emissions are an even bigger problem.
2.2.3.4. Does NET improve road safety?. Unlikely, except to the extent that NET tolling allows earlier availability of motorways in a country: motorways are clearly safer than common two way roads. It has often been claimed that tolled NET roads are better controlled (by means of toll plazas, etc.) than the remaining motorway network. A positive relationship between the introduction of a NET pricing scheme and road safety could not be proved so far. In addition, detour traffic is likely to make the uncharged network less safe due to increasing HGV traffic.
2.2.3.5. Other reasons for choosing NET?. With the EU legislation in force (1999/62/EC), EU member states that want to introduce road pricing for HGV's are allowed to do so only on their motorways (Art. 7/2).

2.2.4. What are the Main Aspects of a Pricing Structure for NET?

According to EU directive 1999/62/EC motorway tolls are to be set to cover average costs of construction, maintenance and operation of motorways or similar roads caused by HGV's above 12 ton gross vehicle weight. In countries without motorways, the primary network relevant for international traffic is meant. In case of private companies involved in the financing and/or the operation of road infrastructure, the full costs of these companies, including certain profit margins and compensation for the risks undertaken, are to be considered in the costs underlying the road user tolls.

The first motivation for introducing NET pricing systems (or for replacing an existing PERM-system by it) is most probably related to fiscal needs. Further, the introduction of NET pricing might be driven by welfare-oriented goals, possibly as an intermediate state towards a DAREA system. If the improvement of social welfare is in the foreground, prices can be differentiated by vehicle characteristics and/or regional settings, possibly using marginal cost information.

According to recent studies, charge levels may range between 0.11 Euro/km (UNITE-Results for Germany) and 0.26 Euro/km (Tariff calculation for Austrian primaries). Current motorway charges of most countries range within this spectrum, but exceptions are possible (Italy: 0.09 Euro/km).

Strictly following the rules set out by the EC Directive 1999/62, motorway tolls for HGV's would vary considerably between countries depending on: (a) local construction costs; (b) traffic volumes; and (c) on the methods used for calculating and allocating infrastructure costs. As the calculation methods are not specified in the directive, the Member states are provided with some degree of freedom for price setting. However, the most important factor determining average costs is the traffic volume accommodated by the road network, which may cause charges to get particularly high in sparely populated regions, unless charges are averaged over wider geographical areas, allowing high traffic parts of the network to crosssubsidise the low traffic parts.

Including average environmental costs would raise tariff levels by 100% or more in case all generally accepted categories of external costs (accidents, air pollution, global warming, noise disturbance) are considered. However, the uncertainties on the level of these costs are very high and in general it is not recommended to add up all cost categories to a single price as the incentive structure would then vanish.

Compared to average costs, marginal costs of road infrastructure for heavy goods vehicles range between 20 and 30% of average infrastructure costs. A similar ratio holds true for the external costs of noise disturbance and the ratio is even more diverse for the marginal external costs of traffic accidents. In contrast, for air pollution and the risk of climate change marginal costs come close to average costs. Consequently, marginal social cost pricing would end in a major deficit for these cost categories. Only in the case of traffic congestion marginal costs are well above average costs and thus MSCP (Marginal Social Cost Pricing), especially in dense areas, recovers total costs.

Directive 1999/62/EC does not forbid a regional differentiation of motorway tolls. This might be done in order to reflect cost levels and thus to include an element of allocative efficiency, or take account of regional characteristics and economic requirements (support of poor regions by lower tolls). The prices may in particular vary when different operators are involved in different parts of the network.

Motorway charges are also allowed to vary by time of day, whereas the highest toll level must not be more than 100% above that of the cheapest time period. Further, the price structures might vary with vehicle characteristics, such as weight and environmental standards. With these permits the price-setting authority is provided with tools to fight congestion, to enhance the environmental standard and the road damage potential of the vehicle fleet.

2.3. Countries with a Motorway Tolling Culture

The IRPS basic form in these countries is a distance-dependent network pricing scheme (NET) which is based on the kilometres driven on a net of interrelated motorways of the same hierarchy-level. The pricing solution most commonly is a flat rate. The possible future system is a distance-dependent area pricing scheme (DAREA). Opportunities and threats of a change from NET towards DAREA are presented in Table 4.

2.3.1. Possible reasons to change from NET to DAREA

2.3.1.1. Does DAREA help finance the road network better than NET?. This is not certain. If a country started road pricing with a NET scheme, probably detour traffic was not a problem at that time. However, due to increasing congestion levels, network density or other reasons, drivers avoiding the tolled network for the free alternative might in the meantime have become a problem with negative consequences for the overall toll revenue. Changing to a DAREA pricing where all roads within a certain area are subject to charge might then be a suitable answer to this problem. Detour traffic would no longer be generated, thus avoiding the negative externalities and losses in toll revenues.
2.3.1.2. Does DAREA reduce congestion more than NET?. Under certain circumstances yes. In dense road networks a switch to DAREA could improve the situation, especially if DAREA is combined with a flexible technical solution that allows for differentiated charging according to the congestion situation. For example, congested parts of the network could temporarily be charged higher in order to set incentives either to switch to other roads or to other times.
2.3.1.3. Does DAREA reduce the environmental burden more than NET?. Probably yes. Compared to a NET solution, DAREA is much more suitable to achieve the objective of environmental care since the whole network is being priced. DAREA sets a clear incentive to reduce the number and length of trips

Table 4. Summary of SWOT Analyses for Countries with a Motorway Tolling Culture (NET).

SWOT Analyses for Countries with a Motorway Tolling Culture: Why Change the Basic Pricing form NET to DAREA?

Strengths	Opportunities
Current system is well-known	Increased revenue generation
Congestion can be reduced to some extent	Internalisation of external costs
External costs can be internalised to some extent	Improved road safety
Enforcement is rather simple	Increased technological flexibility
System is simple and robust	Interoperability
Transparent	Private-public partnership
Flexibility regarding institutional solution	Traffic management
Fair revenue generation is possible	Efficient congestion fighting
Low communication costs	Lower operation costs
Adequate level of data security and privacy	Increased user motivation
	Several possibilities for designing a transparent and
	user-friendly system
	Data security
Weaknesses	Threats
Detour traffic may occur	Substantial initial investment
Management measures inadequate	Termination of current concessions – alternatively
	concurrent operation of different systems
Full internalisation of external costs impossible	Manufacturers influence on the technical solution
Acceptance issues	Risk of private partners bankruptcy
Institutional flexibility	Maintaining government control
Maintaining overall quality of the road	Inexperience with the system reliability and
network	operation in general
Operation costs	Legal issues regarding mandatory equipment
Impossible to eliminate fraud	Uncertainty of future cost of GPS-signals and CN
	Acceptance in general
	User pays for operation costs leads to higher tariffs
	Complexity of pricing scheme affects user
	friendliness and transparency
	Impossible to eliminate fraud
	No guarantee for anonymity

and/or to use the fleet more efficiently. In both charging forms, the differentiation of the charge with regard to the emission category of the vehicle is possible and will encourage the use of cleaner vehicles. Advanced forms of DAREA could even allow for charging more in environmentally sensitive areas (e.g. the Alps).

2.3.1.4. Does DAREA improve road safety more than NET?. With NET, there is always the danger of detour traffic to the non-motorway road network. A change to DAREA results in a reduction of detour traffic of HGV on smaller roads. Such a shift will improve road safety in two ways: (a) Non-detoured HGV trips will be safer because traffic on motorways is more secure; (b) the remaining (passenger) traffic will be safer because of fewer HGV on secondary roads.

2.3.1.5. Is a combination of NET and DAREA schemes a reasonable solution?. For countries with strong regional differences in population density (and therefore in road network density as well), a combination of the NET and DAREA schemes might seem as a way to combine the advantages of both schemes. In the rural, sparsely populated regions only motorways would be charged, whereas in the urban, densely populated areas all roads would be subject to charge in order to prevent detour traffic. However, since the fiscal regime of motor vehicles is identical in all regions of the same country, this combined scheme could give rise to equity concerns between the regions, since the citizens and companies of the DAREA priced areas are negatively discriminated with respect to the rest of the country. On the other hand, this unequal treatment of regions could be acceptable under certain conditions. First, if the DAREA region is clearly richer than the rest of the country (no additional equity problems, but rather a form of compensation). Second, if the tariffs applied for the DAREA scheme are lower than on "NET-tolled roads." Third, if the revenues from the DAREA scheme are used for specific transport infrastructure investments in the DAREA region (e.g. for the development of road and public transport infrastructure).

2.3.2. What are the Main Changes of the Pricing Structure?

The motivation to develop from a NET pricing scheme to a DAREA scheme can either be to reduce the negative impacts of NET prices (detours of traffic to secondary roads), to enhance the basis for setting incentives for a more sustainable transport development or simply to increase revenues.

The average prices per vehicle kilometre on the road network will be higher under DAREA conditions than under NET conditions as the number of vehicle kilometres on the secondary road network usually are much lower than on the motorways. For example, the lower construction costs of the secondary network per road kilometre are over-compensated by the lower traffic volume. Apart from the higher average costs, the cost increase from NET to DAREA is considerable as the possibility to save tolls by using the secondary road network is removed.

The marginal costs of infrastructure use will be higher on secondary roads than on motorways as they are generally designed to accommodate less traffic. Average and marginal costs of air pollution and noise disturbance are also considerably higher on secondary roads than on motorways as they are closer to built-up areas.

The price increases between NET and DAREA will be different between countries, depending on the density of the various network types and on the traffic volumes. Therefore, no general rule can be given how prices develop when changing the pricing scheme.

NOTES

1. For the discussion of the economic foundation of IRPS, see for example Newbery (1998), Rees (2000) and ECMT (2002).
- 2. For a more extensive discussion, see Ecoplan (2001).

REFERENCES

ECMT (2002). Tolls on interurban road infrastructure, an economic evaluation. Summary by J. M. Viegas, pp. 149–156, ECMT round table 118, Paris.
Ecoplan (2001). Possible designs of inter-urban road pricing schemes for heavy goods vehicles, Deliverable 2 of DESIRE, project of 5th Framework Programme (2000 – CM. 10501), July 2001.
Newbery, D. M. (1998). Fair and efficient pricing and the finance of the roads. University of Cambridge. Rees, J. H. (2000). Das Konzept der sozialen Grenzkosten fur eine einheitliche Geb ¨ uhrenregelung ¨ im Verkehrsraum der EU. In: Deutsche Verkehrswissenschaftliche Gesellschaft eV. (Hrsg.) (2000), Grenzkosten als Grundlage f ¨ur die Preisbildung im Verkehrsbereich, S. 10–23, Bergisch Gladbach.
TIS.pt (2003, January). Identification of best functional schemes and implementation issues, Deliverable 5 of DESIRE, project of 5th Framework Programme (2000 – CM. 10501).
Viegas, J. M., Antunes, J., & Neuenschwander, R. (2002). Deriving sound institutional solutions for inter-urban road pricing schemes in Europe. Paper presented at Transportation Research Board Annual Meeting 2002, Washington, DC.

7. POSITION AND RECENT TRENDS IN EUROPEAN COUNTRIES

Chiara Borgnolo, Louise Stewart-Ladewig and Rene Neuenschwander ´

ABSTRACT

This chapter presents a picture of the latest developments of Interurban Road Pricing Schemes (IRPS) in Europe. Specific positions are identified as they are emerging within the two basic options to charge for access to road infrastructure in Europe: either to levy tolls to fund motorway operation and development or make commercial vehicles pay for the use of certain roads.

The review is focused on issues/circumstances in the transition from timebased to distance-based infrastructure charges in Switzerland, Austria and Germany.

Positions of countries located at the periphery of what is becoming the core of a Trans European Motorway network in continental Europe are taken into account on the basis of the planned British scheme for a Lorry Road User Charge.

Rene Neuenschwander and Louise H. Stewart are the authors of the sections on Swiss and German ´ EFC for HGV respectively.

1. FROM TIME BASED TO DISTANCE BASED ACCESS CHARGES

From the mid-1990s, in almost all European countries where motorway development has been traditionally funded through general taxation/public budget, the purchase of a permit was made mandatory for both domestic and international commercial vehicles using national motorways for a given time period (year, month, week or day).

The adoption of vignettes in Europe was pioneered by two EFTA countries – Switzerland and Austria – both determined to cope with quite sustained transit traffic across Alpine motorway road corridors. In Switzerland the measure was complemented with a ban for vehicles over 30 tonnes. In Austria, the adoption of a time based vignette was jointly implemented with ECOPOINT, a scheme conceived to allocate transit quotas against emission targets.

Since 1995, the purchase of the vignette was progressively made mandatory for goods vehicles "having a maximum permissible gross laden weight of no less than 12 tonnes" using motorways in Germany, Holland, Belgium, Luxembourg and Denmark. Sweden had joined the Eurovignette club in 1998. Each national scheme is framed in accordance with key provisions of Council Directive 1999/62/EC on the charging of heavy goods vehicles for the use of certain infrastructures.

The so-called Eurovignette Directive was adopted to complement the creation of a single market for road haulage with a framework to harmonise fixed taxes and infrastructure fees levied by member states. While ruling that tolls and user charges may not discriminate, directly or indirectly, on the grounds of the nationality of the haulier or origin destination of the vehicle, the Directive:

Establishes minimum rates (Euro per year) for fixed taxation components levied by member states – having basically the nature of vehicles excise duties and/or motor vehicle licences.
Defines the type of network "motorway or dual carriageway roads specially designed and built for motor traffic" – where tolls can be levied on all motorised vehicles and user charges can be levied on good vehicles that are registered in a different country than the one where the trip is undertaken.1
Establishes *maximum rates*for user charges (Euro per year, month, week or day).
Defines criteria for setting tolls i.e. payment of a specific amount for a vehicle travelling the distance between two points of the infrastructure.

In the absence of a Community framework for tolled motorway concessions, criteria for weighed average tolls proved quite a broad umbrella to avoid discussing the variety of approaches to determine distance-based tolls for existing motorway concessions used by concerned member states.² The criteria specified in Box 1

Box 1: Criteria for setting and charging distance based tolls and time based user charges in Directive 1999/62/EC (EU, 1999).

*The weighted average toll (art 7(9))*shall be related to the costs of constructing, operating and developing the infrastructure network concerned; the amount shall be based on distance travelled between two points on the infrastructure and the type of vehicle. Without prejudice to the weighted average tolls, provided that toll is not more than 50% above the toll charged for equivalent vehicles with the strictest emission standard or 100% above the toll charged during the cheapest period of the day, tolls can be varied in accordance to vehicle's emission standard and period of the day. In either case, fine-tuning of tolls must be "proportional" to the objectives such measures aim to achieve.

Time based access charges (including administrative costs) shall not be higher than maximum rates the Directive establishes for two classes of vehicles (up to three axes and above) and three "EURO" classes to differentiate "emissions of gaseous and particulate pollutants from diesel engines." User charges rates shall be in proportion to the duration of the use made of the infrastructure (year, month, week or day). A member state may provide that vehicles registered in the member state shall be subject to user charges for the use of the whole network in its territory; it may also apply only annual rates for domestic vehicles.

for relating tolls to the "costs of constructing, operating and developing the infrastructure network concerned" apparently proved not to be sufficiently rooted on sound economic and accountancy principles to trigger the transition from time-based permits to distance based fees levied on Heavy Good Vehicles. With reference to principles that are the building block of Community dossiers for fair and efficient pricing, provisions are limited to the fact that the Directive "shall not prevent" a member state from:

Extending tolls and user charges to other sections of the primary road network where there are safety reasons for doing so or in the absence of a coherent network of dual motorway/dual carriageway.
Applying regulatory charges specifically designed to combat time and place related traffic congestion.
Attributing "a percentage" of revenue from tolls and access charges to environmental protection and balanced development of transport networks.

With the exception of the road pricing system planned in the last decade in the Netherlands, none of the member states has considered so far to differentiate charges according to congestion.

Only the Swiss scheme has been implemented to levy distance-based charges on commercial vehicles also on other roads than motorways. Despite having been conceived independently from Community regulation, the Swiss approach to reinvest the revenue from road charges in favour of a more environmentally friendly transport was eventually assumed as a relevant model for interurban road pricing in both the Commission White Paper on European Transport Policies up to 2010 and the Costa Report to the European Parliament (INI, 2000).

1.1. Electronic Fee Collection of Distance Based Access Charges

Since January 2001, quite a sophisticated system is in operation to charge commercial vehicles as a function of vehicle attributes and distance performed on the whole Swiss road network.

In January 2004, an electronic system began operation to charge HGV for the use of the Austrian motorway network. Within a ten year contract, EUROPPASS has been implemented and is operated by a consortium led by Autostrade Spa, the main motorway concessionaire in Italy who developed the TELEPASS system for EFC at the beginning of the 1990s.

In the same year, Germany has also passed legislation to substitute Eurovignette with charges levied in accordance to mileage of commercial vehicles. As shown in Tables 1 and 2, comparing key features of the three systems, only Germany has chosen to rely on a GPS based autonomous system.

Prior to a review of early impacts from the Swiss scheme and problems met in the adoption of TOLL COLLECT in Germany, key features of new systems for EFC for HGV are summarised below.

1.1.1. National Taxation or Tolls?

In all the three countries, the decision of substituting vignettes with distancebased fees was accompanied by intensive political discussion. The engineering of a time/distance conversion scheme has involved Ministry of Transport and Ministry of Finance/Treasury. As a rule, regional governments have been very active, and quite successful, in negotiation a share of the revenue generated by new distance based charging schemes.

In Germany, the Federal States exerted great pressure on the government to dedicate revenues from road pricing for the maintenance, upgrading and construction of federal transport infrastructure rather than flowing to the national budget. This campaign was very successful, the earmarking of revenues is part of the new law for the introduction of road pricing to Germany.

	G er m an y	A ia tr us	Sw i la d tz er n
To l l r ire ts eq u m en	1 2 is i b le h ic le ig h t p t er m s ve w e	3. 5 is i b le h ic le ig h t m t ax . p er m s ve w e	3. 5 Pe is i b le h ic le ig h t m t ax rm s ve w e
Le l Ba is g a s	M To l l La fo H to o rw ay w r ea vy	Or d in by he Fe de l M in is fo t try an ce ra r	H Tr Ta La ( S V A G ) t ea vy an sp or x w
	Co ia l Ve h ic le ( A B M G ) fro m m er c s m	Tr In io d Te hn lo t, t an sp or no va n a n c o g y	fro 1 9. 1 2. 1 9 9 7 Or d in by he t m an ce
	1 2. 0 4. 2 0 0 2 Or d in by he Fe de l t an ce ra	( B M V I T ) ba d he B S M G A S F I- t t se on	Sw is D f En iro tm t o t, s ep ar en v nm en
	M in is fo Tr Bu i l d in d try t, r an sp or g an	N A G To l l Or d in fro 0 1. 0 9. 2 0 0 3 an ce m	Tr En d t, an sp or er g y an
	H in ( B M V B W ) Re la io t ou s g g n, u		Co ic io ( U V E K ) t m m un a ns
	Fe de l Ro d To l l La 2 0 0 2 ra a w B S M G B G B l. I N 1 0 9 / 2 0 0 2
	( ), t o. fro 6. 1 0 7. 2 0 0 2 m
l l Op To to er a r	l l Co l le G b To H t c m	O A S S E U R P P L K W -M tsy te au s m	io l Cu O f f ic ( O ), Re Z D to g na s m s e
	l l-c l le de to t. w w w o c	G b H t.a t m w w w .g o- m au	l l.a dm in h Be rn w w w .zo .c
Su iso A ho i t ty p er v ry u r	de l O f f ic fo G ds Fe Tr t ra e r oo an sp or	d M Ex to o rw ay an p re ss w ay	io l Cu O f f ic ( O ), Re Z D to g na s m s e
	( B A G ), Co lo ba bu d. de g ne w w w g. n	F in in Co io ( A S F I N A G ), t an c g rp or a n	Be l l.a dm in h rn w w w .zo .c
		V ie f in t nn a w w w .as ag .a
To l le d ds ro a	Fe de l m to ra o rw ay s	M to o rw ay s, ex p re ss w ay s	Co le d k To l l r in te tw te m p ro a ne or ou s
			km
To l le d in km te ro u s	1 2, 0 0 0 km	2, 0 0 0 km	7 1, 0 0 0 km ( 2. 1 % 2 5. 9 % to m o rw ay s,
			in ds 7 2 % he ds ) t m a ro a o r r oa ,
D is ba d l l km ta to nc e se	Fr f l l s j d: ta t o to te te om s r y s m p ro ec ,	Fr 0 1 Ja 2 0 0 4: 0. 1 3– 0. 2 7 3 om nu ar y	Fr 0 1 Ja 2 0 0 1: 0. 1 1– 0. 4 5 om nu ar y
Ca lc la io ba is t	0. 0 9– 0. 1 4 ( ) tu ta no rn ov er x D is l le d, be f a le ta tra	( l. 2 0 % ) tu ta ex c rn ov er x D is l le d, be f a le ta tra	( ) tu ta no rn ov er x D is l le d, is i b le ta tra
u n s	nc e ve nu m r o x s, l lu io la t o n c ss	nc e ve nu m r o x s	nc e ve p er m s h ic le ig h l lu io la t, t ve w e o n c ss
Pr j d l l te to o ec	p 2. 8 b i l l io n	6 0 0 i l l io m n	p 5 0 9. 4 i l l io ( C H F 8 0 0 i l l io ) m n m n
in co m e p er ea r y

	G er m an y	A ia tr us	Sw i la d tz er n
Pa t y m en ho d t m e	C /c d i d, fu l c d, d ire Ba E t c t r, re ar e ar c de b i t	C /c d i d, fu l c d Ba E t c r, re ar e ar	C /c d i d, l c d, d ire Ba E Fu t c t r, re ar e ar c de b i t
hn lo Te c o g y	G S, ire le b i le ( G S ) P M w ss m o	S C du le ( ic in fra d ) D R m o m ro w av e, re	ic hn lo Sp do M te te ro w av e c o g y ee m e r, G P S, D S R C du le ( ic ) m o m ro w av e
A ic lo to t u m a g on	O Bo d U i de d ( j d ts te n- ar n as ne e p ro ec lo 8 0 0, 0 0 0 ) -te to ng rm : u p ap p ro x. In lo te t	A 4 0 0, 0 0 0 O Bo d U i 4 0 0 l l ts to p p ro x. n- ar n , ls ta p or	5 5, 0 0 0 O Bo d U i 2 0 0 l l ts to n- ar n ls ta p or En fo te t
M l lo an ua g- on	rn e g- on 3, 5 0 0 l l s io in ls to ta t te n rm a	N l iv lo i la b le te t o a rn a e g- on av a	Sy s m : rc em en Pa bo hs bo de in t t t y m en o a r r c ro ss g s ( fo fo ig h ic le ) r re n ve s
En fo t rc em en	A 3 0 0 l br i dg 2 7 8 tro p p ro x. co n es fo h ic le t v en rc em en e s	A 1 0 0 l u i ( in d in tro ts te te to p p ro x. co n n g ra he l l p ls ) lu 3 4 b i le t to ta te or p s m o am s	A 1 0– 1 5 l br i dg ( 5 tro p p ro x. co n es ly in io ) t t cu rre n op er a n
O bo d i t n ar un	O Bo d U i ( O B U ) t n- ar n	G Bo o- x	Tr ip C H O B U 1 on -
In l la io ta t s n	D ic ks N in l la io t tru ta t om es c : o s n ire d, l lo io t re q m an ua g- on op n u i la b le av a	D ic ks In l la io t tru ta t om es c : s n ire d /s ic ke ire d fo ks 3. 5 t tru t re q rs re q r c u u =	D ic ks in l la io t tru ta t om es c : s n ire d fo l l ks 3. 5 tru t re q r a c u =
	ig ks in l la io Fo N tru ta t re n c : o s n ire d, l lo io t re q u m an ua g- on op n i la b le av a	ig ks l la io Fo In tru ta t re n c : s n ire d ic ke ire d fo ks 5 /s 3. t tru t re q u rs re q u r c =	ig ks in l la io Fo N tru ta t re n c : o s n ire d, l lo io t re q u m an ua g- on op n i la b le bo de in io t ta t av a a r r c ro ss g s ns
Co t s	O d i fre Bo U t: n- ar n e l la io i d by h ic le In ta t ts s n: co s p a ve ow ne r	O d i 5 Bo U t: n- ar n l la io fre ( ic ke l ie d by In ta t t s n: e s r a p p h ic le /o ) to ve ow ne r p er a r	O d i fre Bo U t: n- ar n e l la io i d by h ic le In ta t ts s n: co s p a ve ow ne r
D is i bu io / tr t n in l la io ta t s n	D is i bu io d in l la io tr t ta t n an s n:	D is i bu io 2 2 0 G O le in tr t ts n: ap p ro x. sa s p o ( lo he dw l l ke bo de t t a a ng ro a ay a y r r , in ) cr os s g s	D is i bu io Re io l Cu tr t to n: g na s m s O f f ic ( O Z D ) e
	In 1 6 0 0 ic try co un : a p p ro x. se rv e tn p ar er s Fo ig 3 5 0 ic tn re n: ap p ro x. se rv e p ar er s	In l la io ic ke l ie d by h ic le ta t t s n: s r a p p ve /o to ow ne r p er a r	Be In l la io 3 7 0 ta t rn s n: ap p ro x. ho ise d t au r g ar ag es

In Switzerland the introduction of a scheme to levy charges on domestic, transit and international road haulage was jointly submitted to a national referendum with a scheme for allocating part of the annual revenue flow to regional governments (cantons) and earmarking 75% of the total revenue to the federal budged to fund the construction of new transalpine rail tunnels.

