World Transport Policy and Practice Volume 18.4 Sept. 2012

ECO-LOGICA LTD ISSN 1352-7614 © 2012 ECO-LOGICA LTD ERIC BRITTON MANAGING DIRECTOR, ECOPLAN INTERNA- PROFESSOR JOHN WHITELEGG TIONAL, STOCKHOLM ENVIRONMENT INSTITUTE AT THE CENTRE FOR TECHNOLOGY & SYSTEMS YORK, STUDIES, UNIVERSITY OF YORK, 8/10 RUE JOSEPH BARA, YORK, F-75006 PARIS, FRANCE YO10 SYW, U.K PAUL TRANTER EDITORIAL BOARD SCHOOL OF PHYSICAL ENVIRONMENTAL & MATHEMATICAL SCIENCES, PROFESSOR HELMUT HOLZAPFEL UNIVERSITY OF NEW SOUTH WALES, UNIVERSITÄT KASSEL, AUSTRALIAN DEFENCE FORCE ACADEMY FACHBEREICH 06 - ARCHITEKTUR, STADT- CANBERRA ACT 2600, AUSTRALIA UND LANDSCHAFTSPLANUNG AG INTEGRIERTE VERKEHRSPLANUN- PUBLISHER GOTTSCHALKSTRAßE 28, ECO-LOGICA LTD., 53 DERWENT ROAD, D-34127 KASSEL GERMANY LANCASTER,LA1 3ES, U.K. TELEPHONE +44 (0)1524 63175 E-MAIL: JOHN.WHITELEGG@SEI-INTERNA- TIONAL.ORG HTTP://WWW.ECO-LOGICA.CO.UK

CONTENTS

Page

Editorial 3

Cover photo description 4

Abstracts and Keywords 5

Don’t Shoot Me, I’m Only The Transport Planner (apologies to Sir Elton John). Jeff Kenworthy 6

Traffic Growth: Modelling a Global Phenomenon. David Gargett 27

Congestion Offsets:Transforming Cities by Letting Buses Compete. Matthew Bradley and Jeff Kenworthy 46

Book announcement 68

2 World Transport Policy and Practice Volume 18.4 Sept 2012 EDITORIAL The three articles in this issue make an should be eliminated. It will be a lasting important contribution to transport de- monument to all those involved in public bate and point clearly to different ways policy and road safety in the period 1945- of viewing some of the key public policy 2012 that large scale human tragedies issues currently underpinning transport are tolerated and human problem solv- and urban planning thinking. Kenworthy’s ing skills that could reduce this appalling “Don’t shoot me, I’m only the transport total are so rarely deployed. One of the planner” subjects the world of traffic and more usual rebuttals to our stance is a ref- transport planning to a detailed forensic erence to road traffic “accident” statistics analysis to show that the design and use and the fact that deaths and serious inju- of transport and land use models, predict ries in most so-called developed countries and provide and other traditional meth- have declined. This would now appear to ods and approaches are fundamentally be slightly off the mark. Recent UK road flawed. Automobile dependence and sup- death and injury statistics have shown an port for higher levels of automobility has increase in death and injury: gone hand in hand with a set of false assumptions and methods that are still re- Key points sistant to challenge even though a clear • The annual number of people killed evidence base for such challenge exists. in road accidents reported to the police Gargett analyses traffic growth data in 25 has increased, by 3 per cent, from 1,850 countries and shows that much of the traf- in 2010 to 1,901 in 2011. This is the first fic growth we have accepted as inevitable increase since 2003. and likely to continue into the indefinite • The number of people reported future has stalled. We are now in a world killed or seriously injured has also in- where traffic growth is far more uncertain creased by 2 per cent to 25,023 from than at any time in the past and the chal- 24,510 in 2010 – the first annual increase lenge for politicians and planners globally since 1994. is to respond to this changed set of cir- • The total number of casualties cumstances with new policies not predi- (slight injuries, serious injuries and fatali- cated on growth. Kenworthy and Bradley ties) in road accidents reported to the po- in their “congestion offsets” article return lice in Great Britain in 2011 continued to us to the need to generate new policy ide- fall, by 2 per cent, from 208,648 in 2010 as and suggest a radical approach to deal- to 203,950 in 2011. ing with congestion. They suggest that • Total reported child casualties congestion should be handled in a way dif- (ages 0-15) have continued to fall, by 0.5 ferent to that employed by the London and per cent in 2011 to 19,474. The number of Stockholm charging regimes and should children killed or seriously injured also fell, be viewed as a phenomenon that is caused decreasing by 4 per cent to 2,412 in 2011, by over use of a common resource (road from 2,502 in 2010. space) by one group and under-use by • Vehicle traffic levels are broadly another group. The under use group are stable after falling for 3 years. The overall financially rewarded for their modest be- casualty rate for accidents reported to the haviour. We welcome comments on con- police per billion vehicle miles continued to gestion offsets and on all articles in this is- fall to 666 per billion vehicle miles, com- sue of World Transport Policy and Practice. pared to 681 casualties per billion vehicle miles in 2010 but the killed or seriously For 18 years this journal has expressed injured rate increased to 82 per billion ve- varying degrees of discontent and disbe- hicle miles. lief around the limpness and weakness of Source: UK Department for Transport, June 2012 public policy designed to deal with over 1 http://www.dft.gov.uk/statistics/releases/reported- road-casualties-gb-main-results-2011/ million road deaths every year globally, equivalent to about 3000 each day. WTPP Any increase in death and injury can be has always taken the view that death, described as a “blip” and ignored but misery and tragedy on this scale is sim- equally it can point to a lack of “bite” in the ply unacceptable in a civilised society and overall approach to dealing with death, in- 3 World Transport Policy and Practice Volume 18.4 Sept 2012 jury and tragedy and we would argue that Cover photo road safety policy in the UK has not yet got to grips with the urgent need to adopt The cover photo was produced by a cycling a total ethical approach along the lines of campaign group in Canberra, Australia the Swedish Vision Zero road safety policy. (Cycling Promotion Fund, 2012). It shows UK road safety policy lacks a clear sense of in a remarkably clear and dramatic way determination to eliminate death and seri- the impact of transport choices on the use ous injury, is accepting of anti-social and of urban space. If a large majority of urban criminal behaviour on the part of speed- residents choose to use a car the street is ing motorists and those who drive ag- overwhelmed and urban life cannot func- gressively and without care for vulnerable tion. Buses, bikes and feet do a much bet- road users and does not implement the ter job and sustain a vibrant urban quality kind of cycling infrastructure to be found of life. These non-car choices look bet- in Denmark or the Netherlands or the kind ter, feel better, sustain attractive places of car-free infrastructure to be found in and cost less to provide and maintain. A Freiburg in southern Germany. It is very similar photo was published in “Gute Ar- clear that driving a car brings with it an gumente: Verkehr” (Seifried, 1990) with element of aggression and predisposition a commentary that left no doubt at all to violence. This was described in 1995 around the significance of space utilisation by Helmut Holzapfel in Volume 1, number concepts and the wastefulness of a trans- 1 of this journal. Violence and aggression port strategy that provided for motorised are not compatible with a civilised society transport. Seifried says “one motorway and deeper thought, determination and intersection in Germany uses as much implementation are needed to get rid of space as would be occupied by a complete these behavioural characteristics. historical inner city area e.g. Salzburg with thousands of residents, 4000 dwellings, So what can we do to eliminate death and 430 businesses, 16 churches, 13 schools serious injury in the road traffic environ- and one university”. ment? We welcome articles and debate on this subject but there are 4 rather obvious We are very grateful to the Cycling Promo- things that have to be done in a European tion Fund in Canberra and Stephen Hodge or North American/Australasian context: for permission to use this photograph

Implement 20mph/30kph speed limits Cycling Promotion Fund (2012) Campaign on every road in every urban area and Photo through every village or small settlement http://www.pozible.com/index.php/ar- chive/index/8424/description/0/0 Recruit and train citizens to do speed mon- www.cyclingpromotion.com.au itoring on their own roads with high quality equipment and recording devices to pho- Siefried, D (1990) Gute Arguemnte: tograph offenders and registration plates Verkehr, Beck’sche Reihe, Verlag C.H.Beck, and speeds. Citizens will take responsibil- Munich, Germany, ISBN 3 406 34003 2, ity for detection and reporting and existing page 46 police and judicial systems will be respon- sible for due process.