With the exception of Switzerland, accounting for environmental damage caused by HGV has been considered as a distinct tax from distance-based access charges. Key features of the Swiss system, starting from the fact that it is both operated and supervised by the Regional Customs Office, would confirm that the scheme has the nature of a national taxation system.

On the other hand, the option of earmarking revenue from EUROPPASS to motorway operation and development would point out that ASFINAG should not be considered much different in scope and nature than a motorway concessionaire with the state as the sole owner (like, say, DARS at present in Slovenia or IRI-Autostrade before privatisation in Italy) (Box 2).

Box 2: Austrian EFC.

The federal Highways financing act of 1996 and the Austrian Infrastructure Financing Act of 1997 gave ASFINAG (a state owned company and executive of the Federal Ministry of Transport, Innovation and Technology) the power to collect distance-based tolls on the highly developed sections of the Austria road network and spend the revenue solely on road and rail related projects.

The Austrian Federal road network (approximately 2000 km comprising motorways and expressways) is covered by the new scheme. All vehicles exceeding 3.5 tonnes gross laden weight, not just lorries, are subject to the charge, which went live on January 2004, following full scale end-to-end testing.

All eligible vehicles must be fitted with an on-board unit – a transponder that is attached to the wind screen. The system uses 5.8GHz microwave DSRC. Achieving the greatest possible degree of interoperability with different toll systems in Europe (particularly Austria's neighbouring countries) was considered to be an important aim of the scheme. The vehicles are charged as they pass under gantries that are mounted on each of the 800 charged motorway segments. The scheme operates as a "non stop" electronic charging systems with no barriers. Each gantry is mounted with an array of microwave beacons. There are 100 gantries that are fitted with enforcement equipment (such as cameras and laser scanners).

Drivers have to press a button on the OBU to indicate the number of axes for any particular journey. There are no provisions for non equipped vehicles.

In all three contexts the conversion from vignettes to distance based fees is resulting in a substantial increase of revenue.

In Switzerland it has increased by a factor of five in average the amount that used to be charged to a commercial vehicle on a "typical route" with the previous vignette system.
In Germany, 411 million Euro from Eurovignette in 2000 are estimated to be multiplied by five, reaching 2000 million Euro in 2005 when TOLLCOLLECT will charge HGV as a function of travelled distance on federal motorways.

Only in Holland – the only context in which road pricing was originally conceived to reduce road congestion and not to fund transport investment – the government was committed to implementing the Mobimeter in a framework of fiscal neutrality (i.e. to compensate the introduction of road charges with a proportional reduction in motor fuel or general taxation). A subsequent Dutch government eventually decided to postpone the implementation of the scheme to the completion of substantial investments to improve capacity and standards of both road and other transport networks.

1.2. Switzerland: Three Years After [Sources for this section: Balmer and Hofstetter (2004), Balmer and Rapp (2003), Swiss Customs Authority (2000), Swiss Federal Office for Spatial Development (2002, 2004)]

The Swiss Heavy Vehicle Fee (HVF) was successfully implemented on January 1st 2001 and replaced a former flat charge which was very similar to the Eurovignette. It is a performance related fee, depending on the distance driven, the admissible weight and the emission standard of the vehicle. The revenues of the fee are mainly used to finance new railway infrastructure. Among the most important projects is the construction of two new rail links across the Alps, the base tunnels through the Saint Gotthard and the Lotschberg. Simultaneously with the introduction of the ¨ HVF the weight limit is increased from 28 tons to 34 tons (and in 2005 to 40 tons). To raise the fee, two systems have been developed: one for users equipped with an On Board Unit (OBU) and one for users without OBU.

Three years after introduction the new HVF is a success. The technical solution worked satisfactorily from the very first day it was in operation and traffic volume of lorries declined. Furthermore the new fee offered strong incentives towards cleaner vehicles and helped fighting air pollution. The financing of the new railway infrastructures is secured to a substantial part by revenues of the new fee. Even at the European level, this pioneering example of cross-financing was recognised. In the White Paper on European Transport Policy the Commission of the EU mentioned it as a possible future way of financing new railway infrastructure.

Operational Experience: The system has worked very well since the very beginning. The main problem was the quality of the vehicle registry data delivered by the 26 decentralised cantonal authorities. This problem could be resolved with a special effort for the harmonisation of this data. The HVF procedures at the border could be added to the existing ones in a very smooth way. The existing customs organisation knew the specific needs of the transport industry very well and was accepted and respected as an authority by this same industry. No system fault in the technical or procedural level occurred. The concept with a secure OBU passed the live test. The fraud in the field is very low, despite the fact that up until now only 12 stationary enforcement stations are in operation at strategic points on the motorways. The whole system including data collection and the assessment practice of the declared data has been approved by court by several appeals. The domestic debtors cause the most administration expense. Different means are foreseen by law, e.g. deprivation of the registration, securing the vehicle, etc. But the prosecution means a huge administrative effort. Foreign debtors don't exist because their vehicles are stopped by customs at the border when they leave Switzerland. Exception handling is the second costly administration part (Table 3).

For a correct assessment of the impacts, the following facts have to be considered.

The HVF is applied on the whole road network and therefore deploys its effects on the whole territory of Switzerland.
The average level of the fee is comparatively high. Per kilometre it is three time as high as the one in the German toll collect project and about 50% higher than the one in the Austrian toll scheme (both tolls levied on motorways only).
The increase of transport costs due to the HVF of (on an average) 19% were more or less compensated by productivity gains due to the higher weight limit of (on an average) 18%.

Table 3. Statistical Figures of the Operation of the Swiss HVF for 2002.

Domestic vehicles equipped with OBU	52500
Foreign vehicles equipped with OBU	1650
Entries Switzerland with Beacon Transactions (OBU)	1122000
Entries Switzerland with Terminal Transactions	1789000
Revenues Total Domestic vehicles Foreign vehicles	500 Mio EUR 78% 22%

Since all effects have to be seen as a result of the interaction between the HVF and the higher weight limit, it is more precise to speak of effects of the new regime instead of effects of the fee alone.

Adjustments of the Fleet Composition: Already prior to the introduction of the fee sales numbers of new lorries increased by 45% compared to previous years due to fleet adjustments. In order to save fees, the haulage business bought cleaner vehicles belonging to the cheapest emission category and adjusted the size of their vehicles to the actual needs of the market. Previously, over-sized vehicles were very common in the business.

Consolidation in the Road Haulage Business: The new regime led to a consolidation in the haulage business, either through mergers or by pushing away smaller competitors. Due to better capacities, larger companies are able to operate their lorries in a more efficient way. For example, this allows reducing unproductive and (due to the fee) expensive empty trips, which offers them an advantage compared to smaller competitors.

Traffic Volume: With regard to traffic volume on Swiss roads the new regime with the Heavy Vehicle Fee and the higher weight limits for lorries has led to a significant break of former growth trends (Fig. 1).

On the national level, annual increases of about 7% prior to the new regime were replaced by declines of around 4% in 2001 and 3% in 2002. In 2003 the traffic volume of lorries remained stable. This decrease recognized so far cannot be attributed to the slow down of economy, because it was also recorded in the first half of 2001, when economic growth was still about the same as in previous years. Interesting enough, the doubling of the flat fee in the year 2000 had shown

Fig. 1. Changes in Road Performance of Heavy Goods Transport in Switzerland 1997–2003.

no effect on growth in transport performance. The break in trend only happened after the change from a flat to a distance related fee.

With regard to the traffic volume through the Alps the new regime has first resulted in a stabilisation of the traffic volume after years of strong increases of up to 10% annually. In 2002, the number of lorries to cross the Swiss Alps was even reduced by 9%. This reduction might mainly be due to the restrictive traffic regulation system introduced after the tragic accident in the Gotthard tunnel on the 24th October 2001. In 2003, the number of lorries in transalpine transport increased slightly, but was still lower than in 2000, the last year under the old regime.

Transfer of Freight from Road to Rail: The transfer of freight from road to rail did not happen in large extent so far. However this is not surprising and was never expected within such a short timeframe for the following reasons.

The share of freight transport by rail in Switzerland is already very high. In transalpine traffic, it amounts to two thirds of the total.
The competitive advantage which the rail gained due to the fact that the new fee changed prices in favour of the rail were outbalanced by the productivity gain of the road sector as a result of the higher weight limit.

A transfer of freight from road to rail is therefore not possible without strong additional efforts in the rail sector. Two factors are crucial: The Railway sector must realise substantial gains in productivity and has to improve its reliability. A specific problem to solve are the delays in transnational transport. The framework for the necessary changes is being established by the Swiss rail reform.

1.2.1. Plans for the Future

In 2005, Switzerland will raise in parallel the level of the rate from now 1 Eurocent per ton.kilometre to 1.6 Eurocent and the weight limit from 34 to 40 tons. Another increase is planned for 2007, when the first new rail link across the Alps, the Lotschberg, will be opened. The fee will then reach the level of 1.8 Eurocent ¨ per ton.kilometre, as it was agreed with the European Union in the treaty on land transport. The expected revenues will increase to almost 1 billion EUR per year. There are, at the moment, no plans to extend road tolling to passenger transport.

1.2.2. Success Factors of the Swiss HVF

Clear political will: According to forecasts, the performance of heavy vehicles on Swiss roads (in kilometers) was expected to increase by 60% within 15 years (between 2000 and 2015). In order to avoid this scenario the expected growth had to be limited, especially in the case of transalpine transport, where Swiss citizens even voted in a national referendum for a marked reduction of the number of heavy vehicles on transalpine corridors. The Swiss HVF is one of the major instruments to achieve this goal. It is supported by a majority of citizens and based on clear and transparent political aims.

Embedding of concerned parties: Opposition from concerned parties against every kind of new charge, tax or fee is obvious. In the case of the Swiss HVF acceptance could be increased by:

Neutral treatment and consistent prosecution of fraud.
The detailed elaboration of the fee collecting concept and the corresponding directive was escorted by all concerned parties.3
Compromises were conceded with regard to the handling of exceptions although this was complex and costly (e.g. for wood transports and for military, agricultural and public transport vehicles).

Technical solution: The technical solution is driven by the specific requirements of a distance related fee on all roads. The OBU records all relevant data, so for the hauliers they are fully retraceable. This secures transparency and trust in the system. In order to prevent or minimise complaints the tachograph is used for distance counting instead of highly sophisticated solutions. This is not very accurate but has been established and accepted as a European legal measurement system for years. At the border stations, distance counting is switched on or off via DSRC beacons. The neutrality of treatment is ensured by a secure OBU for equipped vehicles and very well controlled border procedures for vehicles without OBU.

2. TOLL COLLECT IN GERMANY

In August 2001 the German Government decided to introduce distance related road pricing on the German Federal motorway network for heavy goods vehicles (HGVs) with a weight exceeding 12t GVW (BMVBW, 2001). The overall road pricing strategy was aimed to achieve four goals.

Introduce the user pays principle for HGV road use.
Create better competition between transport modes.
Finance the infrastructure costs of all transport modes.
Make Germany the international expert for innovative road pricing technology (BMVBW, 2004).

In April 2002 the legal changes required to introduce road pricing came into effect (BGB, 2002). In April 2002 the tender for the German IRPS was published. The IRPS was required to function fully automatically causing no stop in the flow of traffic and needing no toll plaza areas. Four consortia (Fela, AGES, ETC and TSR) were the main contenders for the system. TSR, a small German company, could not meet the deadline for the submission and was subsequently excluded from the tendering procedure. The tender of the Swiss consortium Fela Management AG (who successfully runs and manages IRPS for HGVs in Switzerland) was rejected because they could not provide the required bank guarantee for the payment of penalties if the introduction of the system was delayed.

The two systems remaining within the tendering process were from AGES (consisting of Mannesmann (now Vodafone), Siemens/VDO, Aral, Euro-shell, DKV services and UTA) and Toll Collect (then called ETC, comprised of Deutsche Telekom, DaimlerChrysler and Cofiroute). On the 10th of June the Ministry of Transport, Construction and Housing announced that Toll Collect was the preferred tender on the basis that the price of their system was lower than AGES. On the 27th of June, Toll Collect's bid was confirmed as being selected and the company agreed to have the system running on a trial basis in 11 months and completely functional in 15 months (c't, 2003). AGES did not consider that the tendering process was handled fairly and began litigation against the German Government for favouring Toll Collect. However, the Federal Cartel Office ruled that contract was handled in a correct and non-discriminatory way.

AGES turned to the responsible Higher Regional Court in Dusseldorf but the ¨ case did not come to trial as the charges were dropped on the 4th of September 2002. Although the litigation process effectively delayed the start of system construction by several months, the Federal Government and Toll Collect still considered the 31st of September 2003 to be realistic as the date for the complete system start and Germany resigned their participation in the Eurovignette effective on this day (see for example manager-magazin (2004) for a more complete chronology of the introduction of IRPS in Germany).

Many critique points can be made in retrospect, for example, the time frame for submitting tenders seems exceptionally short for such a complex system. The basis of the bid selection was not transparent which led to litigation. The 15 month deadline for the complete system start was critically short and it should have been apparent, especially after the additional delay caused by litigation, that it was impossible to keep to the original schedule.

Technical features. The Toll Collect technical system uses satellite positioning (GPS system) and mobile communication to charge HGVs for the use of Federal motorways. Of central importance to the system is a vehicle on-board unit (OBU) consisting of a GPS receiver, a digital map and a mobile phone providing contact via short messages (SMS) with the Toll Collect charging centre. A DSRC transponder used to check vehicle compliance is also integrated in the OBU. The OBU must be able to identify when the vehicle is travelling on a section of road where road charges are to be paid and transmit pricing relevant data to the payment centre. The system allows charging to be carried out automatically without stopping the flow of traffic. This is particularly relevant for the existing German motorway system as there is no area available to install toll plazas at motorway entry and exit points.

The GPS based system is flexible and can be extended to include other hierarchical road levels or new motorway segments with minimal technical effort. The charging regime is also flexible and could be varied in the future to include payment principles other than distance, for example, time of day, road congestion or geographical area (Toll Collect, 2004). Originally, the OBU was planned to use UMTS (TCP/IP Protocol) to provide "real time" communication between the vehicle and the Toll Collect central. This "real time" system would have allowed a whole spectrum of added services including, but not restricted to, logistical applications. Technical problems combined with the difficulties surrounding the introduction of UMTS in Germany have prevented this so far (Borchers, 2003).

Problems and Delays. Because of problems with the technical system, IRPS could not be started as planned on the 31st of August 2003 and since then the starting date has been moved back several times (see for example Handelsblatt, 2003). The problems arising from this situation are serious and manifold. Since October 2003 HGV infrastructure usage charges could not be raised in Germany. The revenue loss has been quantified as 163 million monthly in 2003 and 233 million monthly in 2004 (Cloes, 2004). The loss of faith in the technical system is so overwhelming, it has changed the meaning of the word "toll" (Maut in German) to become synonymous with anything that does not or will not function. At the European level, the original Proposal for a Directive on the interoperability of electronic road toll systems (EU, 2003) which recommended the TEN-wide (and effectively Europe-wide) kilometre charge be calculated using satellite positioning and mobile communication (GNSS/CN system) for all new IRPS from 2008 and for all charging solutions by 2012, has been amended. Systems based on DSRC technology will now be an acceptable alternative to GNSS/CN systems. As the German system is the first GNSS/CN based system to be used for IRPS, this illustrates the level of mistrust in the system technology at the international level.

Although difficulties with the complete Toll Collect system have been reported, the major problems to date have been with the OBU. Toll Collect used the "best of breed" strategy (use of the best available system parts) for the development of the OBU software components. Lack of communication and cooperation between the teams developing their "best" system components allowed disastrous mistakes, such as the development of different software packages without a common interface, to occur (Filser, 2004). This preventable oversight was a central reason for the postponement of the introduction of the system. Other problems that arose would have been obvious to anyone with experience in the transport branch; for example OBUs that are directly attached to the windscreen of the vehicle must be permitted for use by the German Federal Bureau for Motor Transport, the public authority responsible for all matters relating to the approval and registration of motor vehicles. This formality was neglected and the application for the use of the OBU was submitted so late that the approval could only be granted in October 2003, over a month after the planned system start (c't, 2003).

A similarly preventable problem was the failure to programme the OBUs with the European standard DRSC transaction protocol (CARDME), even though a DRSC link for compliance checks is provided as a basic function of the OBUs. Although this deficit was identified as early as September 2003 in the popular media (Borchers, 2003), there has been no indication from Toll Collect that the problem has been rectified. This effectively means that the Toll Collect OBU is not interoperable with any other road pricing technical system presently in use in Europe. Also, major petrol cards in use in Germany cannot be used to pay monthly invoices from Toll Collect, due to the failure to programme this option into the software system. Problems with other system components also could have been prevented with adequate knowledge of the transport industry and legal requirements for HGV operators: to give one example only, the manual booking/payment terminals did not allow for HGV maximum speed limits or for time at the wheel rest periods required by law and issued motorway use permits with unrealistically short driving times for HGV trips (c't, 2003).

The relationship between the German Government and Toll Collect became very strained at the end of 2003. The major points of disagreement were: providing a guaranteed date for the start of the technical IRPS and payment of compensation for lost revenues. To what extent the compensation for revenue loss caused by system failure or delay had been addressed in the original contract between Toll Collect and the Government is not clear, as the contract details have not yet been made public. On the 17th of February 2004, the contract with Toll Collect was terminated by the Ministry of Transport (see for example Spiegel, 2004). Ten days later, the German Government and Toll Collect had reached a compromise solution. Toll Collect agreed to provide IRPS services in two stages. By the 1st of January 2005 the system must function, albeit with reduced capabilities (called OBU1) or a penalty payment of 40 million per month, increasing by 5 million each month until a maximum monthly compensation of 80 million is reached.

By the 1st of January 2006 the entire system (OBU2) must be completely functional, from this point in time Toll Collect is responsible for paying compensation equal to expected revenues from road pricing if the system fails. The Government has the right to cancel the contract with Toll Collect if either the launch of OBU1 is not successful or the technical development of OBU2 is not completed by June 2005 (if this does happen, the Government intends to reintroduce the Eurovignette and restart the tendering process for the IRPS). Toll Collect will receive a 5% reduced payment in 2005 during the timeframe the OBU1 is in use (yearly payment to Toll Collect by a fully functioning system is estimated to be 20% of the total revenues – 2.8 billion annually – a 5% reduction equals approximately 30 million). A court of arbitration is to review the claims of the Government concerning reimbursement for lost revenues due to the delayed start of the Toll Collect system (BMVBW, 2004; Brychcy, 2004).

Some aspects of introduction of IRPS to Germany have also been seen as problematic by the European Union. The European Commission was concerned that the set kilometre charges were not calculated in accordance with the guidelines for road pricing as set out in the Directive on the charging of HGVs for the use of certain infrastructures (EU, 1999) and requested that they be provided with these calculations before the start of road pricing. In addition to this, the German Government neglected to inform the European Union about the planned rebate on fuel tax promised to the German transport branch as compensation for the increased transport costs arising from the introduction of road pricing. This rebate system could have been considered to discriminate against foreign vehicles, as they would in effect pay a higher per kilometre charge than German HGV operators – as the rebate was only applicable for fuel purchased in Germany. The concern of the European Commission was so great, they wished to hold off the start of road pricing until a solution with the German Government could be found (hib, 2003).

On the 30th of April 2003, the EC DG Competition formally gave clearance to Toll Collect as a joint venture between Deutsche Telekom and DaimlerChrylser on the condition that Toll Collect would not have an automatic monopoly on transport data that could be useful for telematic applications resulting from the IRPS (EC, 2003). It had originally been planned that this data would be used exclusively by the DaimlerChrysler subsidiary company DaimlerChrysler Mobility Management. This data will now be made available to other companies providing telematic services, although it is not clear how this will be done (price setting etc). A postponement of the system start, because of these and other concerns from the EU, became unnecessary as the system was delayed because of the technical reasons described above.

At the time of writing (spring, 2004) the only conclusion that can be made about road pricing in Germany is that none of the objectives of the Government have been met so far: HGVs presently pay no charges for the use the German motorway system which has lead to an even greater distortion of competition between the transport modes. The financing of transport infrastructure construction and maintenance is unsure, which has created political tension at all levels of Government and across all branches of industry, and there has been an incredible loss of faith in German technology in general and specifically in satellite based road pricing technology.

3. TOLLED MOTORWAYS

Distance-based tolls are levied on some 20,000 km of motorway networks manly located in Southern Europe. Countries who historically relied on toll collection to fund motorway development include France, Italy, Portugal and Spain. More recently, also Greece, Croatia and Slovenia have chosen to levy tolls to fund the development of their national motorway networks. In closed, tolled motorway networks each vehicle is identified (either visually or electronically) as it passes through a toll plaza and then charged (either manually or electronically) as a function of vehicle attributes and distance between entry and exit plazas.

On tolled networks as a whole, the average proportion of revenue generated by car and trucks is 80% and 20% respectively. In some combination with the diffusion of pre-paid cards and credit cards, a variety of Electronic Fee Collection (EFC) systems has started being introduced in the last decade to charge vehicles without requiring them to stop. All EFC in use for toll collection are based on Dedicated Short Range Communication systems (DSRC) with quite simple on board units and roadside equipment located at toll plazas. With an agreement on European standards reached only in Summer 2001, the first generation of EFC systems has developed under specification of concerned motorway concessionaires. In countries where different motorways links are operated by several concessionaires, Inter Company Remote Payment procedures have also progressively developed at national basis – TELEPASS in Italy, T.I.S in France and Via-VERDE in Portugal – to enable motorway users to use the same payment means on sections operated by different concessionaires and to allocate revenue among the latter as a function of mileage performed in each section.

3.1. Partnership with European Institutions

At European level, positions and interests of tolled motorway concessionaires are represented by ASECAP. As exemplified in Box 3, being a privileged partner

Box 3: ASECAP' Vision and Mission.

ASECAP is the only European professional association of tolled motorways companies. It gathers and represents 114 organisations that manage more than 20,000 kilometres of toll roads through 14 European countries.

ASECAP's goals consist in:

Defending the toll system and proving to Europe that it is the most efficient way to finance the construction and the operation of motorways and big road infrastructures.
Promoting the image of the toll road sector.
Participating to research and experiment programmes to strengthen the efficiency of the networks and to improve the users' comfort and security.
Facilitating and promoting the exchange of information between the ASECAP members: any kind of information concerning the construction and operation of toll motorways, bridges and tunnels.
Promoting and organising studies meeting at technical, administrative and financial levels.
Collecting technical data and statistics about the European toll networks and providing them to its members and to all interested bodies.

An association definitely focused on Europe, playing a leading role in the development of a harmonised Trans-European network, ASECAP wishes to be the privileged partner of the European institutions (the European Commission, the European Parliament) in order to take an active part in the debate on the future of the road transport as well as in the development of a harmonised Trans-European road network. This contribution to the construction of a common transport policy results for instance in the ASECAP members' participation in the CESARE project, co-financed by the European Commission and whose aim is to develop a common electronic charging system to all members that could be used by other countries of the European Union.

As new charging systems emerge (especially in Germany, the Netherlands and Great Britain), the ASECAP members have to pursue their action and studies on that matter in order to find a way to achieve the interoperability of systems based on different technologies. In this prospect, and to the European Commission's request, the ASECAP members would like to carry out a complementary research in the framework of the research programme initiated par the European Commission that will continue the CESARE and PISTA projects.

ASECAP is an open association and it allows any new member to take part in the studies and to benefit from the experiences of the countries that have been developing a "toll culture" for many years.

These are, in short, the main aspects that will lead to a harmonised Trans-European road network in which ASECAP is highly involved.

Source: Statement of ASECAP President – Remy Chardon – in the opening page of ´ www.ASECAp.com, accessed in July 2004.

in the definition of European policies to charge for the use of roads is a key commitment of ASECAP.

3.2. Toward a New Generation of Tolled Motorway Concessionaires?

In the absence of recent developments in the diffusion of EFC, main European tolled motorway concessionaires have started quite aggressive commercial actions (including discount and promotion of credit cards) to increase the relatively modest share of equipped vehicles (some 20% of networks where EFC is available, in year 2000, basically regular motorway users).

By far most important changes ongoing in "ASECAP contexts" are organisational and industrial in nature.

To reduce the main shortfalls from motorway concessions – over-investment and risk of capture of the state by the concessionaire – in many such contexts new rules have started being introduced/considered to introduce safety/quality/performance targets in contracts with established concessionaires and to make the awarding of new concessions more competitive.

All in all, moves from old approaches to motorway concessions are conceived within Community' legislation for Public Procurement/Works and to favour Public Private Partnership (PPP) in motorway development within project financing schemes.4

These new, more competitive approaches to involve private capital and know how for the construction of new motorway links have proved complex to implement in contexts at early stages in motorway development (including in Hungary where B.O.T schemes for motorway construction have been pioneered in accession countries).