Implement a “3 strikes and you are out” policy so that after a 3rd speeding infringement the vehicle is confiscated

Design, fund and build high quality walking and cycling routes in all urban areas, villages and settlements so that pedestrians and cyclists can maximise the use of traffic free routes

John Whitelegg, Editor 4 World Transport Policy and Practice Volume 18.4 Sept 2012 ABSTRACTS AND KEYWORDS

Don’t Shoot Me, I’m Only The Trans- Congestion Offsets: port Planner Transforming Cities by Letting Buses (apologies to Sir Elton John) Compete

Jeff Kenworthy Matthew Bradley and Jeff Kenworthy

Abstract: Abstract: An historical review of the origins, under- Cities around the world have been unable lying philosophy, development, use and to keep up with demand for road space, abuse of land use-transport or traffic mod- and as a result suffer from congestion. els is provided. It is argued that the way This paper argues that the consequence of these models have been used, and often this congestion has been not just poorly still are used today, has resulted in extreme performing roads, but a structurally inef- automobile dependence with its associated ficient urban transport solution, due to the negative social, environmental and eco- impact of congestion on bus services. As a nomic impacts. Contrasting responses to way of controlling congestion, road pricing the predictions from such traffic models in has been rejected by most cities because cities around the world are reviewed. The it is difficult to sell to the public and diffi- many failures of these models at both a cult to deploy. An examination of both the strategic and technical level are discussed. strengths and weaknesses of road pricing The paper then outlines how cities might is used as a basis from which to develop respond to the shortcomings of traditional an alternative method of congestion con- transport planning which uses such mod- trol, namely congestion offsets. This ap- els. A key underpinning of the paper is proach treats roads as a commons, and the idea that transport planning can no does not argue for the creation of a mar- longer be based on a ‘predict and provide’ ket for road space, as road pricing does. approach, which treats traffic as a liquid. It is argued that congestion offsets would Rather, communities must adopt visioning allow congestion to be controlled in a processes based on a ‘debate and decide’ simple and fair manner, thereby enabling approach, which treat traffic as a gas that buses to emerge as a real competitor to expands and compresses according to how cars in urban areas. The historical absence much road space is provided for it. of congestion control in cities has allowed Key Words: massive distortions to build up; cities Traffic models, transport planning have become awash with both cars and bitumen. Congestion offsets would allow Traffic Growth: Modelling a Global these distortions to be slowly unwound, Phenomenon thus helping to transform cities more into how, it is argued, they should have always David Gargett been. Keywords: Abstract: Congestion, Congestion Offsets, Road The levels of traffic on the road are a key Pricing, Bus Rapid Transit, Transit-Orient- feature of transport systems around the ed Development, Demand Management world. Understanding the determinants of past and likely future traffic growth is thus important in assessing how we handle the challenges that growth presents. A recent report described the different patterns of traffic growth in 25 countries, and their determinants. The current article is an extract of that report, focusing on three of the 25 countries. Key Words: Traffic; oil and petrol prices 5 World Transport Policy and Practice Volume 18.4 Sept 2012 Don’t Shoot Me, I’m Only The Trans- emerging economies have tended to un- port Planner critically import these post-World War 2 (apologies to Sir Elton John) traffic modelling approaches. Partly as a consequence of the conceptual framing Jeff Kenworthy of transport problems inherent in these models, focussing almost entirely on the 1. Introduction impossible dream of congestion removal, many are building freeways at an alarming A big determinant of the shape of cit- rate (e.g. in India and China). ies is the role and direction of transport planning. The different approaches and Counter to this rather negative story, outcomes of this discipline can be clearly there is now clear evidence that other ap- seen in a simple comparison of the conti- proaches are possible, which yield very dif- nental European cities with the American ferent results for cities (Goodwin, 1997). and Australian cities. While these Europe- A fundamental underpinning of the way an cities retain a balance between private that land use-transport models operate and public transport and non-motorised is to treat traffic analogously to a liquid, modes and city centres are mostly intact something that retains its original volume with extensive pedestrianisation schemes, regardless of the container into which it American and Australian cities are domi- is placed, in this case the capacity of the nated by automobile transport and their road infrastructure which accommodates centres are frequently hostile places, be- it. This conceptual treatment makes traf- ing in some cases almost two-thirds roads fic tantamount to immutable. However, and parking space. extensive evidence is now available that traffic is by no means a fixed given, but The reasons for these differences are ob- rather behaves more like a gas, expanding viously complex, including historical and to fill the space available and compressing cultural factors, but transport planning to cope with reduced road capacity (e.g. practices have played an important role http://articles.latimes.com/1997-05-14/ too. In this paper an historical insight local/me-58478_1_freeways-capacity- is provided into how transport planning traffic (accessed April 5, 2012); Siegel, methods and practices, based on what has 2007). been derogatorily termed a “predict and provide” computer modelling approach There is also evidence