On the other hand, in contexts where motorway networks are well developed (mature) new approaches to PPP have started being explored for construction of new motorway links in parallel with a wave in concession renewals and privatisation of established motorway concessionaires. With short-term objectives of national Treasuries set to make cash, the privatisation was favoured by means of extending concession duration, relaxing rules for capping prices and jointly with bilateral negotiations to engage further investment plans (to expand capacity of/further integrate the existing tolled motorway network). Early evidence on performance of recently privatised motorway operators would point out that motorways are among the best performing companies in European stock markets.

More in general, the performance of main European tolled motorway concessionaires would point out the following set of general patterns/specific issues.

Tolling motorway links when traffic is sustained and past investment has been amortised can generate substantial cash flows and surplus revenue.5
As the need for further expansion of national motorway networks is reduced, and expansion of capacity where it is mostly needed (basically around most densely populated conurbation's) is constrained by physical space/environmental concerns, motorway concessionaires are seeking new investment opportunities in other areas/sectors.

Other than in the opening up of international markets for contracts for building/operating roads main concessionaires have entered the market for IRPS development: Autostrade Spa (recently renamed Autostrade per l'Italia) and Cofiroute are key partners of the consortia awarded to develop and operate EUROPASS and TOLL COLLECT in Austria and Germany respectively, and BRISA is the main shareholder of ViaVerde.

Even in the absence of a technology breakthrough, Swiss and Austrian schemes relying on Dedicated Short Range Communications to levy electronic charges on HGV have reduced the gap between increasingly sophisticated pricing strategies and what technology has to offer in terms of both reduced dependence on toll stations and enhanced computing and transmission functions of On-Board-Units.

Jointly with delays and problems in adoption of a GPS based system in Germany, opportunities of "non stop pricing" are also recognised by the European Commission in proposing changes to Directive 41/99 and a new Directive on EFC interoperability favouring the transition towards new approaches in the establishment of open transmission platforms in Europe (and in perspective of truly autonomous ones).

There is also substantial evidence that established motorway concessionaires have started to integrate/optimise local road networks of different standard/capacity and have the know-how for servicing local authorities for both planning of new roads and upgrading/maintenance of existing ones.

Changes in industrial/commercial strategies taking place at motorway concessionaires would also point out reasons for re-thinking conventional tolled motorway schemes in favour of a new generation of "area-wide concessions." In combination with functional circumstances concerning the fact that traffic problems are local in nature (including in terms of strong concentration of short distance commuting and freight distribution traffic) this would allow for optimising investment/upgrading programmes on increasingly integrated regional road networks, with special reference to portions of interurban networks with lower capacity and construction standards, that are second only to urban areas for concentration of accidents and fatalities. Regional concession schemes seem also a pre-condition for introducing regulatory charges to combat time and place related congestion in most densely populated regions.

Institutional processes to devolve powers from central to regional governments can be expected to favour this evolution, including in terms of establishing a Traffic and Demand Manager at regional level with different objective functions, tasks, responsibility than those of the Road Infrastructure Manager(s).

4. THE BRITISH LORRY ROAD USER CHARGE

Lessons learned in the first applications of EFC for HGV have formed the basis for the reform under consideration by the Government of the United Kingdom to modernise the taxation of the road haulage industry. Information on this initiative is naturally much scanter than on those reported above, which have been initiated typically about one decade ago. This shortage of information restricts the depth of the analysis and discussion possible here.

As stated in Progress Report 3 jointly presented in the occasion of the 2004 Budget "It is important that the U.K. learns from the experience elsewhere in Europe. For example, the programme team is making use of industry experts including those with experience of introducing the Swiss Lorry Road User Charge, and advising on the Austrian scheme, to help ensure development for an effective solution for the U.K. Experience in other countries points to the need to build sufficient time into procurement (specifications tendering and negotiation) development, building, promoting, testing and phasing and also the role of a strong public sector authority which will manage a variety of contracts rather than rely on a single supplier with multiple contracts."

As presented by the Government, the driver for introducing a British Lorry Road User Charge (LURC) is not the generation of additional revenue. It is anticipated that the charge will be offset by a repayment of part of the tax on diesel fuel purchased in the U.K., among the highest in the EU, which currently puts the domestic haulier at a disadvantage. The combination of a charge paid by all goods vehicles using U.K. roads and the repayment of part of the tax paid on diesel fuel purchased in the U.K. is expected to put all operators on an equal basis. The charge will be due for vehicles exceeding 3.5 tons. In order to reflect the costs and the environmental performance imposed by the vehicle the charge is supposed to vary according to distance travelled, vehicle type, road type (Table 4).

The technical solution for the levying of the fee is not defined yet. The application of different taxation rates for the use of different roads needs a specific and innovative approach.

Table 4. Indicative Timetables in Adoption of LURC.

Introduction	Procurement
Spring 2004 – Initial Legislation	Spring 2004 – Issue a Prior Information Notice
Spring 2005 – Legislation on structure,	Spring 2004 Publish Official Journal of EU.
collection and administration of charges	Advertisement and Pre Qualification
and key definition	Questionnaire
2006 – Design and build phase e.g. installation of road side equipment	Summer 2004 – Potential Supplier Open day
2006 – Secondary legislation including	Summer 2004 – Preliminary invitation to
regulation	negotiate
2006 – Recruitment and training of staff	Contracts awarded by end 2005
2006–2007 Pilots and testing
end 2006 – Go live for pre-registration services
from 2007–2008 – Equipment installed in
vehicles and start of revenue collection
and fuel duty repayment

Source: Modernising the taxation of the road haulage industry – Progress Report 3, HM Treasury, HM Custom and Excise, Department of Transport, March 2004. Downloaded from http://www.hmtreasury.gov.uk/media//CBC02/lorry road user charge 290404.pdf accessed in July.

NOTES

1. In addition, member states (including those where user charges are levied) can levy tolls for the use of bridges, tunnels and mountain passes.
2. Differing from other public utilities such as energy or telecommunications national governments, either Ministry of Transport or economic Departments, and not independent regulators are responsible for determining the level and structure of motorway tolls as well as other rules for price caps, reinvestment of revenue and concession award/renewal. But motorway concessions were generally created before privatisation of those sectors.
3. In the Swiss case, actors involved in the development of the HVF were e.g. the transport and haulier associations, the Federal ministries of road and transport, the truck and tachograph industry, garages and tachograph workshops, truck drivers association, Swiss cantonal vehicle registration authorities, Swiss cantonal police organisation, parliament commissions as well as insurance associations.
4. Differing from motorways concessions, private investment in a project financing scheme is expected to be remunerated by revenue generated by the operation of the individual, new infrastructure link in the absence of mechanisms allowing for crosssubsidisation among different links operated by the concessionaire. To work – and to compensate higher costs of tapping capital for international capital markets with efficiency gains – the mechanism requires a clear scheme to allocate political risks between the state and the concessionaire and to transfer to the latter all the risks associated to construction and exploitation of the infrastructure.

5. Figures for pre-tax profit/turnover ratios of main European concessionaires would point out that some 40–60% of annual revenue from tolls is spent on maintenance, upgrading and operation of the infrastructure.

REFERENCES

200072030(INI). Report on Transport Infrastructure Charging. European Parliament – Committee on Regional Policy Transport and Tourism – Rapporteur Paolo Costa (FINAL A5-0345/2000).

http://www.ASECAP.com.

http://www.hm-treasury.gov.uk/media//CBC02/lorry road user chargef.

Modernising the taxation of the road haulage industry – Progress report 3, HM Treasury, HM Custom and Excise, Department of Transport, March 2004.

The German system

BGB (2002). Gesetz ¨uber die Erhebung von streckenbezogenen Geb ¨uhren f ¨ur die Benutzung von Bundesautobahnen mit schweren Nutzfahrzeugen (ABMG), 5.04.2002, see: http:// bundesrecht.juris.de/bundesrecht/abmg/index.html.
BMVBW (2001). Federal Ministry of Transportation, Construction and Housing. LKW-Maut auf Bundesautobahnen beschlossen. Press release number 208/01, Berlin 15.08.2001.
BMVBW (2004). Federal Ministry of Transportation, Construction and Housing.Fakten zur LKW-Maut http://www.bmvbw.de/LKW-Maut-.720.htm Accessed 05.05.2004.
Borchers, D. (2003). Und nun noch den Gateway dranpfriemeln. In: FAZ (27.09.2003).
Brychcy (2004). Das Versprechen aus den Chefetagen. Suddeutsche Zeitung, 51/02.03.2004. ¨
Cloes, V. A. (2004). Einf ¨uhrung der streckenbezogenen LKW-Maut in Deutschland.
c't (2003). Verursacherbedingt versp ¨atet. Das fortschrittlichste Maut-System der Welt und die Realit ¨at. c't Zeitschrift 22/2003.
EC (2003). European Commission. Conditional clearance for Toll Collect joint venture between Daimler/Chrysler and Deutsche Telekom. Press Room release DN: IP/03/594, Brussels, 30.04.2003.
EU (1999). European Union. Directive on the charging of HGVs for the use of certain infrastructures. Directive 1999/62/EC.
EU (2003). European Union. Proposal for a Directive on the interoperability of electronic road toll systems. COM(2003) 132 final, Brussels, 23.4.2003.
Filser, H. (2004). Viele Einzelteile, kein System. Suddeutsche Zeitung, 51/02.03.2004. ¨
Handelsblatt (2003). G ¨uterverkehrsamt: Mautstart verz ¨ogert sich. Handelsblatt, 22.09.2003.
hib (2003). heute im bundestag (hib) Opposition mahnt Kompensationsl ¨osung bei Einf ¨uhrung der Lkw-Maut ein. Nr. 054, 12.03.2003.
manager magazin (2004). Maut: Sie kommt bestimmt, bestimmt ... www.manager-magazin.de, 17.02.2004. Accessed 03.05.2004.
Spiegel (2004). K ¨undigung der Mautvertr ¨age. Spiegel online, 17.02.2004 http://www.spiegel.de/ wirtschaft/0,1518,286718,00.html Accessed 03.05.2004.
Toll Collect (2004). Truck toll system. www.tollcollect.de Accessed 04.05.2004.
Wissenschaftliche Dienste des Deutschen Bundestags Reg. Nr.:WFVG–203/03, Berlin, 7.01.2004.

The Swiss system

Balmer, U., & Hofstetter, B. (2004). The Swiss heavy vehicle fee technical solution, operational experience and impact on transport, Article presented at ITS conference in Budapest.
Balmer, U., & Rapp, M. (2003). The Swiss distance related heavy vehicle fee (LSVA) a novel approach to area-wide road charging. Basel and Berne.
Swiss Customs Authority (2000). HVF in concrete terms: The heavy vehicle fee/HVF: Background Fees – Recommendations – Alternatives. Berne, 2000.
Swiss Federal Office for Spatial Development (2002). Fair and efficient the distance-related heavy vehicle fee in Switzerland. Berne, June 2002.
Swiss Federal Office for Spatial Development: http://www.are.admin.ch/are/en/verkehr/lsva/ index.html.

8. HARMONISATION AND INTEROPERABILITY OF NATIONAL TOLLING SCHEMES

Jesper Mertner and Michael Knørr Skov

The "one and only" approach to introduce inter-urban road pricing for heavy goods vehicles in Europe does not exist. But to ensure easy and free movement of goods as well as fair competition among hauliers, some level of interoperability between different systems, operators within member states as well as between member states is desirable. Important stakeholders in this context are road users, operators and manufacturers, national and international government agencies as well as the society as a whole.

Achieving full interoperability of inter-urban road-pricing (IRP) within Europe requires substantial investments in planning and implementation on the following topics:

Technical standardisation.
Procedural and operational harmonisation.
Institutional and contractual approximation.

The European Commission seems also to be moving towards not necessarily one unique IRPS system for Europe but a system of interoperable IRPS systems, both technically, procedurally and contractually as described in the Directive on Interoperability of Electronic Road Toll Systems in the Community (Directive 2004/52/EC) from April 2004 (Council of the European Union, 2004). The principle for the Directive was suggested in the new White Paper "European transport policy for 2010: Time to decide" (European Commission, 2001), stating that users should "have a quick and easy way of paying infrastructure charges, using the same means of payment throughout the network without losing any time at toll stations."

Therefore, aspects of interoperability and long-term convergence of different solutions are discussed in this chapter including whether interoperability is always desirable. The chapter describes interoperability from institutional, contractual, procedural, operational and technical level as well as from a European perspective.

1. WHAT IS INTEROPERABILITY OF TOLLED SCHEMES?

Different EFC systems are fully interoperable if the user of one EFC system can access a tolled road of another EFC system without being treated as a non-equipped user (no manual payment, pre-registration, etc.) and if he is charged for the trip by his domestic operator (preferably) with no aggravation on the price to be paid in comparison with the users registered with that EFC system. In general, systems are fully interoperable when harmonisation in three aspects is achieved: (a) technical; (b) procedural; and (c) contractual (TIS.pt, COWI, ECOPLAN, UNIKARL, ISIS, 2003).

1.1. Importance and Cost of Interoperability

It is important that EFC systems are interoperable across national borders in order to avoid creating new obstacles to traffic flows in Europe, in accordance with the Single Market provisions of the Treaty. Analyses show that many stakeholders benefit from interoperability: road users are interested in easy passage on all roads throughout Europe without artificial barriers. The operators of roads are interested in an increasing share of "equipped" users, as equipped users have lower transaction costs than non-equipped ones. Finally, manufacturers are economically interested in mass production of OBU's and other technical equipment, which makes their markets bigger (TIS.pt, COWI, ECOPLAN, UNIKARL, ISIS, 2003).

On the other hand, road users who rarely drive outside their own country could be less willing to pay extra for their OBU just to allow easier payment of tolls on the occasional trip abroad. Existing operators will also have additional costs to ensure interoperability with systems that are already in operation (for fixed installations and for on-board equipments) compared to the – in many cases rather low – percentage of road users who are requiring interoperability unless each country has several operators, each previously with his own EFC scheme (TIS.pt, COWI, ECOPLAN, UNIKARL, ISIS, 2003).

1.2. Difficulty to Reach Interoperability

Many countries have already invested heavily in systems for passage and/or network tolling while others are planning to introduce area tolling. While new systems may be harmonised according to some agreed international standards it is for existing systems that the convergence towards new interoperable systems presents a significant and expensive challenge.

The existing IRPS have been introduced by different countries and operators, at different times (and levels of technological evolution) and for different political purposes. Compatibility with other tolling systems was in most cases a less important feature. Significant investments have already been made by the operators to set up their EFC systems and by the drivers who use them. Interoperability with other operators would very often require new investments in the first place. However, even for newly designed IRPS, there is no "one and only" technical approach. Besides technical questions, interoperability also concerns issues like procedural and contractual harmonisation of schemes.

1.3. DSRC-Systems and Interoperability

Countries and operators that introduce a DSRC (roadside based) solution can in principle ensure interoperability with other national schemes. Technical interoperability with other DSRC solutions can be reached by sticking to international standards with respect to specification, e.g. EN ISO 14906 (EFC AID, 2003), EN 13372 (EN Profiles, 2003), and EN 12253 (EN L1, 2003) from 2003 generally based on the work of Technical Committee 278 of the European Committee for Standardisation CEN TC278 for DSRC 5.8 GHz. Procedural interoperability – one important point here being the classification of vehicles – is also theoretically solved. The procedures have been defined by the CARDME/CESARE projects (CARDME-4, 2002; CESARE, 2001) and are ready for implementation. For instance, the new Austrian scheme is based on the CARDME specifications and will therefore be interoperable with other CARDME-applications. The aspect of contractual interoperability is being treated at the moment within the EU-co-funded project PISTA (Pilot on Interoperable Systems for Tolling Applications). This project (PISTA, 2004) intends to actually implement an interoperable Electronic Fee Collection (IO-EFC) throughout a corridor belonging to several European motorway networks. But basically, it is within the responsibility and scope of each operator to ensure contractual interoperability through negotiations and contracts with other operators (TIS.pt, COWI, ECOPLAN, UNIKARL, ISIS, 2003). The realistic aim is to ensure full interoperability within the DSRC cluster within a few years, with relatively minor investment needs for drivers and operators.

Drivers equipped with a "simple" OBU for DSRC will be treated as nonequipped on all roads that are tolled exclusively by means of a GPS/GSM solution. There is no room for interoperability in this direction.

1.4. GPS/GSM-Systems and Interoperability

Countries that introduce a GNSS/CN (autonomous) solution, of which GPS/GSM is a version, cannot yet ensure interoperability with other schemes. Up to now, there are no autonomous solutions for EFC in operation. Germany will – when/if the system is implemented – be the first country to rely fully on this technology for the purpose of HGV tolling on motorways. But if, in the future, several autonomous solutions will co-exist in Europe, their technical interoperability should be possible. The same holds for technical interoperability of GPS/GSM-equipment on DSRCtolled roads. However, full interoperability also requires procedural and contractual interoperability. First attempts have been made to develop a standardisation of procedural aspects, but as there are no real-world experiences, these attempts have not produced very tangible results. With respect to contractual interoperability, no action has been taken so far.

For the GPS/GSM technology, no analogous action to CARDME/CESARE has been launched by now. However, ERTICO (a Europe-wide, not-forprofit, public/private partnership for the implementation of Intelligent Transport Systems and Services) intends to carry out an integrated project on road charging interoperability (RCI) in the 6th Framework Program of Research and Technological Development (6FP) in the European Union. The vision of RCI is to establish an open system architecture for road charging systems which will incorporate DSCR based and GPS/GSM based technical building blocks. The EU directive from April 2004 on interoperability of EFC systems (Council of the European Union, 2004) states that OBU's shall at least be interoperable and capable of communicating with all systems in the Member States using one or more of the technologies (satellite positioning, mobile communications using GSM-GRPS standard, 5.8 GHz microwave technology (DSCR)) and they should be made available to interested users.

1.5. Non-Equipped Users and Interoperability

For a specific EFC system, a user with technically or contractually noninteroperable payment equipment is like a non-equipped user. Both nonequipped and non-interoperable users reduce the cost-effectiveness of an IRPS based on EFC due to special payment procedures and/or reduced ability for enforcement.

Non-equipped users can in principle be banned from an electronically priced road, but with several difficulties. It is possible to make the installation of an OBU mandatory for domestic vehicles in one country (e.g. Switzerland). Even mandatory equipment for all users (including foreigners) is possible, as the example of Austria will demonstrate. Mandatory equipment will not face substantial resistance if the OBU's are given to the hauliers free of charge or at a very low price, and installation is quick and easy. However, if an internationally operating heavy good vehicle (HGV) needs to install different equipment for a multitude of European EFC systems, the problem of non-interoperability is not solved, but passed on to the hauliers. This development underlines the importance to improve interoperability between existing and newly developed EFC systems in Europe as a means of reducing the number of non-equipped users. Should a pan-European EFC system be introduced, mandatory equipment of vehicles would certainly make sense.

The discussion takes a different dimension if tolling is to be applied to all vehicles. Through EFC, detection and location of private cars (almost coincident with individual persons) is made, which can raise questions about invasion of privacy rights. However, this can be avoided by recourse to OBU's with anonymous debit card systems. But the legal question of imposing adherence to an EFC system (personalised or anonymous) as a condition to use the motorways or roads in a certain region or country should still be addressed.

An easy and reasonably reliable technical solution for non-equipped users seems to be available: video capture of licence plates, followed by automatic character recognition and direct billing to the vehicle owner. However, there are serious difficulties in general application of such a scheme. Not only will some licence plates be dirty or deteriorated (and thus unreadable), but there is presently not enough cooperation between national authorities to allow direct access by entities (even official entities, let alone private concessionaires) of one country to the national car registries of another. So, foreign vehicles would be uncharged, and that element is sometimes central in support of fairness arguments and to improve acceptability.

Another hard question is that of financing the equipment for all users. If this is passed to the vehicle owners, many will complain; if this is subsidised by motorway concessionaires, it can be quite expensive given the volume of the subsidised set.

1.6. Roaming Charges in EFC-Systems

Roaming charges in EFC systems are surcharges added to the actual cost for using a "foreign" tolled road network with an interoperable OBU (very similar to current pratice in mobile telephone services). Since users with a foreign OBU cause higher costs to the operator, the latter could want to claim a roaming charge. However, the idea of the Single Market based upon non-discrimination of foreign HGVs does not allow charging foreign vehicles higher than domestic ones. Roaming surcharges in road pricing seem thus unacceptable within the EU. An identical agreement has been reached in 2002 concerning use of ATM machines within the EU by bank account holders in an EU country other than that where the machine is installed. But for GSM phones, substantial roaming charges are applied and have been subject to frequent complaints by user groups.

2. INTEROPERABILITY CONVERGENCE FROM A EUROPEAN PERSPECTIVE

The importance of interoperability for an efficient European road network has already been pointed out. The difficulty to reach convergence and interoperability, and the possibilities of the European Union to improve the interoperability of EFC systems throughout Europe within a shorter period of time is addressed in the following sections.

2.1. Recent EU Effort to Improve Interoperability of EFC

So far the European Union (EU) has tried to improve interoperability of EFC through e.g. several research projects and has also tried to set the basis for interoperability between different operators in Europe (CARDME-4, 2002; CESARE, 2001; MOVE-It, 1998, etc.). In its communication COM (1998) 795 final, the European Commission outlined a Community strategy for the convergence of interoperable electronic fee collection systems to facilitate traffic flow and the payment of fees. This communication examined the obstacles to interoperable electronic fee collection systems and put forward certain recommendations for arriving at an appropriate level of interoperability on a European scale.

2.1.1. EU-Directive on Interoperability of Electronic Fee Collection

More recently, the European Council and Parliament approved in April 2004 a directive (2004/52/EC) on the future interoperability of the various systems from the user's perspective (Council of the European Union, 2004). The Directive announced in the White Paper "European Transport Policy for 2010: Time to decide" would establish a single European toll service, based on the principle for users of "a quick and easy way of paying infrastructure charges, using the same means of payment throughout the network without losing any time at toll stations."

The Directive will be the regulatory frame for the deployment of a unique European EFC service in two stages in Member States already having EFC:

HGV (exceeding 3.5 tonnes) and vehicles allowed to carry more than nine passengers at latest by mid-2009.
all vehicles by mid-2011.

This aim is reflected in the EC Directive on the widespread introduction and interoperability of electronic road toll systems in the Community (Council of the European Union, 2004). The laws, regulations and administrative provisions necessary to comply with the Directive should be brought into force by the Member States by end of 2005. The directive lays down the necessary conditions for the implementation of a European electronic toll service, which will be based on the following basic principles:

Based on interoperable system, thus "one single contract, one single on-board unit per vehicle," and later possibly "one single invoice."
It should be available on the whole tolled network.
It is expected to allow the same quality of service in any country, not depending on the country where the contract is signed or the nationality of the driver.
Offering the service will be mandatory for operators and the subscription of the service will be voluntary for the users.

The Directive is not supposed to interfere with the pricing policies of the Member States, but the systems implemented should be capable of handling any charging policy decided at national levels. The directive does not apply to systems using non-electronic means of toll collection or OBU's and on strict local road toll systems.

According to the directive, all new EFC systems brought into service on or after January 1, 2007 on all types of roads in the Community, should use one or more of the following technologies.

Satellite positioning (GNSS).
Mobile communications (GSM/GRPS).
DSRC solutions based on 5.8 GHz microwave technology.

Any new technologies should, when introduced, at least be interoperable and capable of communicating with each other, and existing tolling technologies should ensure full compatibility and interfacing with the technologies defined in the Directive, as well as with each other.

Regarding technological aspects the Directive recommends that new EFC systems brought into service use satellite positioning and mobile communication. A study should be carried out by December 31, 2009, on the possible migration of other systems (e.g. DSRC) towards these systems. All technologies in the Directive should be described including a cost-benefit analysis.

The European Commission has in 2002 and 2003 carried out Consultation meetings with Member countries during the preparation of the Directive on Electronic Fee Collection, according to which all tolled motorway operators should install technical equipment that is interoperable. The problem raised with this proposal is that it dismisses agreements reached at CEN, where existing systems were taken into account, and focuses instead on the interests of "the majority of European suppliers," ignoring the conversion costs for concessionaires and over 2 million users. Besides the doubts about the feasibility of an announced cost reduction of the more complex OBU, this proposal has not been very well received, as it forces the majority of countries with tolling systems, which are based on DSRC, to migrate to a more complex and expensive system, especially on operating costs. Thus, the planned study in the Directive will look at the feasibility of the different technologies.

2.2. EU Possibility to Ensure Interoperability of DSRC and GPS/GSM Solutions

With the support of the already mentioned research projects and the new directive on interoperability, the European Union has played and still plays an important role on the way towards interoperability. Now, it is up to the players in the field to use this spadework and to apply the Directive. Most operators will have an interest to harmonise as far as possible for reasons of cost-effectiveness. New players building up a DSRC solution will certainly use the CARDME standards (even if it's only for the cheaper, off-the-shelf equipment) and the requirements of the new Directive.

Even with the new directive it may be difficult for the European Union to do anything in the near future to improve interoperability of GPS/GSM solutions. In August 2003, Germany was expected to introduce the first area-wide, commercially used GPS/GSM road charging system. It would then be likely that two standards emerge, one based on DSRC and another on GNSS/GSM. Onboard units should be available separately for these two standards, as well as joining them in a single device. However, the German system has – at the time of writing (spring 2004) – been delayed and Germany may, if the new deadline is not kept, return to the Eurovignette system.

Only if the costs of the dual unit are rather low would it make sense to have only one type of OBU, with full dual capability. Otherwise such capability would only be acquired for those vehicles which are expected to travel frequently to countries using one and the other system.