now about “peak (which treats traffic as a liquid), have car use” which undermines the whole his- helped to evolve the automobile-depend- torical trajectory of never ending traffic ent city. This is followed by an explanation increases and road building cycles that of the way technical and computer model- traditional transport planning has facili- based approaches to transport historically tated (Newman and Kenworthy, 2011; Mil- have fallen far short in providing the policy lard-Ball and Schipper, 2010; Puentes and direction and vision required in develop- Tomer, 2009). This paper will attempt to ing well-functioning transport systems in summarise the evidence on these matters, cities. thus pointing the way to a more produc- tive approach to transport planning than In many cases these traditional model- blind acceptance of the output of traffic ling shortcomings endure until today. For models using too simplistic assumptions example, the recent UK pronouncements as the basis of their projections and pre- about the expectation of significant new scriptions. This new approach might be traffic growth over the coming decades termed “Debate and Decide”, which inher- (Local Transport Today 592, March 16-29), ently treats traffic as a gas whose volume the Aberdeen Western Peripheral Route is thus compressible given reductions in (http://www.awpr.co.uk/(accessed April road capacity. 5, 2012,) and the Lancaster Northern By- pass, to name just a few current issues Transport models are useful inventions if in the UK alone, have all the hallmarks of used in a strategic and controlled way. So, this now widely discredited predict and rather than ‘throwing the baby out with provide approach. More extensively, the the bath water’, some suggestions are also 6 World Transport Policy and Practice Volume 18.4 Sept 2012 made in this paper about how the tech- jor transportation study. In the US since nical procedures might be better used to 1962, urban areas over 50,000 people provide solutions to growing motorisation have been required to do land use-trans- in cities and result in more sustainable port studies on a regular basis to quali- transport in cities. fy for Federal road funds. These studies were widely acknowledged as being una- 2. The origins and philosophy of com- shamedly highway-oriented (Brown et puter-based transport planning al, 1972). Governments vigorously pro- moted them partly because they were a Perhaps the most important and influen- high profile vote-winning exercise which tial of the technical procedures in trans- appeared to be tackling transport issues port planning is the land use-transport and partly because of the political influ- modelling process, which emerged in the ence brought to bare on Governments by mid 1950’s as a distinct area of study. The the road lobby and a handful of interna- watershed for land use-transport model- tional transport consulting firms with close ling was the publication in 1954 of Mitchell links inside transport bureaucracies the and Rapkins’ “Urban traffic - A function of world over, who very quickly adopted, and land use” which first drew systematic at- to some extent monopolised, the then es- tention to the fact that traffic arises out of oteric technical procedures. There was a land use (Brown et al, 1972). The implica- huge amount of money to be made from tion of this book was therefore that the fu- “grand transportation plans” during the ture of transport in any city is inextricably 1950’s through to the 1970’s, and trans- linked to whether the city is evolving with port consulting firms were only too eager compact, mixed land uses which support to adapt their technical expertise to fit the public transport, walking and cycling or political expectations of the time (witness sprawling in car-dependent patterns that studies with names like SATS – Sydney ensure the car is needed typically for more Area Transport Study; CATS – Chicago than 80% of daily trips. Area Transport Study; MATS – Melbourne Area Transport Study, ad nauseum). The immediate post-war period, as will be discussed in more detail below, was a time Part of the prevailing philosophy during of huge economic optimism and growth this period was that transport planning when the car was seen as the future of was largely seen as a value free, objective urban transport. The conceptual break- science carried out by equally objective through provided by Mitchell and Rapkin, traffic experts, mostly males. Traffic was accompanied by the rapid evolution of viewed as an independent and unavoid- computing power, led to a meteor show- able, indeed immutable, physical phenom- er of multi-million dollar transport stud- enon and there were few questions raised ies in North America, Australia, Europe about the validity or even the desirability and many other countries. The purpose of of attempting to cater for all projected these studies was to plan for anticipated growth. Technical manuals and standards growth in population, jobs and traffic flows for road design to cope with growing traf- as far ahead as 20 to 30 years such that fic were developed apace and the work of there would continually be an equilibrium transport planning was left mostly in the between the supply of transport facilities hands of single discipline technical ana- and demand for travel as it arises out of lysts. land use. As will be shown, this is very similar to the calls in 2012 from the DfT in The fact was, however, that these highly the UK, which see vehicle miles of travel technical and expensive studies were not in England growing by up to 55% over the scientific nor value free, but were strong- next 25 years (see later). ly influenced by their social setting. The 1950’s and early 1960’s were a very op- The concept of the “grand transportation timistic, certain and prosperous period study” was embraced with enormous en- characterised by growth and consumption thusiasm with virtually every developed orientated lifestyles. Car ownership was city at some point between about 1955 booming and the political expectation, at and 1975 undertaking