Moreover, there is the question of "open specification" or "closed specification" of the technologies involved, and its implications on monopoly of supply of the corresponding OBU's. In Holland, it was decided by the Government before introduction of the Rekeningrijden project, that all specifications had to be open, in order to allow multiple manufacturers to develop and market the OBU's, and possibly also include added-value services on their top models, but in Germany the option has been for a closed specification, thus giving the concessionaires/ manufacturers of the OBU's a monopoly on a virtually compulsory equipment for the hauliers. This seems to be contrary to competition rules in the EU, and will probably be challenged some time in the future, probably closer to the start of operations of the German tolling system when/if this happens. The fact is that there seems to be little sense in interoperability across tolling schemes while one of the key players keeps his specifications (and its market through it) closed. So, a very important contribution of the EU to improve interoperability of GPS/GSM solutions would be to impose that their specifications be open as suggested in the new Directive, thus favouring not only price reductions of the simpler OBU's in the domestic markets, but also earlier and cheaper introduction of interoperable OBU's in the market.

2.3. Harmonisation and Vehicle Classification

Even if harmonisation in this field is not absolutely required in order to achieve interoperability, it would certainly facilitate comparability and simplify the users' understanding of charges if the classification of vehicles is harmonised. To enable EFC systems to fulfil the likely requirements of future charging and pricing legislation, the set of classification parameters includes in addition to the usual vehicle characteristics, environmental attributes, such as emission or noise characteristics. With the definition of the EURO emission categories, a step in the direction of harmonisation has already been taken; some optimism is permitted here because vehicle classification does not seem to be a very sensitive issue for the Member States. Since this harmonisation of classifications has implications on existing concession contracts, price levels in the various classes should be adjusted in each country so that total revenue is not affected.

2.4. Interoperability Without Legal Pressure from the EU

Full interoperability between all European EFC may be expected within the next 10 years depending on the 2009 study suggested in the new Directive on cost and benefits of different technologies and depending on the legal pressure from the EU. However, some countries are presently building up new systems based on DSRC while others are introducing the advanced technology of GPS/GSM. In this situation, the Member States may be less willing to – without legal pressure such as the new Directive – to accept any pressure from the EU which would force them to abandon their solution. The difference in existing technical solutions will therefore persist depending on the outcome of the study mentioned in the Directive 2004/52/EC.

However, it can be expected that the market forces create interoperability where it is economically reasonable. In the absence of an easy-to-use pan-European EFC, or relatively cheap dual-use on-board units, it is even undesirable to have everything interoperable just because it would be technically possible, since each step of interoperability requires costly negotiations and payment procedures. For example, it would be inefficient to make all Portuguese trucks interoperable with the Greek motorway network. It would be cheaper if the few Portuguese trucks that will ever use Greek roads pay their charge manually. If the German GPS/GSM experience proves to be successful and if other big European countries should change to this technology as suggested in the new Directive, it is imaginable that HGVs will be equipped with a standardised GPS/GSM + DSRC equipment ex-fabrica in the future. Such vehicles could circulate on all European EFC roads.

2.5. Interoperability and Europe

To sum up, full interoperability in Europe may be reached within the next 10 years, depending on market forces and the success of the new Directive which can stimulate the definition of standards and will create interoperability where it is economically reasonable, thus making the adequate level of interoperability for many hauliers available within a shorter period of time.

The vast majority of HGV's registered in any European country travel almost exclusively within their own country, and thus under a scenario of multiple technological solutions in place across Europe. It should be possible for vehicle owners to select the spectrum of EFC solutions with which they want each of their trucks to converse fluently. As long as the costs of on-board-units are significantly higher when they are to provide this flexibility than just for the home market, two or more degrees of OBU sophistication should be expected in the market and accepted by member States and the European Commission.

Interoperability of EFC is an interesting feature, but it should not be imposed to the whole fleet and rather made available as the market requires it, for the relevant vehicles. For the European Commission, it should not be a matter of imposing it, but of circumventing or mitigating the barriers that might prevent it as suggested in the new Directive.

3. INTEROPERABILITY AND HARMONISATION

Achieving full interoperability on all three levels (technical, procedural and contractual interoperability) requires substantial investments in planning and in implementation. Therefore, high interest of international bodies and institutions is a prerequisite for investigating all issues involved.

3.1. Institutional and Contractual Approximation

There are basically three different institutional set-ups: all public, all private and a mix of public and private. Harmonisation of institutional arrangements is not necessary to ensure interoperability and approximation of systems in Europe.

To ensure interoperability the participating operators have to enter contractual arrangements between each other to recover the payment for use of their infrastructure. This should also cover sub-contractors responsible e.g. for financial transactions such as banks, credit cards, petrol cards.

Contractual interoperability means the contractual arrangement of every operator system concerning acceptance of third-party OBU's, accounting technique, guarantee of payment, etc. In Europe this level of interoperability is still in an early stage and thus several contracts and/or OBU's could be required. Instead of the users having several contracts with different operators it would be more desirable to reach interoperability through agreements between operators to allow that each operator accepts other operators' contracts and handling of classes as well as OBU's.

Work done in CARDME (CARDME-4, 2002), MANS ( ˚ MANS, 2000 ˚ ) and CESARE (CESARE, 2001) suggest the adoption of common standards and procedures for EFC based transmissions and transactions in Europe. The focus is on voluntary agreements between the owners of a variety of national IRPS – Memorandum of Understanding – and contractual conditions for allowing road users to subscribe to different EFC systems and thus to roam.

In the process of preparing the new Directive on interoperability of EFC approved by the European Council and Parliament in April 2004 it has been suggested that the operators should enter contractual agreements allowing users of tolled roads the same conditions everywhere and the possibility to subscribe on the whole network. The contractual agreement could be based on the Memorandum of Understanding produced by CESARE II (CESARE, 2001).

There is ongoing work to define and establish an international Memorandum of Understanding (MoU) for motorway tolling operations e.g. through different EU supported R&D projects. The MoU provides the framework for contracts between road operators, EFC, operator, bank and credit card issuers as well as a standardised contract between the contract issuer (e.g. toll operator) and the user (e.g. operator-operator, operator-user, operator-payment service provider, payment service provider-user).

Contracts would mainly contain details about payment forms and equalisation payments. If the operators use different technical systems (e.g. road side vs. autonomous) there are some technical implications to deal with. Therefore, the MoU may also include technical specifications for the OBU, RSE and the wireless communication (DSRC) and security architecture for existing types of systems. A future MoU may also include specifications for e.g. GNSS and/or CN (Global Navigation Satellite Systems and Cellular Networks). Basically the payment systems will remain the same.

Final contract arrangements should in the future offer the possibility to support local contracts as well as international ones. Contracts between operators across borders would ideally be based on the above mentioned MoU. Users might be offered the possibility to sign multiple contracts within the MoU. Contracts between operators and users will allow the user only to have a contract with the "home operator" and then to use the other systems being part of the MoU. This should in the future be possible for road charging systems in all European countries, similar to what has occurred in the development of the vehicle insurance system: some years ago car insurance was only valid in each country separately, and travelling abroad required an extension (and additional payment). Currently, EU drivers receive a "green card" through which their car insurance is valid throughout Europe.

It is important that national authorities as well as the EU take early initiatives to use legal instruments to set the evolutionary avenues without undermining useful technological developments (CARDME-3, 2001).

Some examples of interoperability between fee collection systems are already operational in e.g. Helsinki and Florence where interoperability has been achieved between different applications such as fee collection for road usage, public transport, and park and ride. Also in Portugal, the Via Verde EFC system is applied not only across different motorway operators' networks but also in parking garages, access control to protected urban areas, petrol stations, etc. Full interoperability between Portuguese and Spanish EFC schemes is expected for 2006.

Changes in legal and contractual functions are much larger and deeper in the case of DAREA schemes, as this will in most cases affect not only motorways, but also national, regional and local roads within a certain perimeter. This implies first of all some agreement from the Government, involving the various levels of public administration, regarding the overall justification for tolling (as an option weighted against other policy alternatives), the possible exceptions, the allocation of responsibilities for the various operational functions and of course the distribution of revenues. And, naturally, where tolled motorway concessions are in place, conversion to a DAREA scheme would always be difficult and probably cause the existing contract to be denounced by the Governments, with full indemnity paid to the concessionaire.

3.2. Procedural and Operational Harmonisation

A correct workable toll procedure has to be created and therefore a minimal functionality should be defined to achieve procedural interoperability.

Procedural and operational harmonisation across Europe may be more difficult to obtain due to the many existing payment means and methods to users of tolled roads. Therefore the users should first of all be able to extend their existing payment means and method to cover other operators.

For new countries, operators or concessions interoperable payment methods may be introduced from the outset. The new on-board equipment should be designed to work in different operational environments. Norms CEN TC278 WG1 and CEN ISO 14906 define data elements and the standardised functionality. The procedural interoperability depends on what data (such as black lists, EFC-Context Mark, distribution and management of keys etc.) are exchanged and how they are exchanged between the operators.

How the user data are managed by the operator is partly a local issue (e.g. personalisation and transaction), but also an issue between operators (i.e. the required data for the transaction at the charging point is available in the OBU). If interoperability between different road networks is considered to be an important requirement, questions of transparency and uniformity of procedures related to the user become critical, as well as those related to charges between operators, and possibly even to relations between States.

With respect to procedural aspects for IRPS the main problems are handling of non-equipped users, acceptance and handling of exemptions. If mandatory equipment with OBU is not possible, the operator is forced to design a Dual System. System access and handling of non-equipped users will remain the most difficult issue to be solved in an EFC system. In particular, DAREA schemes with complex tariff schemes with dynamic fee levels according to daytime, congestion, etc., are not feasible with non-equipped users.

3.3. Technical Standardisation

There is no single fee collection system but almost as many systems as there are applications. Different IRPS technologies are already in use and new ones are planned to be introduced by different operators and for different purposes. To ensure interoperability across Europe some technical standardisation is necessary. When looking at technical standardisation of similar tolling systems with the same purpose – e.g. financing of roads – it seems possible to ensure some interoperability at the European level during the coming years.

For different types of systems with different and sometimes contrary purposes – e.g. financing and traffic management – it is more difficult and only certain parts may be standardised in the coming years, e.g. designing OBU's to work in different operational environments. New systems may from the outset use standardised technologies while existing systems initially will exist as they are and over time develop towards more standardised technologies as suggested in the new Directive.

DSRC systems seem appropriate for a big number of equipped users or for mandatory equipment of all users. Due to the need of extensive road side equipment, DSRC will primary be used for NET schemes. DAREA schemes can only be equipped in special circumstances like in Switzerland, where the access to the area is precisely defined, with a small number of "gates" in relation with the toll application area, and domestic vehicles are legally forced to be equipped with an OBU.

Based on the DSRC technology every third-party system should be able to communicate with a national system to guarantee technical interoperability. Therefore, there exists a European norm for DSRC communication. The experience has shown that the level of technical interoperability is practically possible. However, there are several technical issues to consider, e.g. to implement DSRC protocol options (e.g. one or two sub-carriers associated with Profiles 0 and 1, retransmission procedures), to implement EFC functions (including security algorithm and keys) and data elements (including their coding, and qualification of the recognised user equipment).

The following are some examples of equipment that interoperate within "one system:"

T.I.S (Liber-T in France): RSE from Saab CombiTech (CT) and Thales, OBU from CSSCI, Thales and Q-free.
Sydney Harbour Bridge Tunnel (SHBT): equipment from Kapsch, CT and Q-Free (also discussions on interoperability with Tran-urban City Link Melbourne, Eastern Distributor, M5 etc.).

Cross border/system interoperability is not implemented anywhere yet. One of the first examples could be the usage of the Swiss LSVA-OBU on the LKW-Maut in Austria, but other bilateral cases will probably occur in the near future.

The need for interoperability should be considered carefully as the average private driver is mainly travelling within a small region. However, urban systems in coexistence with NET systems should have similar technical systems for the convenience of the drivers. NET schemes can be covered with both types of technologies (DSRC & GPS/CN) as seen of the two planned systems in Austria and Germany. The system in Austria is a "classic" DSRC system providing gantries and beacon on each tolled motorway section while the system in Germany is expected to be the first big scale autonomous system using GPS and CN for location and communication. The road side equipment will be limited to enforcement stations.

Autonomous systems are most appropriate for complex tariff schemes (e.g. different tariffs depending on road hierarchy). The more complex and flexible the tariff scheme is planned, the more appropriate an autonomous scheme will be as the road-side equipment for DSRC Schemes would become far too expensive and even with high investment costs, a coverage of time, congestion and/or level of road dependant tariff would not be feasible. For DAREA schemes adapted with traffic management purposes and/or with tariff differentiation in space or in time, basically only autonomous systems are feasible. DAREA is practical for commercial transport with heavy goods vehicles (HGV, example Swiss LSVA) but may in some cases not be practical for all vehicles. A domestic requirement for compulsory OBU would face a significant opposition and for handling the nonequipped foreign users there is not enough space and personnel at the borders (and at many borders within the EU, e.g. between countries in the Schengen agreement, no personnel at all).

Full roaming can only be achieved by providing both technologies (DSRC & GPS/CN) with all necessary features and covering international standards in one OBU. An on-board equipment with a high level of functionality (e.g. DSRC, GPS, GSM, odometer-reading, human-machine interface) offers theoretically the possibility for added value telematic services such as traffic information, fleet management systems, emergency systems and others. In the (near) future, many HGV's engaged in international transport will probably be equipped with fixed installed equipment like GPS-, CN-, DSRC-module and digital tachograph etc., which could cover/support the requirement for full roaming in connection with an existing OBU. Some efforts are needed to define/standardise the interfaces. It is likely that, until all vehicles come equipped ex-fabrica, introduction of interoperable OBU's will be made gradually, with voluntary adhesion (and extra payment) only by drivers who frequently make use of road networks outside their own region.

4. CONCLUSION

The "desired" system is a combination of (distance dependent) systems which are interoperable both technically and contractually, e.g. one OBU (operational in the whole Europe), one contract and only payment to "own" operator. This is in line with the EU objectives in the new Directive and as stated in the White Paper "European transport policy for 2010: time to decide: that users should have: a quick and easy way of paying infrastructure charges, using the same means of payment throughout the network without losing any time at toll stations." However, to reach this objective a number of issues are critical at both the European level and at national level.

Interoperability is an important element of convenience and efficiency for international hauliers, but the variety of starting conditions raises the costs of providing it across the board. Given that most hauliers are dedicated to domestic markets, and others are dedicated to specific foreign markets, interoperability should be provided "a la carte" as the market forces find it interesting. The same ` majority containment to domestic road networks applies to private cars.

However, there are possible important contributions from the EU in the short and medium term:

To impose open specifications for on-board units (OBU) in all countries.
To encourage convergence of standards in the GNSS/GSM platform (as it has done for the DSRC platform) and as suggested in the new Directive 2004/52/EC.
To encourage convergence of vehicle classification.

4.1. Critical Aspects at European Level

A new Directive (2004/52/EC) was approved in April 2004 to ensure EFC interoperability on the European road networks. According to the Directive all new EFC systems should use one or more of the satellite positioning, mobile communication and/or DSRC systems and be able to communicate with each other. A time frame is also suggested in the Directive.

However, the absence of a European framework and harmonisation of IRPS until April 2004 has made it critical to move to an interoperable standard. This is clearly seen for the mature countries planning to introduce an IRPS soon. There are movements in different directions with regard to the basic form (Austria/Germany, NET; Netherlands, DAREA). Also with regard to the technical solution the mature countries are moving in different directions (Netherlands/Germany, GPS; Austria, DSRC). To ensure interoperability, the EU is required to provide the necessary framework e.g. through minimum technical standards ensuring interoperability as well as a framework to ensure contractual interoperability or roaming as it has now done with the new Directive on interoperability for EFC systems.

The traditional tolling countries are all based on national systems decided in the past that are now moving (or have already moved) towards national interoperability. To ensure that new concessions on new toll roads or new concessions on previously concessioned roads in traditional tolling countries move towards international interoperability a European framework is also required. The new European framework (Directive) will support that both traditional tolling countries as well as the mature countries planning to introduce an IRPS will move in similar direction.

The introduction of national IRPS by several EU countries has led to different technical and institutional solutions being used within the EU. This would not be a problem if harmonisation of the legal framework and technical standards had emerged at EU level. With the growing number of national systems and different concessions and public-work contracts established or planned there is a growing complexity that must be taken into account when implementing the new EU framework for harmonisation (Directive 2004/52/EC). It seems to be taken for granted that "new rules for pricing and funding transport infrastructure do not have to undermine current concession systems." But if the existing systems – new and old – are not allowed to limit the flexibility when specifying standards at EU level these systems will have to be upgraded to European standard which will require additional costs for operators and users, as well as possible re-negotiation of concessions. Alternatively, the time horizon for concessions to be renewed will affect the implementation/harmonisation.

4.2. Critical Aspects at National Level

The different objectives of IRPS in the countries may be a critical issue to move to a future interoperable situation. Countries wishing to manage transport demand through e.g. IRPS require more complex systems and toll differentiation (frequently implying a DAREA scheme) than countries using IRPS to cover infrastructure costs (NET).

Even with the introduction of the DSRC standard on 5.8 MHz, full interoperability is still not ensured. The OBU's for the new systems following the standard may not be immediately interoperable but will still need a change of some of the protocols. Interoperability between OBU's for new and old DSRC systems may not be possible as well as interoperability between OBU's for DSRC and GPS systems. New GPS OBU's may be made interoperable with DSRC systems by adding a DSRC interface but it will be difficult for simple roadside based DSRC OBU's to become interoperable with GPS OBU's.

The users should be able to have a contract with a "home operator" and then use the other systems part of the MoU, which in the future should be possible for all road pricing systems in Europe as described in the new Directive. The MoU provides the framework for all levels in an EFC system and includes details about payment procedures and equalisation but could also include technical aspects including specification for autonomous systems. One of the really critical issues of international interoperability is the question of the possible roaming surcharge discriminating foreign users to the national.

The newcomers (countries who have not yet decided to introduce an IRPS) may in principle move to any of the future situations especially as long as the European framework was not in place. The newcomers could basically decide either to join or follow the systems or paths of e.g. the mature countries, they could also decide to take the lead in the European development, choose to establish their own national system or finally they may decide to keep the Eurovignette and wait and see and use the new European framework and standard. It would be desirable that the newcomers are not introducing their own national system independently of other existing systems, but that they follow the new European framework.

REFERENCES

CARDME-3 (2001, September). Strategy for the introduction of interoperable electronic fee collection in Europe.

CARDME-4 (2002, June). The CARDME Concept, Del. 4-1.

CESARE (2001, December). Constrained Estimation of Accuracy with Recursive method. CARDME Steering Committee.

Council of the European Union (2004). Common position adopted by the Council with a view to the adaptation of a Directive of the European Parliament and of the Council on the interoperability of electronic road toll systems in the community (Directive 2004/52/EC) presented in the Official Journal of the European Union 30th April 2004.
EFC AID (2003). EN ISO 14906, Road Traffic and Transport Telematics (RTTT) Electronic fee collection – Application interface definition for dedicated short range communication, 2003.
EN L1 (2003). EN 12253, Road Transport and Traffic Telematics (RTTT) Dedicated Short-Range Communication (DSRC) – Physical layer using microwave at 5.8 GHz.
EN Profiles (2003). EN 13372, Road Transport and Traffic Telematics (RTTT) Dedicated Short-Range Communication (DSRC) – DSRC profiles for RTTT applications.
European Commission (2001). White Paper, European transport policy for 2010: Time to decide, COM(2001) 370, Brussels 2001.
MANS (2000). Doc. 2.1. Guidelines on Automatic Payment Using Central Account. ˚
MOVE-It (1998). Motorway Operators Validate EFC for Interoperable Transport. Activity 031: Operator Specification for Contractual Interoperability Deliverable 3.2.
PISTA (2004). Pilot on Interoperable Systems for Tolling Applications, (www.pistaproject.com).
TIS.pt, COWI, ECOPLAN, UNIKARL, ISIS (2003). Recommendations on IRPS for HGV, DESIRE Deliverable, 6th May 2003.

9. ECONOMIC AND ENVIRONMENTAL IMPACTS OF ROAD TOLLS FOR HGVs IN EUROPE

Lorenz Kleist and Claus Doll

1. INTRODUCTION

The reform of pricing regimes in transport, as in other economic sectors, will not be possible without consequences. These consequences can be positive or negative, may take place on the short term or in longer terms and will depend on the condition of transport supply and demand relations as well as on economic parameters. For the acceptability and the appropriateness of particular pricing schemes under specific conditions it is of utmost importance to get an understanding of possible dangers arising from the introduction of road pricing systems. Thus, this chapter provides an assessment of various scenarios of road pricing systems using several sophisticated tools in order to capture as much impact dimensions as possible. (Compare Doll et al. (2003) for a more detailed description.)

This chapter is organised as follows: Section 2 provides an overview of possible impacts on transport flows, the economy, traffic safety and on the environment. Section 3 then presents the instruments to analyse these impacts and Section 4 presents the scenario assumptions and the results by type of impact. Finally, Section 5 draws some conclusions.

2. IMPACT DIMENSIONS AND PATTERNS OF REACTION

2.1. Adaptive Behaviour Within the Haulage Business

The most direct reaction of shippers on the introduction of road user charges on particular network parts or areas is to detour, avoiding these network parts as much as possible in order to save money. In case of raising tolls on motorways traffic will detour to the secondary network and in case of a national pricing system on all roads hauliers on international trips may choose routes through other countries. However, the length and the traffic conditions of alternative routes will limit these reactions. Route shift effects belong to the negative impacts of road tolling as secondary roads are usually more sensitive to pavement degradation, congestion, environmental and noise pollution than the motorways.

IWW (2001) estimates, that under the scenario of tolling HGVs on the German motorway network by 0.13 Euro/km the overall share of traffic, which will be displaced to the trunk road network, averages 7% of regional trips (up to 150 km) and 3% in long-distance haul. Most remarkably, also in an alternative scenario, where the entire inter-urban road network is priced equally, a displacement of motorway traffic to the trunk road network of 2% was estimated. This can be explained by the fact that the road tolls let the distance-related component in the hauliers' decision function get more important, while the time-component remains constant.

In a more strategic way the haulage business can react on cost increases by making better use of their vehicle fleet. This may either happen by improved tour planning, by vehicle pooling or by mergers. While the pure increase of vehicle load factors clearly is a positive development, the merging of firms often constitutes the consequence of smaller haulage companies vanishing from the market as they can no longer withstand competition.

IWW (2001) concludes that by merging of firms, by the establishment of vehicle pools and by improved tour planning, increases in vehicle load rates and reductions in empty running up to 15% can be achieved in cases of a high area pricing scheme. The additional costs of those measures for the hauliers are balanced by productivity increases in the management sector. However, the traffic-relevant productivity increases caused by road tolling are likely to be very small as further increasing competition in the future anyway will force hauliers to make better use of the available vehicle stock and labour force.

Moreover, the shippers can react on changing pricing regimes by shifting towards more cost-effective vehicles. In case of a tolling system for HGVs over 12 t vehicle gross weight, costs can be reduced by using lighter vehicles, which are not subject to tolls, or by employing heavier vehicles, which are more cost effective per ton of goods. While these effects are usually of a negative nature, the adaptation of the vehicle fleet composition as a consequence of emission-differentiated charges is a positive development supporting the policy on regulating emission standards of new vehicles.

One can also observe cost saving activities, which are beyond legal forms of competition but not severely penalised in Europe. This concerns the overloading of vehicles, the violation of regulations and social dumping practices. It can be expected that a higher cost pressure through road user charges has also an impact on such practices. This does not necessarily mean that road user charging tends to increase the illegal operations in the haulage market. There are also clear indications that employment of low cost drivers or cabotage practice of firms located in low wage countries is reduced. This is due to the fact that cabotage and dumping freight traffic activities imply longer distances on the road network. As soon as there are distance-related charges this incentive is reduced.

2.2. Shippers' Choice and Intermodal Competition

It is the hope of politicians that through the introduction of road pricing systems the cash-strapped railways will gain back considerable market shares. However, price is not the only variable determining modal decisions. In particular the drop of the demand for bulk goods, the increasing importance of unitised cargo, justin-time production structures and the regional dispersion of manufacturing and demand locations push flexibility, reliability and speed of shipment alternatives into the foreground of the decision framework. The fact that, even though road congestion was steadily growing over the past decades, truck transport has become the predominant freight transport mode and is still increasing its market share shows, that the adoption to the change in the freight transport market demands considerable efforts from the European railways. While most national rail carriers do not seem to take this challenge, the example of the Trans-Alpine railway market from Denmark to Italy which is served by a single company demonstrates, that offering one-stop-shops and fast services meets market requirements and is an important instrument to stop the decline of the railways' market shares.

Therefore, IWW (2001) investigates modal split reactions with and without an improvement of the railway's service quality. In the model it is assumed that shippers only react when a particular level of cost increases is reached. Thus, in the moderate scenarios of pricing only on motorways the growth of rail freight volumes is hardly above 1%, while in the high area price scenario combined with a quality improvement by railways this mode can increase its transport volume by 14%. These reactions of course differ by goods classes and distance bands.

Shippers basically have two alternatives of reacting on changes in the cost structure of transport: Reducing the amount of shipments through replacement of purchased goods by in-house productions or changing locations of their company, of their suppliers or their customers in order to reduce travel distances. Eberhard (2000) has shown, that the share of transport costs in relation to production costs and to the costs of re-location is too little for the introduction of road pricing systems to have a significant effect on location decisions. However, this statement depends on regional economic characteristics and on the goods market in question.

2.3. Product Prices and Productivity

Several investigations lead to an expectation that between 20% and 40% of the costs imposed on the hauliers by additional HGV tolls can be absorbed within the transport sector. Some of these effects are of negative nature for society (e.g. traffic shifts to the secondary network), but most of the effects are positive, as they reduce vehicle kilometres driven, and thus reduce environmental loads, noise disturbance, accidents and congestion.

The remaining 60%–80% of costs need to be carried by the hauliers and forwarders, i.e. by the shipping industries or by the consumers by increased product prices. It has already been stated that the shipping industry might react by changing modal choice or by re-structuring purchase, production and distribution mechanisms. Since the cost effectiveness of these measures could not be estimated in the framework of the DESIRE project it was assumed that their net effect is not more than a 5%–10% reduction of the extra costs caused by the new tolling structures.

The possible reactions of industries on price changes are as follows:

Increase of consumer prices.
Reduction of demand for input factors.
Location change.
Reduction of tax payments, etc.

Of course, the revenues from road pricing can be used to compensate some of these effects.

3. THE DESIRE MODELLING APPROACH

3.1. Selection and Linkage of Two Economic Models

The previous section has shown, that the introduction of road pricing schemes entails numerous reactions by different actors in all levels of the economic decision-making and that some of these impacts are inter-connected to each other. Thus, the quantification of impacts on the one hand requires the consideration of micro-economic behavioural patterns to understand what happens inside the transportation sector and the related sectors of economy, and on the other hand a macroscopic perspective must be taken in order to capture the total level of impacts on Europe's national economies.

The complexity of these issues is addressed by the application of two different models: The VACLAV road transport network model and the ASTRA-T system dynamics platform. VACLAV is a European multi-modal transport generation and network assignment model. Within the DESIRE project the model was applied to compute hauliers' reactions on the tolling of particular infrastructures across Europe, taking into consideration local network specificities, such as congestion, border crossing times or speed and weight limits.

ASTRA-T is a system-dynamics model, which computes the development of numerous variables and indicators over time by considering their inter-dependency. The model is working on a European scale by differentiating between countries and four functional zones within each country. The transport module of the ASTRA-T model distinguishes between several distance bands, travel purposes and goods types of all relevant transport modes and is thus capable to predict changes in modal choice. However, it does not contain physical transport networks, which makes it difficult to forecast impacts of certain network-related policies. To overcome this gap, the ASTRA-T model was linked to the VACLAV network model by passing information on the change of transport times, distances and costs entailed by specific pricing scenarios from VACLAV to ASTRA-T. ASTRA-T then computes changes in demand structures, which are used to periodically update the transport generation module of the VACLAV model. Figure 1 presents the linkage of the two models in more detail.

Instead of adding a third analysis tool, microscopic information on the behaviour of various actors was used to define the behavioural functions within the two models. This comprises information on threshold cost levels, triggering specific adaptation reactions of hauliers, shippers and firms, affinities of market segments to specific transport modes or the ability to pass on cost increases from hauliers to shippers, producers and to the final customer. An overview of the two models and their specification is presented in the subsequent sections.

Fig. 1. Linkage of VACLAV and ASTRA-T for Policy Analyses.

3.2. Direct Reactions Within the Road Sector: The VACLAV Model

Originally based on a model for Germany VACLAV has been enlarged to a Europewide forecast model for long distance passenger and freight traffic. A network flow model has been created, which follows the classic four-step approach of trip generation, trip distribution, modal choice and trip assignment. In the framework of the DESIRE project, however, the model has been used for the assignment of freight transport volumes to the road network and for the calculation of times and costs between the regions only. The steps of generation, distribution and modal split are performed by the ASTRA-T model, which makes use of the impedance on the relations calculated by VACLAV.

Initially the model was restricted to the area of the European Union. Within the project for the traffic forecast on the ten PAN-European corridors (NEA et al., 1999) the model has been extended to Central- and East Europe. Almost all projects have been handled by international consortia. Therefore it was possible to extend and refine the databases with traffic or socio-economic data in most countries with the help of local organisations.

The geographical area covered by VACLAV comprises the whole Europe. The zoning system is based on the NUTS-III nomenclature within the European Union. Due to differences on the data availability the zoning in the CEE-countries ranges from NUTS-III to NUTS-0. Overall there are 1200 zones defined in VACLAV.

VACLAV covers all inter-zonal long distance trips (trips length at least 50 km). The demand is segmented by three trip purposes and four modes. Although the first purpose of VACLAV has been the forecast of passenger demand, also models for freight forecasts are included especially the road freight assignment step is well developed with support of the NEAC model. Within the projects "TEN-STAC" (NEA et al., 2003), "Forecast2020" (IWW & MKmetric, 1999) and "Forecast on the 10 Pan-European Corridors of Helsinki" (NEA et al., 1999) carried out for the European Commission in order to prepare the extension of the TEN-T networks to central and eastern Europe, the common assignment for freight and passenger transport has been applied.

The road network has been developed and refined to a great level of detail over time. Parts of it were applied and tested in several research projects. Currently, the network representation is based on the GISCO networks published by the EU. These networks have been enriched by data on link types and capacities, gradients and curvature collected in national and international research projects over a period of more then 10 years. Thus, the networks exceed the current state of the art, which makes the VACLAV model an appropriate tool for analysing infrastructure-related policies. The current representation of the European road network in the VACLAV model is presented by Fig. 2.

Speeds on road links (and therefore their contribution to user costs) depend on the traffic load situation. This relationship is represented by specific speed-flow and cost-flow functions for each link type. Link types are formally distinguished with

Fig. 2. The VACLAV Road Network for Europe.

respect to technical standards like the number of lanes and the terrain type (slope, bends). Beyond a characteristic load value ("capacity"), the increase of link flows persistently leads to a speed reduction and hence an increase of user costs, which is called "congestion effect." Speed flow functions are defined for cars and trucks and for every road type according to FGSV (1997). In order to take account of the congestion effect, the assignment is applied in an iterative way. The overall traffic demand is split and assigned step by step, where the link speeds are adjusted according to the new traffic loads after each step. As the assignment of traffic between the NUTS-III zones can only explain a fraction of total link loads, information of local traffic volumes is added to the networks. This is done by a procedure based on GIS-information on settlement densities developed by Schoch (2003).

Within the EU-funded projects SCENARIOS (INRETS, 1999) and SCENENES, (ME&P, 2001) databases with socio-economic data sets for the whole Europe on a regional scale have been developed. In the framework of the current impact analysis, this data is partly replaced by inputs from the ASTRA-T model, but parts of the data remain useful for the disaggregation process of the matrices provided by ASTRA-T to the regional level needed in VACLAV. The VACLAV database contains information on motorization, population by age classes, households, GVA and employment by Sectors, unemployment and GDP per capita in PPS.

The model is calibrated on the basis of manual or automatic counting post data, which are available for a wide range of countries. The flow maps generated by assigning the O–D matrices to the network finally indicate the sections where further corrections are needed and corresponding adjustments are necessary.

3.3. Macro-Economic Perspectives: The ASTRA Model

ASTRA, a system dynamics model, was developed under the 4th research framework programme of the European Commission in the project on Assessment of Transport Strategies (IWW et al., 2000) and is improved by the Institute for Economic Policy Research (IWW), Germany, and Trasporti e Territorio (TRT), Italy, in the TIPMAC project in the 5th framework programme (IWW & TRT, 2003). The objective of the ASTRA project was to develop a tool for the assessment of the long-term impacts of the European transport policy with respect to the economic, environmental and social implications. The technical challenge of ASTRA was to demonstrate that the applied system dynamics methodology is appropriate for such long-term policy assessments.

Since the successful completion of the ASTRA project at the end of the year 2000 the so-called ASTRA family of models has been developed. The first model ASTRA-I covers the Italian nation and enables an assessment of the Italian transport policy. The second model, ASTRA-E, is developed to assess the employment impacts of different technology policies for the European Union. The third model ASTRA-C incorporates an extension of the assessment capabilities with ASTRA towards a dynamic cost-benefit analysis framework. ASTRA-T, that is the model that is developed in the TIPMAC project, stands for a major increase of complexity of the model. To give an impression of this increase of complexity, the number of objects in the original ASTRA model and the ASTRA-T model can be taken for comparison. In the original ASTRA about 120.000 objects were included. In ASTRA-T currently nearly 5,000,000 objects are implemented. This enables for an analysis of transport impacts on a much more detailed level. For instance, the spatial differentiation has been extended from four macro regions in the European Union to 14 countries (Belgium plus Luxembourg form one region) plus a nested zoning system of four functional zones for each country that comprises four groups of NUTS II zones. In total there are 61 zones now covered in ASTRA-T, of which 17 are equal to one specific NUTS II zone while 44 zones are composed of more than one NUTS II zone in dependency of their population density and settlement patterns. The economic analysis is extended from 12 economic sectors to 25 economic sectors that are coherent with the EUROSTAT Input-Output table system that is based on the NACE-CLIO coding system. Currently, the model is extended to the new EU Member States and to Switzerland within the LOTSE project. For the current analysis of road tolling policies the ASTRA-T implementation is applied.

The model consists of eight sub-modules, which are illustrated in Fig. 3. The Population Module (POP) provides the size and age structure of population development for the 15 European countries, depending on fertility rates, death rates and migration. The Macroeconomics Module (MAC) provides the national economic framework, which imbeds the other modules. It incorporates neo-classical elements like production functions. Keynesian elements are considered like the dependency of investments on national income extended by some further influences on investments like exports or government debt. Elements of endogenous growth theory are incorporated like the implementation of endogenous technical progress as one important driver for the long-term economic development.

The Regional Economics Module (REM) calculates the generation and distribution of freight transport volume and passenger trips. The number of passenger trips is driven by employment situation, car-ownership development and number of people in different age classes. Freight transport is driven by the production value and the monetary Intra-European trade flows of 15 goods producing sectors. The Foreign Trade Module (FOT) is divided into two parts: trade between the EU15 member states and trade between the EU15 countries and the rest-of-the world, that is divided into 12 regions. Both models are differentiated into 25 economic sectors and relationships between country pairs.

Fig. 3. Structure of the ASTRA-T Model.

Major input of the Transport Module (TRA) constitutes the demand for passenger and freight transport that is provided by the REM in form of ODmatrices. Using transport cost and transport time matrices the transport module is performing the modal-split for five passenger modes and three freight modes. The Vehicle Fleet Module (VFT) is describing the vehicle fleet composition for all road modes. Vehicle fleets are differentiated into different age classes based on one-year-age cohorts and into different emission standard categories.

Major output of the TRA provided to the Environment Module (ENV) are the vehicle-kilometres-travelled (VKT) per mode and per distance band and traffic situation respectively. Based on these traffic flows and the information from the vehicle fleet model on the different vehicle fleet compositions and hence on the emission factors, the environmental module is calculating the emissions from transport. Finally, in the Welfare Measurement Module (WEM) major macroeconomic, environmental and social indicators can be compared and analysed.

The model outputs are presented by developments of selected key variables over time. In contrast to neo-classical general equilibrium models or standard CBA procedures, the development of output variables is considered even more important than their value at a particular point of time, as the intermediate performance may decide on the viability of a selected policy in terms of acceptability by certain actor groups.

4. RESULTS FOR EUROPE

The impact assessment investigates the development of transport-related and economic key indicators in three policy scenarios against a reference case between 1997 and 2015. The countries covered by the analysis comprise the EU member states plus Switzerland, Norway and Hungary. The central question that has to be dealt with is:

Which effects on the European transport sector and the national economies have to be expected when implementing the main basic forms of IRPS – NET and DAREA – on a European scale?

Questions dealing with the technical realisation and/or the institutional arrangements chosen together with the implementation of IRPS have not been part of this assessment. In fact, the assessment focused on the question, how the impacts on the transport flows of heavy goods vehicles on the European road network and on the national economies – with a special focus on the transport sectors and on the national budgets – vary for different scenarios of IRPS implementation within Europe. The results for the different scenarios should not be interpreted as a general forecast on economic development in general and the development of freight transport volumes on roads for the time horizon of 2015. Rather, the relative influence of different HGV tolling schemes on the economy and on freight transport flows is the central question.

4.1. Four Scenarios of Future European Road Tolling

The impact assessment investigated the development of transport-related and economic key indicators in three policy scenarios against a reference case between 1997 and 2015. In order not to mix these effects of road pricing solutions with the impacts of other policies only one common infrastructure scenario had been defined, reflecting the current plans of the of national governments on road and rail infrastructure construction as well as the projects of the Trans-European Transport Network (TEN-T) in the status of revision of October 2001. It has been assumed, that the funds for these projects will be raised independently of the IRPS scheme to be implemented in the different scenarios.

The Reference scenario describes the planned future development of inter-urban road pricing of HGVs, including the introduction of motorway toll systems in Germany and Austria as well as the further increase of the Swiss mileage-dependent heavy vehicle fee (LSVA), while all other countries are assumed to maintain their current systems. Against the reference case the following three policy scenarios have been analysed:

(1) NET: All countries except Switzerland introduce motorway tolls for HGVs as from 2005. Switzerland maintains its network-wide charging system.
(2) DAREA: All countries introduce HGV-tolls on all inter-urban roads.
(3) NET/DAREA: Motorway tolls for HGV in all countries except Switzerland, combined with a HGV-toll on all inter-urban roads in selected high-density areas.

The definition of the scenarios had to cover the following core aspects:

(a) The pricing schemes in place in the year 2015 and the path of their implementation

Notwithstanding the recent activities of implementing new inter-urban road pricing schemes in several European countries the situation of road pricing for HGV across Europe is still quite heterogeneous. Some countries have a long tradition of road user charges on motorways, while others are planning to introduce road pricing for HGV on motorways in the near future. And whereas Switzerland has already introduced a DAREA pricing scheme for HGV in the whole country, other countries do have no actual plans at all as regards introducing IRPS or introduce IRPS only for urban areas (U.K., Denmark). In order to assess the impacts of the introduction of different basic forms of IRPS for HGV on a European scale, it is assumed that the road pricing situation in the European countries will converge. This implies that in the policy scenarios at the forecast horizon 2015 almost all countries will have the same scheme implemented, either the scheme NET (Policy Scenario I) or the scheme DAREA (Policy Scenario II). In a third policy scenario, the general scheme NET is accompanied by the implementation of a DAREA scheme for certain high-density areas. Different implementation paths with the steps 2005 and 2010 reflect the differing initial maturity as regards the implementation of IRPS of the European countries. The policy scenarios are assessed against the Reference Scenario, which includes the implementation of a new IRPS only for Austria and Germany (NET) in the year 2005.

(b) The range and the inter-country variation of the toll levels

For each of the three basic scenarios, two different price levels (Low/High) have been defined. While Low-values are derived from rough estimates of infrastructure costs, High-values include additionally external costs. All price levels are based on average costs, which implies generally higher tolls in the DAREA cases due to the lower traffic density on the secondary road networks compared to motorways. Average price levels are 0.16 d/vkm (Low) and 0.50 d/vkm (High) in the NET scenario and 0.32 d/vkm (Low) and 0.76 d/vkm (High) in the DAREA scenario.

For the Reference Scenario, in some countries the differentiation of charge levels by road type has been further differentiated by single network sections (e.g. single tolled bridges and tunnels in France, Austria and the U.K. etc.). For all other scenarios a more uniform level of charges has been assumed for reasons of simplicity.

Vehicle weight classes or environmental standards do not differentiate charges. For all scenarios a standard 40 t truck-trailer in combination with EURO-5 emission standard has been assumed. The reason behind this simplifying assumption is, that in practice charges are assumed to meet a given budget goal. The toll setting authority is expected to react to any action of the hauliers to reduce the load of charges to be paid (e.g. by using more environmentally friendly or lighter vehicles) by adjusting the charge structure such, that the average charge level is maintained (Table 1).

(c) The use of toll revenues

In the way the road tolls are defined in the scenarios, they constitute a budgetindependent source of funds for financing transport infrastructure. The central question to solve was how to deal with the revenues created by the tolling schemes in the different scenarios. Although a part of the revenues has to be used for covering costs of maintenance and renewal of the respective national road infrastructure, the additional funds remaining allow in principle for a reduction of indirect taxes like VAT and fuel tax as well as of non-wage labour costs. Knowing that the reduction of indirect forms of taxation has impacts virtually on all sectors of the economy, but are not easy to quantify because of its very unspecific character, the approach chosen consisted of three elements:

| Country | | Policy Scenario I: NET
Only Motorways | | Policy Scenario II: DAREA
all Inter-Urban Roads | | | | | | | |-------------|-------------------------|------------------------------------------|--------------------|----------------------------------------------------|-------------------|--------------------|--|--|--|--| | | Introduction
in Year | Low
Values [d] | High
Values [d] | Introduction
in Year | Low
Values [d] | High
Values [d] | | | | | | Austria | 2005 | 0.26 | 0.48 | 2005 | 0.49 | 0.49 | | | | | | Belgium | 2005 | 0.16 | 0.60 | 2005 | 0.33 | 0.33 | | | | | | Denmark | 2010 | 0.16 | 0.42 | 2010 | 0.32 | 0.32 | | | | | | Finland | 2010 | 0.18 | 0.46 | 2010 | 0.35 | 0.35 | | | | | | France | 2005 | 0.16 | 0.59 | 2010 | 0.31 | 0.31 | | | | | | Germany | 2005 | 0.16 | 0.59 | 2005 | 0.34 | 0.34 | | | | | | Greece | 2010 | 0.16 | 0.39 | 2010 | 0.32 | 0.32 | | | | | | Hungary | 2005 | 0.16 | 0.46 | 2010 | 0.32 | 0.32 | | | | | | Ireland | 2010 | 0.20 | 0.46 | 2010 | 0.41 | 0.41 | | | | | | Italy | 2005 | 0.09 | 0.45 | 2010 | 0.31 | 0.31 | | | | | | Luxembourg | 2010 | 0.25 | 0.63 | 2005 | 0.49 | 0.49 | | | | | | Netherlands | 2005 | 0.14 | 0.47 | 2005 | 0.28 | 0.28 | | | | | | Norway | 2010 | 0.22 | 0.78 | 2010 | 0.45 | 0.45 | | | | | | Portugal | 2005 | 0.12 | 0.49 | 2010 | 0.33 | 0.33 | | | | | | Spain | 2005 | 0.12 | 0.42 | 2010 | 0.28 | 0.28 | | | | |

Table 1. Definition of Toll Levels of the NET and DAREA Scenarios.

(1) All revenues from road pricing are calculated and accounted by country.

Sweden 2010 0.18 0.53 2010 0.37 0.37 Switzerland 2001 0.75^a 1.10^a 2001 0.75 0.75 U.K. 2010 0.14 0.45 2010 0.28 0.28

(2) A part of the revenues created by road pricing is earmarked for maintenance and renewal of transport infrastructure.
(3) The remaining part of the revenues is used for the reduction of indirect taxation.

Because of the concept of using only one infrastructure scenario for all road tolling policies (in order to separate network infrastructure effects from pure tolling effects), the amount of network upgrading measures could not be a variable of the revenues collected in each scenario. This and other questions related to the use of transport pricing revenues are dealt with in the EU-funded research project REVENUE (ITS et al., 2004). In fact, it was decided to make the following incorporation of the treatment of revenues in the policy scenarios:

The revenues created in the Reference scenario at the toll level *Low (Infrastructure Costs Recovery)*from 2005 on are totally reinvested into the renewal of road infrastructure and defined as the Infrastructure Costs for Dynamic Renewal of Transport Infrastructure.

aAll roads.

All additional revenues caused by the policy scenarios and by the toll level High (Infrastructure Costs + External Costs) are used for the dynamic decrement of indirect taxes.

For the countries without road tolling in the Reference Scenario (U.K., Benelux, Scandinavia, Greece) it has been assumed, that the funds for the renewal of road infrastructure are generated by vehicle taxation. Therefore, the additional revenues from road tolling in the policy scenarios are used completely for dynamic tax reduction.

This strategy of combining the different IRP schemes in the road pricing scenarios with the toll levels and the treatment of revenues provides the following advantages:

The effects of road tolling are not mixed with network effects of new infrastructure when comparing the scenarios.
A clearly defined part of the tolling revenues is earmarked for transport infrastructure.
Additional revenues of the policy scenarios are used for a reduction of indirect taxes and thus influence the general economic performance.
For the countries with no IRPS system in place in the Reference scenario, the total effect of pricing roads for HGV and reducing indirect taxes are appraised by comparing the Policy Scenario I with the Reference scenario.
The same infrastructure scenario is used in all model runs.

In the following, the main properties as regards IRPS schemes of the proposed scenarios are described:

Reference scenario:

HGV road tolling is only included for roads that are already tolled in the year 2002 and for the countries, where the decision on the implementation of new tolling schemes has already been made. This is true for Austria and Germany. Although the governments of Germany and Austria planned to start the new tolling scheme until 2004, for modelling purposes it was assumed that tolling on the networks begins in 2005, which turned out to be quite realistic with respect to Germany while in Austria the system has indeed been put into operation in January 2004. For Hungary, road pricing on motorways was assumed to be equal on all motorways on a low level starting from the year 2005.

With the exception of Switzerland, there is no DAREA pricing scheme in this scenario. Switzerland itself, of course, retains the DAREA scheme of the LSVA introduced in 2001 in the Reference Scenario as it does in all policy scenarios.

Policy scenario I, <NET>:

It is the basic property of the policy scenario I that in the year 2015 all countries will have implemented the NET pricing system. This means that in every country a distance-depending road pricing system for HGV is introduced on the primary road network whereas the secondary road network is free of charge. An exception is again made for Switzerland, which certainly retains its DAREA tolling scheme. Since the initial situation of the countries as regards road pricing for HGV on motorways is different, there are three different paths of implementation for three groups of countries. The following steps of implementation over time are part of the definition of IRPS in policy scenario I:

(a) Countries that already have their motorway network tolled, will keep a NET scheme until 2015. Parts of the primary road network that are not part of the pricing scheme today will be included in the scheme starting in the year 2005. The countries that belong to this group are: Portugal, Spain, France, Italy and Hungary.
(b) The geographically more centrally located countries Austria, Germany as well as the Benelux states are introducing road user charges for HGV in the year 2005.
(c) All other countries, with the exception of Switzerland introduce a NET pricing scheme for HGV in the year 2010. This group of countries includes the U.K., Ireland, the Nordic Countries (Denmark, Sweden and Norway) and Greece.
(d) In the year 2015 all countries, with the exception of Switzerland that retains its DAREA scheme, therefore have a NET pricing scheme in operation for at least 5 years, which ensures that relevant impacts of the scheme are covered within in the time horizon of the model.

Policy scenario II, <DAREA>:

The basic property of the policy scenario II is the implementation of a full DAREA road pricing scheme for HGV by the year 2015 in all countries that are part of the modelling work. In order to take into account the differences in maturity towards such an advanced IRPS and in order to distillate the specific differences of impacts not only compared to the Reference Scenario but also to policy scenario I, again three groups of countries are identified with the following steps of implementation of IRPS over time in policy scenario II:

(e) In the year 2005 countries that actually do impose tolls on HGV on their motorways extend tolling to the whole primary road network. If the whole primary network is already tolled, the tolling is kept. The countries that belong to this group are: Portugal, Spain, France, Italy and Hungary.

(f) In the year 2005, Germany, Austria and the Benelux countries implement a DAREA pricing scheme. This means, all roads relevant for inter-urban traffic are priced for HGV.
(g) In the year 2010, all countries that will not have introduced a DAREA pricing scheme before implement a complete DAREA pricing scheme. This implies that the countries not having installed any kind of IRPS in 2005 (U.K., Ireland, the Nordic Countries Denmark, Sweden and Norway and Greece) will implement a full DAREA scheme in one step while the countries of group 1 upgrade their existing NET pricing scheme to DAREA.
(h) In the year 2015 all countries therefore have a fully operational DAREA pricing scheme implemented for at least five years, which ensures that relevant impacts of the scheme are covered within in the time horizon of the model.

Policy scenario III, <NET/DAREA>:

In the third policy scenario, the basic forms of IRPS – DAREA and NET – exist in parallel in the countries. While all countries are assumed to have the NET scheme in operation, for a subset of High Density Areas (HDA) out of the totality of NUTS2 regions a DAREA pricing scheme is implemented.

4.2. Impacts on Transport Flows

The direct network-effects caused by various road pricing policies have been estimated by applying the VACLAV network model, which is based on a European road network database. In most parts of Europe, including Central and Eastern European countries, the network is very detailed as it covers all major interurban roads relevant for long-distance traffic. Thus, the model allows the computation of various alternative travel routes between pairs of geographical zones depending on free-flow travel speeds, road occupancy, geographical characteristics and on road user charges.

4.2.1. Traffic Shifts

The impacts on transport behaviour and the resulting change of transport flows embrace route shifts as the most direct reaction of hauliers on price changes on different network parts or territories, modal split reactions as medium-term adaptations of forwarders to changing shipment costs and the reduction of total transport demand by the shipping industries. Behind these observable effects a number of strategic adaptation processes within the haulage business need to be taken into account as they can most likely not pass on all costs to their customers. Based on IWW (2001) and ECOPLAN (1999) it is estimated that loading factors may increase up to 5% in case of high area-wide charges, while pricing on motorways only would have a maximum effect of 1.6%. It is estimated, that the additional administrative costs of improved tour-planning can be compensated by making use of managerial processes for saving costs via financial measures and the co-operation or merging of companies. The resulting traffic shifts have been calculated by the network model for each scenario, taking into account network capacities and congestion effects on the entire European road network. They are presented in percentage changes of the kilometres travelled by HGVs on motorways and on the entire inter-urban network.

Reference scenario

In the Reference scenario 2015, where the Austrian toll system, the full stages of the Swiss mileage-dependent HGV-fee and the planned German motorway toll are put into operation, the model predicts a shift of transport flows from Germany (−4%) and Austria (−2%) to the Benelux countries (+2%), France (+1%) and Switzerland (+6%). On top of that, shifts away from the motorway to the secondary networks are predicted for Germany (−3.5%) and for Austria (−2%). The reactions in the policy scenarios now have to be interpreted against this initial situation.

Scenario I and II: NET and DAREA

In case low motorway charges are introduced in all western European countries, the development initiated by the reference case is partly turned back. As traffic flows now tend to avoid the Benelux networks by early shifting to Germany, the initial increase of demand in Switzerland is nearly completely shifted back to the Austrian and French Alpine crossings. A remarkable model output is, that in the case of tolling for high motorway charges, 2.5% of road transport demand is pushed out of the area of investigation, i.e. to Eastern Europe. The model reveals the expectation, that the introduction of high motorway tolls in countries, which had no pricing system in the reference case, causes considerable shifts of traffic demand towards the secondary networks. Introducing area-wide charges is able to gain some volume back to the motorways in case of existing motorway charges in the reference case. However, the introduction of area charges also may cause route shifts towards the secondary network as the hauliers simply try to save overall travel distance.

Traffic shifts away from the motorways range in general between approximately 5% in the NET-Low case and up to more than 25% in the NET-High case, depending on the reference toll system and on the density of the national road networks. More remarkable is the fact that also in the DAREA-scenarios the traffic volume reduction on the motorways is more pronounced than on the whole network.

The different levels of demand shift from the motorways to the secondary network in the single countries can not be explained by price changes alone. One of the important factors determining the model reaction is the density of the secondary road network (Fig. 4).

Scenario III: Combined NET/DAREA pricing:

The main purpose of scenario III was to find out whether route shifts from the motorways to the secondary networks can be avoided by introducing DAREA pricing only in designated high-density areas. When looking closer into the transport markets of densely populated regions, we can identify three main aspects affecting the hauliers' route choice decision:

Network density: In high-density areas the availability of secondary roads is much higher than in (rural) low-density areas. This provokes hauliers to leave the priced networks in order to save costs.
Traffic loads: Within dense areas the traffic loads on the secondary network are much higher than on inter-urban roads. Traffic loads are a decisive variable in the hauliers' decision function as they mean much worse travel conditions on non-priced network parts. Generally, high traffic volumes reduce the share of traffic leaving the priority network in order to avoid paying road tolls.
Transport distances: High-density regions are naturally characterised by a dominant share of local and regional freight traffic, while in inter-regional routes long-distance traffic plays the main role. Short-distance movements react much more sensitive to price changes than long-distance deliveries. This can be explained by the access time from and to the motorways, which is relatively more important for local and regional shipments than it is for long-distance hauls.

The model results indicate that the motorway traffic loads increase more or decrease less in high-density areas compared to the entire country. This is particularly remarkable as first, the hauliers' readiness to shift to secondary roads is higher in agglomeration areas, and second, the high DAREA-tolls should additionally cause a decrease of motorway traffic volumes. The desired effect of increasing motorway shares by levying DAREA-prices in densely populated areas appears most clearly in Germany, France and Austria, but also in the U.K. and in the Benelux countries. These countries are generally characterised by dense networks, which implies a high danger of route shifts under the NET pricing regime. In summary the thesis that combined NET and DAREA pricing helps reducing the negative side-effects of a pure motorway-based toll system in dense areas could be confirmed. This is particularly true for countries with dense road networks and when motorway charges are high.

4.2.2. Modal Split

The model approach used for DESIRE generally allows for the differentiation of the influence of the European road pricing policies in the scenarios NET and DAREA on the modal share of transport volumes according to three goods categories and four distance bands. Since for the short distance traffic of the Local Distance Band there is no real alternative to road transport, we only have to look at the longer distances. The classification of transport volumes into distance bands and goods categories are presented in Table 2.

The model results show that modal split is only slightly affected by the assessed changes in pricing policies. While in the NET-Low case the decrease of the market share of road haulage hardly exceeds 1% in any of the distance- and goods-classes, the DAREA-High scenario brings about a reduction between 1.6% for unitised goods and 9.4% for bulk goods. For the railways this means 4.1% increase in demand of bulk goods, +9.6% in general cargo and +5.8% in unitised goods. Modal shift is most significant in long distance transport (road:−12%, rail:+10%), while the effect is small in regional and local traffic.

Table 2. Differentiation of Results for Distance Bands and Goods Categories.

Distance Band	Token	Description
Long Distance Band Medium Distance Band Regional Distance Band Local Distance Band	LGD MED REG LOC	>700 km 150–700 km 50–150 km <50 km
Goods Category	Token	Description
Bulk	BLK	Mass products, crude materials. High volume-to-value-ratio: Solid fuels and ores Petroleum products Crude building materials Base chemicals
General Cargo	GCG	Semi-finished products. Medium volume-to-value-ratio: Cereals and agricultural products Metal products Fertilisers, plastic and other chemicals Large machinery
Unitised Goods	UNT	Consumer products. Low volume-to-value-ratio: Consumer food Conditioned food Cement and manufactured building materials Small machinery Miscellaneous manufactured articles

| EU15 Modal | Road | | | | Rail/Inland Waterways | | | Short-Sea Shipping | | | |-----------------------------------------------------------------------------|------------|------------|------------|------------|-----------------------|------------|------------|--------------------|------------|--| | Split 2015 | BLK
(%) | GCG
(%) | UNT
(%) | BLK
(%) | GCG
(%) | UNT
(%) | BLK
(%) | GCG
(%) | UNT
(%) | | | Modal split by goods category 2015 | | | | | | | | | | | | Reference | 25.8 | 47.5 | 72.7 | 25.9 | 17.5 | 13.6 | 48.3 | 35.1 | 13.7 | | | NET-Low | 25.6 | 47.2 | 72.6 | 26.0 | 17.6 | 13.7 | 48.4 | 35.3 | 13.8 | | | NET-High | 24.2 | 44.9 | 71.9 | 26.6 | 18.6 | 14.1 | 49.2 | 36.6 | 14.0 | | | DAREA-Low | 24.9 | 46.0 | 72.2 | 26.3 | 18.1 | 13.9 | 48.8 | 35.9 | 13.9 | | | DAREA-High | 23.4 | 43.7 | 71.5 | 26.9 | 19.1 | 14.4 | 49.7 | 37.2 | 14.1 | | | Relative change of modal share by goods category to Reference scenario 2015 | | | | | | | | | | | | NET-Low | −0.6 | −0.6 | −0.1 | 0.4 | 0.6 | 0.2 | 0.1 | 0.5 | 0.4 | | | NET-High | −6.3 | −5.5 | −1.1 | 2.6 | 6.4 | 3.7 | 1.9 | 4.2 | 2.0 | | | DAREA-Low | −3.5 | −3.0 | −0.6 | 1.6 | 3.5 | 2.0 | 1.0 | 2.4 | 1.2 | | | DAREA-High | −9.4 | −8.0 | −1.6 | 4.1 | 9.6 | 5.8 | 2.9 | 6.1 | 3.0 | |

Table 3. EU15 Modal Split 2015 per Goods Category.

Note: BLK = bulk goods, GCG = gemeral cargo, UNT = unitised goods, NET = network pricing, DAREA = distance-based area pricing.

Tables 3 and 4 present the modal shares by goods categories and by distance bands and their relative changes in the four scenarios considered.

A detailed analysis of the influence of increased variable transport costs by the policy scenarios on the modal shift reveals the following:

EU15 Modal Road Rail/Inland Waterways Short-Sea Shipping Split 2015 LGD MED REG LGD MED REG LGD MED (%) (%) (%) (%) (%) (%) (%) (%) Modal split by distance band 2015 Reference 28.2 66.1 88.9 11.1 26.2 11.1 60.8 7.7 NET-Low 27.9 66.0 88.6 11.1 26.3 11.4 60.9 7.7 NET-High 25.8 65.1 87.9 11.9 27.0 12.1 62.3 7.9 DAREA-Low 27.0 65.5 88.1 11.5 26.7 11.9 61.6 7.8 DAREA-High 24.9 64.5 87.1 12.2 27.5 12.9 62.9 8.0 Relative change of modal share by distance band to Reference scenario 2015 NET-Low −0.8 −0.2 −0.2 0.5 0.3 1.9 0.3 0.4 NET-High −8.3 −1.6 −1.1 7.1 3.0 8.8 2.6 3.0 DAREA-Low −4.2 −0.9 −0.8 3.5 1.8 6.7 1.3 2.2 DAREA-High −11.6 −2.5 −2.0 10.1 4.8 15.8 3.5 5.1

Table 4. EU15 Modal Split 2015 per Distance Band.

Note: LGD = long distance, MED = medium distance, REG = regional transport, NET = network pricing, DAREA = distance-based area pricing.

In the NET-Low scenario of modest pricing of motorways only, the loss of transport share of road is negligible for most countries. This holds, of course, especially for the countries that have already introduced motorway charges for HGV in the Reference scenario. The decrease of the modal share of road in the European average is about 0.2 percentage points (see Table 4). The most significant decrease of road transport shares can be observed for Finland, Ireland and Sweden with a relative decrease of the road modal share between 1.4 and 2.7%.

Significant effects can be observed especially for the high toll levels, both for the NET and the DAREA scenario. Depending on the goods category and on the distance band the modal share for rail and ship shows a relative increase between 3.0 and 15.8% (see Tables 3 and 4) For the DAREA-Low scenario the values are in between the NET-Low and the NET-High scenario.

There are significant differences of the modal shift between the goods categories and distance bands. In general, for unitised goods the road sector can keep most of its transport volume with a loss of maximal 1.6% in the DAREA-High scenario whereas for the lower value goods Bulk and General Cargo the alternative transport modes gain up to 4.1% for Bulk and 9.6% for General Cargo (Table 3).

Differentiating the change of the modal split of the policy scenarios in comparison with the Reference case to distance bands it occurs, that the most significant effects with a decrease of road transport share of 11.6% for the DAREA-High scenario and 8.3% for the NET-High scenario can be observed in the Long Distance Band whereas the changes for the Regional and the Medium Distance Band are similar and significantly lower (Table 4). We can conclude that the main reason for this difference is the relative strength of the alternative modes of transport for the long distances. Although one could guess that also another reason holds for this effect, namely the higher share of vehicle kilometre costs of total transport costs for the long distances, we will see below that this reasoning does not hold in this case. Quite on the contrary, the modelled increase in transport costs per ton kilometre is significantly higher for the regional and medium distances than for the long distances, which is basically caused by the lacking modal alternatives in the regional distances that do not allow for a significant modal shift on these transport relations together with differences in the other transport costs per vehiclekilometre.

4.2.3. Total Freight Transport Demand and Prices

The following driving factors are of special importance when investigating the joint impacts of the policy scenarios on the development of total transport volumes compared to the Reference case.

An increase of road transport costs leads to higher transport prices for road transport that can not be completely compensated by productivity increases in the haulage business through strategic adaptation processes (increasing load factors, reducing empty trips, mergers, optimisation of tour planning).

The resulting higher freight rates for road transport cause an increase in total transport costs because they can not be fully compensated by the shipping industry by modal shift to alternative modes or better logistics processes and must therefore be partly passed on to the consumers by an increase of product prices.

The increase of product prices is, according to the definition of the policy scenarios, compensated by a reduction of indirect taxes that lead to a contrarian increase of the disposable income and thus to a relative increase of total demand. The extent of this effect depends on the percentage of this income effect that is sapped by the increase of goods prices. Since we assumed that the introduction of new road pricing systems leads to adaptation processes in the transport sector and an increase of its productivity, the net effect of the policies on total demand is positive.

A change of total demand leads to changes of total transport volumes in the same direction compared to the Reference scenario. This statement does not necessarily exclude the possibility of a decoupling of the development of transport volumes and GDP over time, since such a decoupling would not be caused by the introduction of road user charges for HGV.

The decreasing effect of road tolls on freight transport volumes depends on the relative increase of total transport costs and varies across distances and goods categories.

The model results show that the total amount of transported tons is approximately constant in all scenarios, but we can observe a decrease of freight traffic [t] with the transport distance especially for the high toll levels. This result is quite intuitive because an increase of variable transport costs is an incentive for the shift of freight transport volumes from long distances towards shorter distances.

As regards the total freight performance [tkm] we observe an increase for the long distances of up to 1.3% for the DAREA-High case and a decrease for the regional distances of up to −0.6%. The total increase of ton-kilometres for this scenario compared with the Reference scenario is with 0.5% significantly higher than the increase of total tons transported. The fact that the transport performance [tkm] increases for the policy scenarios must be interpreted as an increase of the average transport distance, by the change of the matrix of transport-flows. Another important aspect is the fact that, for the long distances, the quality of the alternative transport modes rail, inland waterways and short-sea shipping is better than for shorter distances, for which there is no real alternative to road transport.

Table 5 shows the ASTRA-T results changes in transport volumes by distance bands and by goods categories 2015.

			Change of Freight Volumes [t] per Distance Band to REF		Change of Freight Volumes [t] per Goods Category to REF
	LOC (%)	REG (%)	MED (%)	LGD (%)	BLK (%)	GCG (%)	UNT (%)
NET-Low	0.1	−0.1	−0.1	0.0	0.0	0.0	0.0
NET-High	0.5	−0.1	−0.3	−0.8	0.1	0.0	0.2
DAREA-Low	0.3	−0.2	−0.2	−0.3	0.0	0.0	0.1
DAREA-High	0.7	−0.3	−0.5	−0.9	0.1	0.0	0.2

Table 5. EU15: Change of Freight Transport [t] to Reference Scenario.

Note: BLK = bulk goods, GCG = general cargo, UNT = unitised goods, LGD = long distance, MED = medium distance, REG = regional transport, NET = network pricing, DAREA = distance-based area pricing.

The effect, that the transport performance [tkm] per goods category increase for Bulk Goods and decrease for Unitised Goods is also in line with the development of transport costs per tkm together with the increase of total transport volume [t] and is caused by the lower volume-to-value ratio of high-priced unitised goods compared to bulk goods.

4.2.4. Freight Transport Costs and Strategic Adaptation in the Haulage Business

While environmental advocates usually expect that the major reaction of pricing road transport is a shift from the road to environmentally more friendly modes, in reality the main reaction is an internal adjustment within the road freight transport sector. However, the manifold reactions of hauliers and shippers to reduce transport costs and by this to partly compensate for the user charges cannot be exactly forecasted.

As mentioned earlier, there are different forms of rationalisation of transport operations and adjustment of the vehicle fleets. These forms include the optimisation of round trip tours, the formation of vehicle pools or the co-operation between transport haulage firms, as well as possibly some illegal but not severely penalised behaviour like overloading of vehicles or the violation of regulations and social dumping practices.

4.2.5. Tour Planning and Loading Factor Improvement

Productivity increases of the hauliers forced by raising operating costs are investigated in IWW (2001) by observing current developments in Switzerland and Germany and by consulting experts in the transport business (ECOPLAN, 1999). The study concludes that through the merging of firms, through the establishment of vehicle pools and through improved tour planning, increases in vehicle load rates and reductions in empty running can be achieved. The additional costs of those measures for the hauliers are net out by productivity increases in the management sector. However, the traffic-relevant productivity increases caused by road tolling are likely to be restricted as increasing competition in the future will force the hauliers anyway to make a better use of the available vehicle stock and labour force. The overall increase in load factors is estimated to be no more than 0.7% for forwarder goods and 0.2% for bulk goods when charges of 0.13 Euro/km are raised solely on motorways. When raising the toll level to 0.20 Euro/km and levying it on all roads, increases of 1.8% (forwarder goods) and 0.7% (bulk goods) are estimated.

These estimates are based on two decisive assumptions:

(1) The decision base for hauliers is the increase in direct vehicle operating costs rather than the total company expenses allocated per vehicle kilometre. This means, that overhead costs are to be excluded from the vehicle operating costs. The Calculation Handbook for Hauliers (BGL, 2002) reports, that 60% of operating costs per time unit are overheads. Considering the share of the value of time in the hauliers cost structure, the costs have to be increased by 70% in order to get the relevant cost level for strategic decisions.
(2) The elasticity of reactions on cost changes is far from being constant. It is estimated that real behavioural changes can be expected only when cost increases reach a level of 15–20%. This holds for internal decisions in the transport sector as well as for decisions taken by shippers and forwarders.

The average price elasticity is estimated to vary considerably with the type of goods transported and the transport distance. On average a price elasticity of the gross load factor of 0.04 seems to be realistic when estimating across all market segments. The gross load factor is defined as the total transport performance (tkm) divided by the total HGV mileage, including empty hauls. Changes in the gross load factor are directly translated in to changes of the vehicle mileage.

The results of the load factor indicate that in the low variant of the NET scenario no significant influence of the road tolls on vehicle load factors can be expected. Only for those countries not having a road pricing system introduced in the REF case, increases of the average vehicle load factor of around 0.2% (Luxembourg: 0.8%) are calculated. However, they are dominated by the high transport volumes of those countries applying road charging already in the Reference case (Germany, France, Italy, Spain, etc.). Due to the general increase of charge levels in the NET-High scenario an increase of the vehicle loading factors and thus a decrease of vehicle kilometres travelled is calculated for all countries. However, with values between 0.1% and 3% these effects are very moderate. The European average results as 1.6%. Higher charge levels and the pricing of all roads in the case of the DAREA scenarios result in higher loading factor increases than in the NET scenarios. While the effect estimated in the DAREA-low case (1.3%) is comparable to the NET-High case (1.6%), the DAREA-High case brings about a more pronounced increase in load factors (4.3%).

The model estimates show that the effect of road user charges for HGVs on vehicle load factors and thus on traffic volumes is weaker than it is generally assumed. This mainly stems from fixed package sizes required by production industries, the generally not extendable carrying capacity of special vehicles (e.g. for liquid fuels or chemicals) and from customers' requirements concerning the frequency of deliveries. On the other hand, the consequences of merging activities induced by road pricing for particular companies might be significant. Big haulage companies will be able to adopt to the new pricing environment as their customer stock will be big enough to alter the use of the existing fleet and the regular purchase of new lorries further allows to adopt the fleet in order to lower costs. Small companies will not have these possibilities. The consequence is that small companies will feel the pressure to merge in order to make use of economies of scale and scope and to hold their ground in the toll-driven increase of competition intensity. This development could be observed in Switzerland after the introduction of the LSVA in 2001.

After the discussion about the possibilities of the haulage business for a strategic adaptation according to the new road tolling schemes we are now dealing with the results of the DESIRE model approach for the change of total freight transport costs of the policy scenarios in comparison with the Reference case. It has to be mentioned that in this approach neither the possibilities for a change of location for production and warehouses nor toll-level-specific load factor improvements are included. Figures 5 and 6 present the relative cost increase by goods category, distance band and policy scenario in 2015 compared to the Reference scenario.

Depending on the scenario the cost increases reported by the model are of high significance and average up to 48.1% for Unitised Goods and 105.2% for the regional distances in the DAREA-High scenario. Compared with the changes in the other policy scenarios the price increases of the NET-Low scenario are negligible. For Unitised Goods the increase of transport costs is approximately three times higher than for Bulk Goods and 50% higher than for General Cargo. This can be explained by the differences of the load factors between the goods categories which are much lower for the high-priced but "light-weighted" Unitised Goods than for mass products.

Looking at the cost developments for different transport distances the results of the model seem to be quite counter-intuitive at the first sight as the boost of transport costs per tonne-kilometre is much higher for short distances than for long distances. The main reason for this result is the fact that, besides differences

Fig. 5. Policy Influence on EU15 Freight Transport Costs per Distance Band 2015.

in the non-tolling costs per vehicle-kilometre, there are no or only few transport alternatives to road transport for local and regional transport distances. Thus, the shipping industry can adapt to the road pricing scenarios by shifting parts of their transport volume to other modes and thus restrict the growth of their average transport prices. Against this background we can conclude that the introduction of road pricing, especially in the high toll levels and in the DAREA-schemes that do not allow for reducing transport costs by avoiding the motorways, create a pressure for reducing costs both for long distance and short distance transports. As long as the pricing scheme is not extended to HGV<12t the regional and local

Fig. 6. Policy Influence on EU15 Freight Transport Costs per Goods Category 2015.

road transport, that has few possibilities of modal shift, can react by using smaller vehicles. With an extension of road pricing to trucks <12t the tested IRPS can serve as an effective incentive against an individualisation of transport processes by use of modern communication techniques.

The high cost pressure for short transport distances also promotes mergers of local haulage companies in order to optimise route planning processes by the use of economies of scale and scope which mean improved logistical resources. On the other hand, for long distances the share of transport costs on total costs is higher than for short distances, which means that the same percentage increase of road transport costs causes a relatively higher percentage increase of total costs. Thus, the cost pressure felt by haulage and forwarder companies on the Long Distance Band is stronger than it may seem when looking at the differences of road-pricinginduced cost increases per ton-kilometre between long and short distance bands.

4.3. Impact on National Economies

Influence on national GDP values

It can be estimated that between 20% and 40% of the costs imposed on the hauliers by additional HGV tolls can be absorbed within the road transport sector. Some of these effects are of negative nature for the society (e.g. traffic shifts to the secondary network), but most of the effects are positive, as they reduce vehicle kilometres driven, and thus reduce environmental loads, noise disturbance, accidents and congestion. The remaining 60–80% of costs need to be carried by the forwarders, i.e. by the shipping industries or by the consumers by increased product prices. Forwarders might react by changing modal choice or by re-structuring purchase, production and distribution mechanisms. The cost effectiveness of these measures can not be estimated with the present tools and thus we need to assume that their net effect is not more than a 5–10% reduction of the extra costs caused by the new tolling structures.

Using input-output techniques it could be shown that for most economic sectors the share of transportation of the total production costs ranges between 2% and 5%. It follows, that the order of magnitude of increases in production costs is well below 1% for the NET-Low scenario, but ranges up to 5% in the other scenarios, depending on transport distance and type of good transported.

These values vary by country and by economic sector, a distinction that is decisive for discussing the influence of road pricing on companies' cost structure, since the situation is different for bulk goods markets and markets for higher value products. Bulk goods in general are more problematic than high value goods because of the following reasons:

Bulk goods are usually shipped by special vehicles, which do not provide as much degree of freedom in altering load factors than unitised goods do. Bulk goods are not distributed within networks of many clients, but more frequently between single origin and destination locations. This limits the possibility to reduce empty headings. Bulk goods are generally more transport intensive than higher value goods.

This means the possibilities for a compensation of additional costs imposed by road pricing is limited and the remaining cost increase is more relevant for bulk goods than for higher value goods. Estimating the intensity of transport costs by 15% of production costs and the potential to compensate for cost increases within the hauliers' market by only 10% leads to an increase of production costs up to 1% in the NET-Low case and up to 10% under the DAREA-High pricing regime.

According to the scenario description the following approach for the use of the additional toll revenues and its consequences on aggregated real demand had been implemented in the model.

All additional revenues created by road-pricing in the policy scenarios compared to the Reference case are used for the reduction of income tax. Thus, because of a raise of the nominal income of the households the aggregated demand increases and induces positive effects on production and thus on GDP and employment.

Since the relevant factor for the economy is not the change of nominal but of the real demand caused by the policies, the reactions of consumer prices to the increased transport costs was of importance here. Results of econometrical studies indicate a price elasticity of 70% of consumer prices, which means that 70% of the additional income of the households is eaten by product price increases. Thus, the remaining 30% of the respective additional national toll revenues of the policy scenarios are available as an increase of real demand (Fig. 7).

Figures 8 and 9 take a closer look at the impacts of the NET-High and the DAREA-High tolling scenarios on the GDP development per country.

The model results for the development of the GDP draw the following picture. Effects on the GDP can be observed for the high toll level scenarios with a relative increase of 0.16% (NET-High) and 0.14% (DAREA-High) compared to the Reference case. When applying the lower toll levels the positive effects of internal adaptation processes of the transport sector and of the increased aggregated demand approximately equal the effects of increased transport prices.

When introducing new tolling schemes we can generally observe a slump in the economic performance that varies across countries depending on the pricing system installed, the year of introduction, the year of introduction of the scheme in the neighbour countries, the old scheme that has been already in place (NET or no IRPS for HGV), the respective economic structure of a country and interaction processes between the situation of the countries in general.

Fig. 7. Influence of the Policy Scenarios on GDP for the EU15 Countries.

This slump is generally followed by a new trend that varies between country and implemented scheme that causes a compensation process caused by adaptation processes in the transport and the transport-intensive sectors that in the most cases lead to a positive impact on the general economic performance until the year 2015.

Despite the positive effect of road pricing on the average economic performance, the relevance of the economic slump after the introduction of new pricing schemes,

Fig. 8. Policy Influence on GDP per Country for the NET-High Scenario Compared to Reference Scenario (%).

Fig. 9. Policy Influence on GDP per Country for the DAREA-High Scenario Compared to Reference Scenario (%).

should not be underestimated and could well be capable of reducing the political feasibility of new IRPS significantly, especially in periods of general weak economic performance. Countries that decide for a unilateral introduction of new tolling schemes with high tariff-levels risk a worse reduction of economic growth compared to a situation of coordinated implementation paths. This indicates the necessity for a coordination of national road pricing policies that may be hard to accomplish.

The model approach implemented does not take into account the costs for the technical and institutional implementation of the tolling schemes as well as license fees and the costs for enforcement. Furthermore, it can be expected that, especially in the case of technologically advanced schemes the system may need some time to be fully operable and reliable and their may be risks of malfunction that have to be carefully calculated. On the other hand, new HGV-road pricing schemes could bring about considerable revenue to the governments. For all EU member states, the introduction of the DAREA-High scheme would result in additional gross toll revenues of 560 billion d in 15 years (ca. 37 billion d/year) compared to the Reference scenario.

We can conclude that in general and in the long run, we can expect positive effects on economic performance from the implementation of IRPS, but that there are also certain economic risks that have to be carefully assessed. Since after all – not taking into account the increased stability of financial resources for transport infrastructure investments when shifting these resources from general budget to earmarked road user charges – the economic benefit mainly relies on efficiency gains in the transport sector, significant effects can only be expected when the toll-level is above values currently in discussion in Germany or Austria and above the actual tariffs in mature countries like France, Spain, Italy and Portugal. To avoid unwanted impacts on national economies a coordinated European path of implementation seems to be favourable. It may need years to compensate for the costs of restructuring in the transport sector and the costs for implementation and enforcement of new systems. Also, the reliability of the new system is obviously of high importance.

In addition to the discussion of the impacts of the DESIRE road pricing scenarios on the general economic performance the additional revenues from road-pricing that can be expected in the single countries compared to the Reference scenario have been calculated. Naturally, the amount of additional revenues depends not only on the total transport volume of a country but also on the situation in the Reference case. For countries that already have a tolling scheme for HGV in place in the Reference scenario (Germany, Austria, France, Spain, Portugal, Italy) the additional revenues are certainly relatively small compared with countries that start without any inter-urban road user charges at all (Fig. 10).

4.3.1. Influence on Sectoral Production Values

In addition to the discussion of the effects of the DESIRE policy scenarios on the general economic performance that has concentrated on the GDP as the main indicator in the previous chapter, we present next the model results for the effects

Fig. 10. EU15: Additional Gross Toll Revenues of the Policy Scenarios Compared to Reference Scenario.

| | EU15 Production Value for Most Influenced Sectors | | | | | | | | | |------------------------|---------------------------------------------------|--------------|---------------|----------------|--|--|--|--|--| | | NET-Low (%) | NET-High (%) | DAREA-Low (%) | DAREA-High (%) | | | | | | | Construction | −0.13 | 0.23 | 0.03 | 0.35 | | | | | | | Trade | −0.02 | 0.17 | 0.07 | 0.24 | | | | | | | Catering | 0.00 | 0.34 | 0.16 | 0.46 | | | | | | | Food | −0.13 | 0.23 | 0.03 | 0.35 | | | | | | | Other market services | −0.02 | 0.65 | 0.30 | 0.91 | | | | | | | Textiles | 0.00 | 0.34 | 0.16 | 0.46 | | | | | | | Communication | −0.07 | 0.02 | −0.03 | 0.05 | | | | | | | Energy | −0.04 | −0.29 | −0.22 | −0.54 | | | | | | | Vehicles | −0.02 | 0.65 | 0.30 | 0.91 | | | | | | | Transport air maritime | −0.12 | 0.41 | 0.12 | 0.60 | | | | | | | Transport Inland | 0.03 | 0.34 | 0.18 | 0.49 | | | | | |

Table 6. EU15: Policy Influence on Production Value per Sector Compared to Reference Scenario (2015).

on the sectoral production value of the EU15 countries. These results have been obtained by the implementation of national input-output tables for 25 economic sectors in the ASTRA-T model and are presented for a selection of most influenced and most interesting sectors with respect to the scope of the model. Table 6 presents the change in production values for the most affected economic sectors.

When focusing on the most influenced economic sectors we observe the highest increases in the sector "construction," especially for Italy, followed by Greece, France and Finland. Especially for Italy and Greece the effect on this sector is of high importance for the effect on the total production value.

Of special interest is a view on the sector "vehicles." In contrast to the positive effects on the most sectors, we observe here for the majority of the countries negative effects in the year 2015. This can be explained by the reduction of road transport demand because of higher transport costs and shifts to alternative modes of transport for the long distance relations. But we should take into account that for the years following the introduction of new schemes, a differentiation of the tarifflevels according to weight categories and emission classes induces an anticipated renewal of the vehicle fleets and thus a temporary increase of demand in this sector.

Taken into account that the sectors presented in Table 7 are the most influenced ones we observe relatively high effects on sectors with a close relationship to final consumption like "catering," "food," "trade" and "market services" whereas mass products as metals, plastics and chemicals show only less pronounced reactions. We can trace back this effect to the implementation of the revenue use in the model by the reduction of income taxes and thus to an increase of the disposable income of the households.

D A R E A ig h H	Co io tru t ns c n ( % )	Tr de a ( % )	Ca in te r g ( % )	Fo d o ( % )	O he t r ke M t ar Se ic ( % ) rv es	Co ic io t m m un a n ( % )	En er g y ( % )	Ve h ic le s ( % )	Tr t an sp or ir A i im ( % ) M t ar e	Tr t an sp or la d In n ( % )
A U T	0. 5 4	0. 7 4	1. 1 3	0. 7 5	0. 7 1	0. 9 1	0. 7 8	2. 3 4 −	0. 1 3	0. 3 5
B L X	1. 3 6	0. 8 0	1. 0 1	0. 4 7	0. 7 8	0. 4 4	0. 4 3	0. 0 3	0. 0 2	0. 0 6
D N K	0. 6 6	0. 3 8	0. 4 2	0. 1 6	0. 3 5	0. 2 4	0. 2 1	1. 8 0 −	0. 0 7	0. 1 3
E S P	1. 6 8	0. 7 5	0. 8 2	0. 6 1	0. 7 7	0. 5 9	0. 5 4	0. 9 5 −	0. 3 0	0. 2 1
F I N	1. 8 8	1. 2 7	1. 4 7	1. 0 5	1. 1 9	0. 8 8	0. 8 8	0. 5 3 −	0. 2 1	0. 4 2
F R A	1. 9 4	1. 3 7	0. 9 6	0. 8 5	0. 5 3	0. 3 7	2. 0 7	0. 0 8 −	0. 0 2	0. 1 0 −
G B R	0. 2 5	0. 9 1	1. 0 2	0. 8 6	0. 6 5	0. 5 9	0. 3 6	0. 4 1 −	0. 1 1	0. 2 0
G E R	0. 4 3	0. 8 1	0. 7 6	0. 8 9	0. 6 8	0. 8 1	0. 4 0	0. 3 4 −	0. 1 8 −	0. 1 3 −
G R C	2. 3 0	0. 7 8	0. 5 8	0. 6 1	0. 5 2	0. 5 6	0. 4 5	7. 4 4 −	0. 1 5	0. 2 1
I R L	1. 4 5	0. 7 8	0. 7 8	0. 3 4	0. 9 6	0. 6 2	0. 2 3	0. 0 2	0. 4 2	0. 1 9
I T A	2. 9 7	1. 0 8	1. 1 0	0. 8 7	0. 9 6	0. 8 5	0. 3 7	1. 3 3 −	0. 3 4	0. 4 4
N L D	6 0. 3	0. 3 2	5 0. 2	0. 2 2	0. 4 2	0. 3 8	0. 4 3	6 1. 8 −	0. 0 1	0. 1 1
P R T	5 1. 4	6 0. 4	5 0. 2	5 0. 4	5 0. 3	5 0. 1	0. 4 7	5 0. 0 −	0. 2 1	6 0. 2
S W E	0. 5 9	0. 8 3	0. 5 7	0. 3 7	0. 8 3	0. 6 1	1. 0 7	0. 0 0	0. 4 9	0. 6 0

Finally we have to state that the effects on the transport sector are relatively small. While for the sector "transport inland" we can easily understand that increased road transport costs are not likely to have a increasing effect on this sector (which is nevertheless positive for most countries since we have an increase of total transport volumes) the small reaction pattern for the sector "transport, air maritime" is quite surprising.

4.4. Impacts on Environment and Road Safety

The adaptation processes in the haulage business and the decisions of the shipping industries triggered by the implementation of new road pricing schemes also imply changes of environmental emissions of the transport sector and of road safety indicators. Here, we shortly present the expected influence of the policy scenarios on the basic transport-caused emissions of the air pollutant NOx and the greenhouse gas CO2 as well as on the number of road fatalities and the structure of the truck vehicle fleets according to the EURO-standard emission categories.

Not surprisingly, the reduction of total road transport performance [tkm] helps reducing emissions of air pollutants and greenhouse gases. The model results indicate a reduction of CO2 emissions between 0.5% for the DAREA-Low scenario and 1.4% for the DAREA-High scenario and a reduction of NOx emissions between 0.3% for the DAREA-Low scenario and almost 0.8% for the DAREA-High scenario. It should be noticed, that this reduction takes place against the background of an increase of total freight transport performance (all modes). In addition, despite certain shifts of transport flows to the less safe secondary road network, the model results indicate a decrease of road fatalities of up to 0.5% in the DAREA-High scenario in comparison to the Reference case, caused by a reduction of total vehicle-kilometres of HGV performed on the European road network (Table 8).

On the development of the emission indicators over time we see that their relative decrease is closely connected with the date of introduction of the new

| | Change to Reference Scenario 2015 | | | | | | | | | |------------|-----------------------------------|--------------|---------------|----------------|--|--|--|--|--| | | NET-Low (%) | NET-High (%) | DAREA-Low (%) | DAREA-High (%) | | | | | | | CO2 | −0.08 | −0.90 | −0.49 | −1.35 | | | | | | | NOX | −0.04 | −0.54 | −0.27 | −0.76 | | | | | | | Fatalities | −0.02 | −0.35 | −0.17 | −0.53 | | | | | |

Table 8. Policy Influence on Emissions and Road Fatalities.

Fig. 11. EU15: Policy Influence on CO2 Emissions.

road pricing schemes. In reality, we can assume a smoother development of the emissions that takes place already before a new tolling-scheme is in operation because the anticipation of proposed tolling schemes with significant tariff-levels and a differentiation of tariffs according to emission categories lead to an early restructuring of the vehicle fleets by the road transport companies. Since in the DESIRE scenarios the differentiation of toll levels to emission categories was not included explicitly, this effect is not included here (Figs 11 and 12).

Fig. 12. EU15: Policy Influence on Yearly Transport Fatalities.

IWW (2001) estimates for Germany, that diversifying average motorway tolls of 0.13 Euro/km for HGVs by EURO emission standards would decrease the share of vehicles equipped with Euro-5 standard by about 8% between 2003 and 2010, while the effect of pricing the entire network with 0.20 Euro/km is roughly twice as high. However, for the assessment of environmental effects the exposure of inhabitants by air pollutants must be considered rather than the simple atmospheric emissions of vehicles. Air pollutant emissions at motorways are usually less harmful than emissions on secondary roads, which are usually built closer to settlements. This implies that traffic shifts caused by pricing motorways only reduce the environmental advantages of differentiating charges according to emission standards. This negative impact of the NET scenarios does also hold true for indicators of traffic safety, as accident rates on secondary roads are usually clearly higher than on motorways.

To measure the net impact of these effects, IWW (2001) has used a monetary indicator, which is based on the estimates of marginal social costs of air pollution, greenhouse gas emission and accidents for road and rail traffic. The study has concluded that the net effect on environmental performance and traffic safety is very close to zero under NET-Low conditions. The high route shifts in the NET-High cases imply a significant increase of social costs. In contrast, the DAREA scenarios are considered by the study to reduce external social costs on the roads between 4% (0.20 Euro/km) and 6% (0.40 Euro/km).

5. CONCLUSIONS

The analyses indicate that each type of reaction on the introduction of road pricing schemes entails positive effects but also bears the risk of undesirable outcomes. While on the one hand hauliers are forced to make better use of their vehicle fleet and thus increase loading factors and reduce empty headings, the negative side of the emerging cost pressure are shifts of traffic demand to the more sensitive secondary networks and possible bankruptcies of haulage companies. Concerning the economic development in the medium- to long-term we observe positive impulses from the cost-driven increases in transport efficiency, but in the short term the extraction of financial resources from the economy will in some cases cause a drop of economic performance indicators.

From an economic and environmental perspective and looking into the long term, the high price scenarios are superior to the low price scenarios for Europe in total. However, the acceptability of such pricing regimes might not be very high. Among the high price scenarios the NET variant is to be preferred, because it lets some freedom to local and regional traffic, which is usually not as adaptive as long distance haulage. Regarding the low price scenarios we found, that the NET variant does hardly cause any change to the development of economic performance indicators compared to the reference case.

Considering the environment we state, that simply raising prices for road transport has only the little effect of deeming the growth of traffic volume on the road and to shift some demand to the more environmentally friendly modes. However, only a differentiation of prices according to the environmental friendliness of vehicles will create the incentive to restructure the vehicle fleet towards cleaner technology. Though this question was not within the focus of the DESIRE project, it is emphasised here that the combination of tough regulations on emission standards of new vehicles and financial incentives to modernise vehicle fleets form a powerful instrument to constantly reduce emissions and to increase the quality of life for today's and forthcoming generations.

REFERENCES

Doll, C., Kleist, L., Schade, W., & Schoch, M. (2003). DESIRE (Designs for Interurban Road Pricing Schemes in Europe). Deliverable 4: Impact Analysis and Assessment. Project funded by the European Community under the 'Competitive and Sustainable Growth' Programme. Lisbon.
ECOPLAN (1999). Auswirkungen des bilateralen Landverkehrsabkommens. Bern.
FGSV (1997) Empfehlungen f ¨ur Wirtschaftlichkeitsstudien an Strassen (EWS). Forschungsgesellschaft fur Strassen – und Verkehrswesen e. V. Bad Godesberg. ¨
INRETS (1999). Socio-economic external developments, spatial dynamics and their relations to transport. Deliverable C1 of the project SCENARIOS, financed by the 4th framework programme of the European Commission. Contract no. ST-96-AM.104. Project co-ordinator, Lyon: INRETS.
ITS/TIS.PT/KULeuven/ECOPLAN (2004). State of the art and conceptual background. Deliverable 1 of the EU-funded research project REVENUE (Revenue Use from Transport Pricing), 5th RDT framework programme. Project co-ordinator, Rome, Italy: ISIS.
IWW (2001). *Anforderungen an die Umweltorientierte Schwerverkehrsabgabe f ¨ur den Stra*ßeng ¨uterverkehr. Project No. 200-96-130. Berlin: German Environment Agency.
IWW & MKmetric (1999). Final report on passenger transport of the research project "Transport Forecasting Passenger and Goods to 2020" financed by the 4th framework programme of the European Commission. Project co-ordinator, Delft: TNO-Inro.
IWW & TRT (2003). ASTRA model update: Description of the ASTRA-T model. Deliverable D3 of the project TIPMAC (Transport infrastructure and policy: A macroeconomic analysis for the EU) financed by the 5th framework programme of the European Commission. Project co-ordinator, ´ Cambridge, UK: Cambridge Econometrics.
IWW, TRT, ME&P, & CEBR (2000). ASTRA methodology. ASTRA Deliverable 4 plus Annex, IWW, TRT, ME&P, CEBR, Project on Behalf of the European Commission DGTREN 4th FP, Karlsruhe. http://www.iww.uni-karlsruhe.de/ASTRA.
ME&P (2001). Final report of the project "SCENES European Traffic Scenarios" funded by the 4th Framework Programme of the European Commission. Contract No. ST-97-RS.227. Project co-ordinator, UK: ME&P.
NEA, IWW, & INRETS (1999). Final report of the study "Traffic Forecast on the Ten Pan-European Corridors of Helsinki". NEA-IWW-INRETS, project funded by the European Commission.
NEA, IWW, MKmetric, NESTEAR, & others (2003). Description of the base year 2000 and forecasts 2020. Deliverable D3 of the project TEN-STAC (Scenarios, Traffic Forecasts and Analysis of Corridors on the Trans-European Network) funded by the 5th framework program of the European Commission. Contract-No. ETU/B5-7000A-SI2.346797. Project co-ordinator, Rijswijk: NEA.
Schoch (2003). Verwendung feinraeumiger geographischer Informationen in aggregierten Verkehrsprognosen. Karlsruhe Papers on Economic Policy Research No. 15. Nomos Verlagsgesellschaft Baden-Baden.

10. THE EVOLUTION OF THE EU DIRECTIVE ON ROAD CHARGES AND TOLLS FOR TRUCKS

Jose M. Viegas ´

The proposal for amendment of EU Directive 1999/62 presented by the European Commission in July 2003 to the European Parliament and to the European Council (European Commission, 2003) had the main purpose of responding to various points on which member states did not feel in agreement with the terms of that Directive, by providing a revised framework for road charges and tolls that would respond to the requests of internalisation of external costs in a unified way, thus avoiding fragmentation of the EU in this field of charges for road transport activity.

The submission of this proposal was welcomed by the EU Council in December 2003, which also urged that agreement be reached in time for approval in March 2004.

This is an area which requires co-decision by these two bodies, which means that they may have independent reviews and amendments of the version proposed by the Commission in two consecutive readings. If agreement is not reached in this way, a "conciliation committee" is created, with equal numbers of Council and Parliament members, whose works are followed by Commission representatives. After agreement in this committee, the agreed text must then receive formal approval by both parties.

Whereas the current Directive supported the application of charges (time-based prices, i.e. vignettes) as well as of tolls (distance-based prices), the proposed amendment is much more dedicated to the application of tolls, as they are the

instrument favoured by most countries, as they bring increased revenue and are more closely related to the actual usage of the road, thus being more fair and efficient.

This proposal targets the set of vehicles and roads that are relevant for the internal market, thus defining the lower weight limit at 3.5 ton (instead of 12 ton in the 1999 version) and the main itineraries, included in the Trans-European Road Network (TEN), with the possibility for extension to parallel roads where there would be a danger of detour traffic to avoid paying the toll. It also leaves to Member States the right to apply tolls on roads of lower hierarchy as well as smaller vehicles, as long as they comply with the rules of the EU treaty.

It also defines that the tolls are to be computed so that they cover the relevant costs, defined as the construction, operation, maintenance and development costs of each road, avoiding costs that have already been amortised, but allowing full respect of existing concession contracts. It suggests that the main purpose of these pricing measures is not to collect additional revenue, but rather to stimulate behavioural adaptations by hauliers and their drivers that improve efficiency in the use of roads. As such, it suggests that national governments could introduce compensation to hauliers by reducing annual vehicle taxes.

Governments would be able to vary toll levels based on the vehicle type (weight and emissions classes) and on time of day (linked to level of congestion), and would be forced (after 2008) to impose such variation based on environmental sensitivity and population density of the area crossed by the road, and on its accident risk.

Specific mark-ups of the tolls were foreseen, up to 25% of their average value, in sensitive areas (specifically mountain areas) as an instrument for financing new infrastructure of European interest in the same corridor and transport zone, after approval on a case by case basis by the European Commission.

Earmarking of the toll revenues was mandatory for the benefit of the transport sector, either on the roads or more generally on the transport sector as a whole so that alternative modes of transport could be developed. Finally, an independent infrastructure supervision authority was proposed to oversee the operation of the various national systems and their application of the rules, namely on the application of the revenues.

This proposal contained several items that gave it virtually no chances of being approved in its original state either by Parliament or by the Council. The most obvious such items were those related to the independent supervision authority (unacceptable by sovereign governments) as well as the full earmarking of the revenues, which raises acceptability by the end users but reduces the degrees of freedom of governments and may ultimately be unjustified if the costs of a given motorway have been supported by the public budget and no significant investment in transport infrastructure of European interest is needed in the same corridor and region.

At the time of writing (June 2004), this proposal has been approved by the European Parliament on April 20th with some changes, but failed to get approval from the Council. In spite of the earlier request by the Council for an agreement among its members until March 2004, that agreement has not been reached in the Council, neither in March nor in June.

Following the European elections in May 2004 and the end of the term of the current Commission, a new Commission and Parliament will be in place in the second semester of 2004, and it is not clear what direction this Directive amendment could take.

This chapter aims at the analysis of the arguments developed in the discussion in the Parliament and in the Council, trying to understand the tensions between the various positions and to identify possible compromises towards producing an acceptable version.

The European Parliament introduced significant changes in the document it approved on April 20th (European Parliament, 2004). Besides deletion of the article related to the independent authority, the earmarking of revenues was changed from imposition to recommendation. An important change in the opposite direction was that non increase of the financial burden to the transport sector became an imposition instead of a recommendation.

The other main group of changes was related to the variation of tolls in relation to the various aggravating factors. In this respect, the European Parliament resolution imposes that the European Commission must present within a maximum delay of two years after entry into force of the new Directive, a "generally applicable, transparent, and comprehensible model for the assessment of all external environment-, congestion-, and health-related costs to serve as the basis for future calculations of infrastructure charges." This model, to be developed with the assistance of a committee with representatives of the Member States, should considerably reduce the current variability in the values attributed to those external cost elements across the EU, and thus reduce the risk of arbitrary fixation of the toll levels on the basis of such values.

Still in this group of changes, it states that the variations of tolls for a number of reasons (vehicle weight and emission class, time of day and day of week, environmental sensitivity and population density of the area, accident risk of the road) must not affect the weighted average of the tolls (with which total costs must be covered). This means that the weighted average of the tolls must be computed considering these variations as intrinsic elements of that average, and not that these variations would be allowed as additionals to that weighted average. Moreover, the highest rate should not be higher than 100% above the value of the lowest rate.

An important detail in these changes is that, although the European Parliament supports the mark-up of tolls in sensitive areas by as much as 25% to finance new

infrastructure of European interest, it constrains this new infrastructure to be of environmentally friendlier (than the road) modes.

Although the Council uses in its first reading the version submitted by the Commission and not that resulting – in a parallel process – from the discussions in Parliament, some of the evolutions visible in the version put forward by the Irish presidency for discussion in the Transport, Telecommunications and Energy Council of June 10–11, 2004 show some different concerns from those of Parliament.

In the text sent by the Committee of Permanent Representatives of the Member States (COREPER) to the Council on 3rd March 2004 (Council of the European Union, 2004a), a good synthesis of the situation was presented. It was stated that the main point of tension was between peripheral countries on one side and transit countries on the other, about the level of the tolls, the first group insisting that tolls "precisely reflect infrastructure costs" and the second fearing harmful effects of international commercial traffic on congestion and on the environment, and wishing to raise toll levels as a disincentive to the use of certain roads and in certain cases as a means to fund alternative infrastructure.

Other issues were also relevant: countries can also be grouped into two sets, according to whether they have developed their motorway networks at earlier dates and under public budgets, or at later dates and under concession contracts with private companies. The former group is now introducing tolls on heavy goods vehicles for coverage of (mainly) maintenance costs, but the latter has had tolls on all vehicles for construction and maintenance. Since the proposal of the Commission only allows consideration of construction costs as long as they are not amortised or included in a concession agreement, this could lead to very different toll levels in the countries of one and the other set.

Recognising these diverging interests, the COREPER proposed a compromise package, knowing that not all elements would appeal to all countries, but the package would be found by all as an improvement over the current Directive. The main elements of that package were:

A set of core calculation principles, later evolving into a common methodology for cost calculation, applicable only to future tolling arrangements.
The availability of mark-ups in specific cases for relief of congestion or environmental problems through construction of new infrastructure, but only where existing tolls are not higher than would result from the application of the core principles.
A more precise definition of construction costs, limiting the extent of the historic costs that can be passed on to the user in the form of tolls.
A wider set of possibilities to vary tolls away from the average level, but with a maximum ceiling on the degree of variation upwards.
A harmonised regime regarding discounts for frequent users, protecting against discrimination in the internal market.

Most of these elements go in a direction similar to the resolution of the European Parliament, although the ceiling for the variations from the average level was defined as 100% above that average and not (as in the Parliament resolution) 100% above the minimum rate.

But other points also deserved specific treatment, namely the scope of application of tolls, both in terms of the roads and of the vehicles where they can be applied. For the case of roads, the suggested compromise of COREPER was that each Member State would simply designate the parts of their TEN network where tolls would be applied, without having to define the criteria for exclusion of certain parts. For the vehicles, the basic proposal of the 3.5 ton lower limit was to be maintained, but with the possibility for Member Countries to apply tolls only above the current 12 ton limit.

Also on the issue of earmarking of revenues a compromise was sought, trying to keep the logic of the Commission proposal but reducing its scope of application. The compromise suggested was that earmarking would only be applied to the revenue above that corresponding to the infrastructure costs, i.e. to the mark-ups when they are imposed.

An interesting point introduced by the COREPER is related to the possibility of application of discounts on tolls to frequent users. Since one of the purposes of the tolling system is to manage the amount of heavy goods traffic on the tolled roads protection of frequent users seems strange. In this version, three reasons are given to allow such discounts:

Savings in administrative costs made by the operator.
Competition against other routes or modes, so that the loss in total demand and revenue do not have a significant impact.
Social acceptability of tolls by local community, but imposing two further conditions in this case: these discounts may not be passed to the other users as toll increases, and they should not distort competition in the internal market.

Certainly the most used of these three reasons will be the third one. But the second reason is a bit awkward as if toll revenue from each road was a significant concern. This will be the case in private concessions, but in those contracts the risk of improvement of adjacent roads or of opening of new or improved rail connections has always been passed to the concessionaire without any protection clause in the contracts. But here, even for the roads under direct administration by the State, it

would seem that revenue stability (or maximisation) would be a stronger concern that containment of traffic volumes.

The Irish Presidency of the Council overtook this text and presented on 28th April to the Working Party on Land Transport a slightly modified text as the basis for its meeting of May 4th. The most significant modification was exactly on earmarking, where it reaffirmed that under the principle of subsidiarity "Member States shall determine the use to be made of revenue from charges for the use of road infrastructure," adding immediately after a recommendation of use of that revenue to "benefit the transport sector and optimise the entire transport system," thus coming to a very similar position to that taken by the European Parliament on April 20th.

This proposal was still worked by the Working Party on Land Transport, with some changes introduced on the point related to mark-ups. According to the document released by the Press Office of the Presidency immediately before the meeting of the Council on Transport, Telecommunications and Energy of June 10th four important changes to the mark-up compared to the text examined by the Council in March:

The scope of the mark-up has been limited to mountainous regions.
The level of 25% has been retained only for the specific case of the Brenner tunnel.
The level for any other mark-up is reduced to 15%.
The possibility to combine the mark-up and toll variation on a particular route is constrained.

In the words of the press release after that Council meeting (Council of the European Union, 2004b), "the Council had an exhaustive debate" and "delegations expressed differing views on some key aspects of the Presidency compromise proposal," but the Presidency concluded that it was impossible to reach agreement on the revision of the existing "Eurovignette" Directive at this stage and referred the file back to COREPER for further examination.

What can be concluded from all these steps? The Statement of the EU Council in December 2003 welcoming the proposal and urging its approval for March 2004 is an indication that most Member States expected to reach a better situation with a revised Directive than they have with the current one, but the subsequent difficulties show that such a majority of Member States probably had different and incompatible solutions in their minds when such a statement was issued.

There is no publicly available information on what were the "key aspects" of the Presidency compromise proposal that led to expression of different views by the delegations, but the earlier text of the COREPER, and the adjustments made by the Presidency over the text prepared by COREPER seem to indicate that, at the Council level:

The formulation now reached for the definition of construction costs and the soft approach on earmarking should prove acceptable by most.
The development by the Commission of a unified methodology to calculate external costs can provisionally get support from most Member States, although the process of its development, accompanied by a Committee with direct representation of the Member States, could be more contentious.

but two deeply dividing issues remain:

(a) The extent of the variations and mark-ups, based on multiple factors, beyond the tolls levels that reflect the construction, operation, maintenance and development costs.
(b) By how much the different age of the infrastructure may influence, through the ratio of its construction cost that is allowed for inclusion in the toll, the overall level of that toll.

Then, in divergence with the Parliament resolution there is still the issue of non increase of the financial burden for the transport sector, and the constraint that the revenue from the mark-ups be used for infrastructure of friendlier modes.

In the relatively short-term there seems to be only one way to get agreement in the Council, which would consist of the following solutions for the two dividing issues:

(c) To start by applying a relatively low maximum level of variation above the average toll level, until the unified model for the computation of the external costs has been developed and accepted by a majority of Member States, and following the results of the model, accepting that this might imply increasing the maximum level of variation.
(d) To accept that in all roads, independent of their age, a minimum level of construction cost could be included in the average toll level (for instance as a provision for reconstruction), so that the difference between the tolls in a new and in an old road (all other factors being equal) would not be more than say 30%.

All countries virtually depend on road transport for their trade, but some are more affected by foreign trucks on their roads. Moreover, the generalisation of road tolls across Europe represents a significant change of the pattern of costs for hauliers, especially those who work mostly in countries currently not applying tolls. Since margins are quite small in this very competitive business, many hauliers will fear that their competitors might be in a better position to reorganise production and

make efficiency gains that would allow to internalise some of the additional costs without passing them to their clients, thus gaining further competitive advantage.

So, governments are keen to receive additional revenue from this sector, especially from foreign trucks who currently pay only a small part of the costs they impose; but they do not want to have toll levels so high that their own industry is disrupted.

A compromise for a two-step solution as suggested above, might allow the introduction of the new principles and a "soft landing" of the toll variations on the cost structure of the industry. This would be the kind of compromise that so often has been necessary to reconcile the diverging interests of various European countries. The alternative is to keep fighting for more radical solutions, delay the approval of the amendment and continue to operate under the current Directive 1999/62.

REFERENCES

Council of the European Union (2004a). Document 6985/04, Inter-institutional File 2003/0175 (COD), Addendum 1, from the Permanent Representatives Committee (part 1) to the Council, on 3rd March 2004, Brussels.
Council of the European Union (2004b). Press Release after the 2589th Council Meeting on Transport, Telecommunications and Energy, Luxembourg, 10–11 June 2004. Accessed 22nd June 2004. http://ue.eu.int/ueDocs/cms Data/docs/pressData/en/trans/80931.pdf.
European Commission (2003). Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Directive 1999/62/EC on the charging of heavy goods vehicles for the use of certain infrastructures, COM(2003) 448 final, 2003/0175 (COD), 23rd July 2003, Brussels.
European Parliament (2004). Legislative resolution on the proposal for a European Parliament and Council directive amending Directive 1999/62/EC on the charging of heavy goods vehicles for the use of certain infrastructures (COM(2003) 448-C5-0351/2003-2003/0175(COD)) – (Codecision procedure: First reading), Brussels.

11. CONCLUSIONS: PAYING FOR ROAD USE, THE WAY AHEAD

Jose M. Viegas ´

As pointed out in the introduction, the European Commission has been very active in promoting pricing as an important instrument for efficient use of transport infrastructure, acting in three fronts: policy documents – Green Papers for consultation, White Papers for statement of the Commissions policy intentions, Research Projects, and Directives.

In both road and rail there have been some difficulties, because direct payment for use of infrastructure is not in the tradition of these modes, albeit for different reasons: indirect payment through taxes in the case of roads and integration between infrastructure and operations in the case of rail.

In the first part of this book (Chapters 1 to 5) we have shown the basic motives, and the conceptual and technological solutions for imposing road tolls on trucks, discussing the legal, contractual and institutional issues and reviewing the acceptability arguments for this kind of pricing.

In the second part (Chapters 6 to 10), our attention turned to the situation in the European Union, starting by a general consideration of the multiple existing situations and possible solutions, discussing in which cases each solution would be a better choice, and then looking in more detail at the discussion and trends in various countries, as well as at the Union level. This has shown how diverse the national perspectives on the theme can be, opening the discussion about the relevance and necessary reach of interoperability of electronic fee collection schemes. The results of a complex modelling exercise have been presented next, showing the scale and geographical spread of economic and environmental impacts

of the imposition of these tolls, for three different tolling policy scenarios. Finally, a review has been made of the recent evolution of political discussions surrounding the amendment of the existing Directive on road tolls and charges for trucks in the EU.

In balance, we see that:

A multiplicity of motives is invoked to justify the application of road tolls to trucks, mostly around the user pays principle, and in connection to road maintenance costs.
There are good arguments and technological means for imposing tolls on trucks for using some roads, but no immediately available solution to apply tolls on all roads with different levels of price depending on the type of road, traffic conditions and environmental sensitivity.
Some countries are moving towards application of tolls for trucks on motorways, but some do not consider it advantageous or justifiable.
Generally speaking, the application of tolls will have very little effect on modal shift of freight, but a clearer effect on route choices (sometimes through different countries) depending on where they are applied.

It is also clear that this discussion is taking place in a period of strong change, with two main dimensions affecting freight transport by road: on the one hand, the recent EU enlargement (as well as subsequent ones) will create the basis for trade growth, with goods travelling over longer distances; and on the other international transport is carried out on the youngest part of the fleet, with remarkable and sustained reductions of all types of emissions. Also, as trucks share the motorways with private cars, and car ownership continues to increase, congestion around urban agglomerations is not likely to decrease, on the contrary.

The European Commission continues to strive for intermodality on freight transport, but facing low levels of success for the effort, largely because it goes on being easier to the road hauliers to adapt to the business needs of its clients than to the operators of the other modes to approach their production and business models to those of the road industry.

These pressures have created an environment in which high expectations have been placed on the application of road tolls and charges to trucks.

The European Commission White Paper on Transport Policy (European Commission, 2001) emphasises that these road charges proposed do not constitute an additional cost for the road users, rather they transfer what previously was a fixed cost to a variable cost. But it is not clear that this view is shared by national governments, who are not committed to that statement, particularly when the measures included in the Directives only affect part of the vehicles on part of the road network. The fact that the text under discussion in the Council

(see Chapter 10) does not mention this issue – in opposition to the version approved in the European Parliament – confirms this non commitment.

This double partiality – tolls only on some roads and only for some vehicles – is normally justified on the grounds that trucks are responsible for the vast majority of maintenance costs. However, if the total costs of construction, operation, maintenance and development of the motorways, plus the external costs related to environmental impacts and congestion are considered (which is the cost package that is being targeted in the Directive amendment), the picture is no longer the same: the number of lanes and the radius of curves of motorways are largely determined by the number and driving patterns of private cars, and the contribution of private cars for congestion and pollutant emissions is in most cases at least as great as that of trucks.

So, the reason for this choice of tolling only trucks on motorways (with the recent addition of parallel competitive roads) has to be sought somewhere else, namely on the grounds of social acceptability and of an expected modal shift from road to rail. A simple and rough pricing tool is being used to pursue multiple and complex combinations of objectives, inevitably generating misfits between the "hammer" and the "mould." The existence of other objectives besides the application of the user pays and polluter pays principles, i.e. correct allocation of costs to their originators creates distortions that make any solution much more difficult to accept by all or even the majority of Member States.

On the acceptability dimension, it is obvious that any pricing measure is always more acceptable if it is the others who have to pay. Trucks are an easy target inasmuch as they are frequently depicted as the bad guy in the development of European Transport: not only is their share of environmental impacts at the local range (noise, vibrations, particles) clearly much bigger than that of cars, thus making a strong impact on popular perceptions of damage attributable to each of these groups, but also in many regions a significant part of the trucks driving through are foreign, so obviously part of those others. Moreover, national hauliers and drivers are not so numerous that they would represent a serious risk in electoral terms in any particular region or country.

Concerning the hopes for modal shift in favour of rail, even considering these tolls as fully additional charges (no reduction of the fixed costs of owning and operating a truck), what earlier research (SOFTICE, 1999), the modelling work presented in Chapter 9, and the Swiss experience with road charges show is that:

There is some capability of hauliers to improve organisation of their services in order to absorb part of the charges without passing it fully as additional price to their clients.

The level of charges justifiable by the full cost of road maintenance is not enough to divert significant amounts of freight from road to rail.

To a large extent this is a consequence of the protected economic environment in which the railways have been allowed to live, as it has eroded their capability to provide services as required by their clients and thus preserve their historic market shares. To achieve significant modal shift from road to rail it clearly is not enough to intervene on the price of road transport: determined and effective action is needed towards greater responsibility of the railways for the economic results of their activity.

So, even if considerable additional revenue is made available for the States through these charges, thus relieving the public budget, one of the main goals of the policy is missed, and considerable problems remain in terms of consistency of these measures.

The discussion process on the amendment of Directive 99/62 has been difficult and is not complete at the time of writing, but it has been centred on a few particular items, not on the key points of long term inconsistency of the current practices of pricing for the road sector.

As the preceding chapter may have shown, this is no longer about creating an efficient pricing instrument for the use of the road by trucks, but really about getting (more) money out of an industry that has been very successful in gaining market share over the railways and so should be able to pay a bit more. If foreign trucks are involved, it is even easier to justify before the national audience the political correctness of the measure.

Some points in the recent discussion deserve special attention:

Mark-up for congestion is an appealing concept, but it really makes very little sense if tolling is not applied to all road vehicles, at least at the periods of heavier congestion. Most congestion occurs in and around urban agglomerations, and with a percentage of heavy vehicles in the traffic flow that rarely exceeds 15%. Charging more to HGVs in periods of heavy congestion may amount to pricing them out so that private cars can enjoy the road all for themselves, but without paying.

Moreover, this mark-up is supposed to be used for construction of new infrastructure, but there is a non negligible risk that construction of that new infrastructure may be opposed with arguments that it will induce more traffic and thus create congestion or environmental problems downstream. During the possibly long period of discussion, when there is no guarantee that the new construction will proceed, should the mark-up still be active? And why should new construction be financed mainly by HGVs if they are not the main cause of congestion?

Earmarking of revenues for the transport sector is, as already discussed in Chapter 5, an element that generally improves acceptability. In general terms, if the value of the tolls really corresponds to costs, this money should be used to cover those costs, as is the case with private concessions. But if the component related to environmental costs becomes significant, there is no special reason why the corresponding revenue should be spent in the transport sector: the amounts generated are quite high, and rather quickly the situation would be reached where very low value for money would be obtained from compulsory investments in the transport sector, just because the money was there.

In urban road pricing projects it has been found that using those revenues to invest in public transport is seen positively by many, but this is possibly because these citizens fear being priced out of their own cars, and so would benefit of an improved public transport system, which they would be using with some frequency. But in the case of road haulage, the industry and drivers will not be considering a conversion to railways, and so this perspective of improved acceptability by financing the alternative modes is generally not applicable.

However, the strange fact is that cross-subsidisation of the railways by road tolls is put forward without any imposition of improved performance from the railways, so, in fact, risking that this money is transferred really a fonds perdu ` . In many cases across Europe, the reasons for the decline of the market share of railways have little to do with lagging infrastructure, so investments in this expensive part of the system risk being of little impact if the other factors are not attacked in parallel. But for this, the traditional automatic and indefinite financial guarantees from member States towards their national railways has to be curtailed.

Environmental mark-ups are a reasonable instrument, but it makes sense to impose a ceiling to their relative value, at least until adequate paths can be made available for the critical corridors directly affected. This applies to sensitive regions, mainly the Alps and Pyrenees, where there is a need to balance the economic perspectives of European development through trade with the defence of the environment, avoiding an over-representation of possibly selfish interests of those sensitive regions, who may just want to minimise transit traffic.

Even a full economic evaluation of the environmental impacts of road traffic (using the unified methodology promised within two years) inevitably reflects the valuation of a series of amenities by the affected residents, but not of a series of other positive goods which become less available to the peoples of the regions whose trade is affected. So, we should not be na¨ıve to the point of admitting that the answer to this problem is only a question of finding the right scientific methodology: it has to do with the political balance between the interests of

different kinds of peoples in different regions. So, it would be sensible to define the upper limit of the mark-up for a given region in connection to the relative quality of the alternative paths through less sensitive regions.

The possibility of also charging for the use of "normal" roads that are parallel to the higher quality roads targeted in this Directive has to do with the risk of detour, as noted in earlier chapters, even if the current version of the Directive proposal does not go as far as suggesting a DAREA scheme, where all roads within a certain perimeter are subject to tolls. Even if the problem is real, there is enough experience in southern European countries with tolled motorway concessions to show in what kind of circumstances it gets serious, and how to mitigate it through physical measures.

The limited solution proposed in the last version discussed in the Council may be of very difficult application as it imposes a significant cost for even local trips in some regions, in a clearly different regime to that prevailing in other parts of the same country. The counter measure suggested is discounts for frequent users, which is an economic signal to the exact contrary of the intentions of tolling for demand management: it creates an incentive to use more, as the unit cost is reduced.

As mentioned earlier DAREA schemes are politically acceptable for wholecountry application, but very hard to pass for limited parts of a country, since it imposes very different fiscal regimes for companies (or even citizens if such tolls were applied to private cars as well). In this area as well, some common sense would be welcome, by recognising that some problems can be better tackled with instruments other than pricing.

It is interesting to note that some of the greater difficulties in this discussion have to do with the attempt to define a new pricing layer on (some) transport activities, trying to have a very rational new layer but not daring to touch the underlying parts, which have plenty of inconsistencies, but are already there, with all citizens used to paying and not questioning why they came to be defined as they are.

The main reason for this political inertia is the very significant weight that road transport related taxes represent in State budgets: in a sample of nine EU countries studied in (TISpt, 2002), it was found that these taxes represented between 4.6% and 10.2% of the total taxation of those States. Any change of the fiscal and pricing approaches to road transport thus has to bear in mind the high social costs that sudden movements of that income flow would impose, and accordingly has to be planned with transition strategies that minimise the risks of such turbulence.

To address the problem of those inconsistencies of pricing policies, a general move towards variabilisation of motoring costs is needed, i.e. introducing changes

in the fiscal treatment of road transport such that a smaller part of the vehicle costs are fixed (linked to ownership) and a larger part of those costs are linked to its actual use.

In spite of the difficulties of such a change, there are signs of policy trends in this direction in Europe and the U.S. While in general these trends are mostly justified by the objectives of demand management and environmental preservation, in the case of the European Union the argument of mitigation of barriers and distortions to the single market for cars is also present. In all these cases, arguments are geared towards private cars, but it is difficult to see why these changes would not equally cover the case of trucks.

Significant documents published in Europe include a report for the European Conference of Ministers of Transport (INFRAS, 2000), a Communication and a Consultation document from the European Commission (European Commission, 2002, 2004). In parallel, the Commission has launched since the late 1990s a series of R&D projects in which the concept has been studied and advocated as an effective tool.

In the U.S., several pilot projects are under way, promoted and financed by the Office of Policy of the Federal Highway Administration, under the general title of "value pricing." Whereas most of these have an element of tolls or charges related to congestion levels (HOT lanes and cordon tolls), there are also projects related to mileage-based insurance and variabilization of fixed auto costs (FHWA, 2004).

As we move towards this more rational pricing system, it is normal that road tolls for trucks could become more sophisticated, with multiple components defining the price per kilometre at each place in space and time. As the consumption of road transport resources and services approaches the pricing patterns of most other goods and services in the market, we should expect that price levels vary within a polygon defined by the following basic principles:

Costs of provision of those resources and services have to be covered, in order to make them available over the following activity cycles.
When the provision or the consumption of these services generates (external) costs beyond those directly involved in the exchange, these costs must also be covered and passed on through prices to the consumer, so that adaptation of behaviour can be expected by introduction of new processes for provision, as well as by adjustment of the volumes and patterns of consumption.
Prices should be higher when the costs of provision are higher or when there is scarcity of available resources.
When consumers have a choice, the price charged has to bear a relation with the quantity and quality of the goods or services provided.

These principles show that, taking a long term view of the preservation of the transport system but also a short-term concern for its efficient operations, the conceptual basis for prices has to be average cost pricing (including external costs), but allowing for marginal cost based adjustments and, very importantly but frequently forgotten, providing value for money.

These principles, with the exception of the latter, are already largely included in the proposed amendment of Directive 99/62, in all versions (initial proposal by the Commission, resolution by the European Parliament, and under discussion in the Council). The controversy is mainly on the extension of the allowed increases over the average.

Based on all these elements it seems correct to recommend that a stronger general move towards variabilisation of motoring costs be made, affecting cars as well as trucks, and the amendment of Directive 99/62 be made under this general direction.

At this stage the amendment should contemplate the basic improvements suggested over the earlier version, like lowering the weight limit and extending the road set under pricing to the whole Trans-European road network, clarifying what road costs are allowed and respecting existing concessions, but adopt relatively low ceilings for the various types of mark-ups until a more comprehensive pricing reform of the transport sector (at least covering all road vehicles) can be prepared and adopted.

Imposing high variability of road tolls on top of not only high fixed costs of truck ownership but especially of almost free access to the same roads by private cars is certainly unfair and risks creating a negative environment that would jeopardize the process of evolution towards the necessary variabilization of costs for all road vehicles. This is the real challenge on the road ahead towards a better transport system, not the bargaining between peripheral and transit countries about how much one's trucks should pay to use the other's roads.

REFERENCES

European Commission (2002). Communication on taxation of passenger cars in the European Union (COM (2002) 431). Accessed August 9th 2004. http://europa.eu.int/comm/taxation customs/taxation/vehicles taxation/.

European Commission (2004). Consultation on proposals for passenger car taxation (14th July 2004). Accessed August 9th 2004. http://europa.eu.int/comm/taxation customs/taxation/ consultations/car taxation en.htm.

FHWA (2004). Description of pilot projects on value pricing. Accessed August 9th 2004. http://www.fhwa.dot.gov/policy/otps/projdesc.htm.

INFRAS (2000). Variabilisation and differentiation strategies in road taxation – Theoretical and empirical analysis. Final report. Paris: ECMT Group on Transport and the Environment.

SOFTICE (1999). Survey on freight transport including a cost-comparison for Europe. Final report. European Commission Transport Research Fourth Framework Programme.

TISpt (2002). Study on vehicle taxation in the member states of the European Union. Study conducted for the European Commission – DG Taxation and Customs Union, January 2002 Accessed August 9th 2004. http://europa.eu.int/comm/taxation customs/taxation/ car taxes/vehicle tax study 15–02–2002.pdf.

viegas10evolutioneudirective2005.md 513 KB История Исходник

status: converted

INTERURBAN ROAD CHARGING FOR TRUCKS IN EUROPE

RESEARCH IN TRANSPORTATION ECONOMICS

Recent Volumes:

INTERURBAN ROAD CHARGING FOR TRUCKS IN EUROPE

Photocopying

Derivative Works

Electronic Storage or Usage

Notice

CONTENTS

LIST OF CONTRIBUTORS

GLOSSARY

FOREWORD

NOTE

1. INTRODUCTION: PAYING FOR ROAD USE

REFERENCES

2. BASIC ROAD PRICING SOLUTIONS

Heike Link and Louise Stewart-Ladewig

1. OBJECTIVES OF ROAD PRICING

1.1. Efficient Use of Infrastructure (Demand Management)

2. INTERNALISING THE EXTERNAL COSTS OF TRANSPORT

2.1. Financing Investment in Transport Infrastructure

3. CHARACTERISTICS OF ROAD PRICING STRATEGIES

4. THE COSTS TO BE INCLUDED AND THE CHARGING PRINCIPLE

4.1. The Vehicle Types to be Charged

4.2. Distance Versus Time Related Charging Schemes

4.3. Price Variation

4.4. Charging Technology

5. BASIC SOLUTIONS FOR INTERURBAN ROAD PRICING AND THEIR SCOPE

5.1. Distance Dependent Network Pricing (NET)

5.2. Distance Dependent Area Pricing (DAREA)

5.3. Driving Permit or Area Licensing (PERM)

5.4. Road Pricing Schemes with Different Incidence

5.4.1. Passage tolling (PAS)

5.4.2. Cordon pricing (COR)

REFERENCES

3. TECHNICAL AND PROCEDURAL DESIGN ELEMENTS

Andrea Felix

1. GENERAL DESIGN ELEMENTS

2. BASIC TOLLING DESIGN ELEMENTS

2.1. Charging Principle

2.2. Traffic/Road Situation

2.2.1. No Impact to Existing/Nominal Geometrical Profile of Road and Traffic Flow is Accepted

2.2.2. Impact to Existing/Nominal Geometrical Profile of Road and Traffic Flow is Accepted

3. TARIFF CLASS DETERMINATIONS AND TARIFF MODULATION ELEMENTS

3.1. Tariff Class Determination and Tariff Modulation

3.1.1. Vehicle Characteristics

3.1.2. Tariff Modulation

4. TECHNICAL OPTIONS FOR EQUIPPED USERS

4.1. Localisation of Vehicle and Initiation of Charging Process of Equipped Users

4.1.1. Road side Infrastructure/Infrastructure External of Vehicle

4.1.2. On Board Based System

4.2. Classification of Equipped Users

4.3. Charging Equipped Users

4.3.1. Central Account

4.3.2. On-Board Account

5. TECHNICAL OPTIONS FOR NON-EQUIPPED USERS

5.1. Initiation of Charging Process of Non-Equipped Users

5.1.1. By Road Side Infrastructure

5.1.2. By User of the Charging System

5.2. Classification of Non-Equipped Users

5.3. Charging Non-Equipped Users

5.3.1. Using Non-Stop Lanes

5.3.2. Using Stop Lanes/Service Area

6. ENFORCEMENT CONCEPT AND TECHNICAL APPROACH

6.1. Enforcement Concept

6.2. Enforcement of Equipped Users

6.3. Enforcement of Non-Equipped Users

6.4. Equipment Used for Enforcement

4. LEGAL, INSTITUTIONAL AND CONTRACTUAL ISSUES

Rosario Mac ´ ario ´

1. INTRODUCTION

2. STRUCTURING THE FUNCTIONS OF AN IRPS

3. PERFORMANCE ASSESSMENT OF DIFFERENT INSTITUTIONAL SETTINGS

A "Mixed" institutional model (Fig. 2) assumes the following:

Table 6. (Continued )

4. CONTRACTUAL DIFFICULTIES AND CONTRACT EVOLUTION

5. CONCLUSIONS

REFERENCES

viegas10evolutioneudirective2005.md 513 KB

История Исходник