at least one ma- least in the US and Australia, and for a 7 World Transport Policy and Practice Volume 18.4 Sept 2012 time in Europe, was that the automobile their resulting traffic, modelled on the would be the future of urban transport. baseline set of data developed under the Planning was based on a standard of one or previous step; two cars per family. Priority was therefore (4) Analysis of alternative transport net- given to automobiles and road construc- works to cope with predicted travel and; tion, while the pillars of transport planning (5) Evaluation of various alternatives ac- became mobility and speed. Accessibility cording to costs, benefits, impacts and largely took a back seat to mobility, mean- practicality, followed by recommendations. ing in simple terms that a 200 metre walk at 5 km/h to a corner shop for a litre of These tasks, however, are by no means milk became a 5 km drive at 50 km/h to value-free, objective technical procedures a ‘big box’ shopping centre. The approach and there are numerous ways that they was to “construct away” any problems of can be biased to facilitate certain direc- the car, especially the traffic queues and tions. The transportation studies of this increasing number of traffic accidents and early era pioneered large-scale road and deaths through better road systems (Gun- highway planning and in the process public narsson and Leleur, 1989). transport, especially rail, was glossed over and effectively eliminated from considera- Thus right from the outset land use-trans- tion in many cities, (e.g. Denver, Detroit, port studies tended to be strongly asso- Phoenix, Houston and most other US cit- ciated with planning for roads and cars ies). Stopher and Meyburg (1975) show rather than a balance of transport modes, this clearly when they comment about how and most of the US and Australian land public transport was dealt with technically use-transport studies pioneered the build- in the modal split stage of the US studies: ing of elaborate highway and freeway systems (Brown et al, 1972). “The earlier in the process that transit trips could be estimated and removed The first major transportation plans to ap- from further consideration, the more effi- pear were the Chicago and Detroit Area cient would be the resulting highway trav- Transportation Studies, which were very el forecasting procedure.” much along the lines just described. These two studies pioneered the techni- The analysis would then proceed with most cal procedures we know today as the land forecasting based on private transport use-transport modelling process (Black, growth and land use patterns to match. 1981). These technical procedures have Once such low density land use is in place been refined and tuned over the years but public transport is more often than not have evolved into what is generally known based on an inefficient and infrequent bus as the “conventional” land use-transporta- service (often merely demand responsive tion planning study. It has in fact been said buses in very low density areas), walking that there is “a generalised international is reduced to taking the dog around the urban transportation planning process” or block, and bikes are hardly used at all, UTP (Ben Bouanah and Stein, 1978). even by children. Car usage proliferates and a massive increase in road funding is This process can be characterised by the then needed to provide the highway ca- following major tasks (see Black, 1981): pacity for the “grand plan” needs and to (1) Formation of goals and objectives; keep the system from perceived collapse. (2) Inventories of the present situation (mainly Household Travel Surveys (HTS) There are other ways that the large-scale to determine a population’s trip making transportation studies using 4-Step traffic frequency, purposes, modes and other models have played an important role in characteristics), which are then used to facilitating automobile dependence in cit- undertake the four key mathematical ies (Newman and Kenworthy, 1984). For steps - trip generation, trip distribution, example, the modelling process does not modal split and traffic assignment (hence address walking and bicycling in any mean- the other commonly used term: 4-Step ingful way due to the fact that the models Gravity Models); are set up with a set of Origin-Destination (3) Forecasting of new land use plans and Zones (O-D zones) and are designed to 8 World Transport Policy and Practice Volume 18.4 Sept 2012 measure trips from one zone to another. One of the papers reviewed by Atkins Shorter intra-zonal trips, which are pos- sums up the experience with traditional sible by foot and bicycle, are hardly dealt transport modelling in the following rather with, but are placed in a kind of ‘throw damning way: away’ category called “centroid connec- tors” and not modelled effectively. In ad- “It might be said (with due apologies) dition, the smaller scale effects of local of computer-based transportation model- density increases or mixed land use within ling that ‘never before in the history of hu- an O-D zone are not brought out in the man conflict has more money been spent model results. The density of population by more people with less to show for it’” and jobs within each zone is an average (Drake, 1973; 1) and therefore considered to be uniformly distributed. The distribution of that densi- 3. Self-fulfilling prophecies ty within a zone is therefore not modelled properly or reflected in policy conclusions Building large road systems changes the (e.g. high density, mixed use TODs around nature of the city into a more automobile- rail stations, which can greatly alter over- dependent one. In general, modelling has all travel patterns). assumed that land use is “handed down” by land use planners and that transport The shortcomings of these traditional planners are merely shaping the appropri- 4-step models are dealt with in a compre- ate transport system to meet the needs hensive way by Atkins (1986) through a of the land use forecast. This is not the detailed review of over fifty critical studies. case. One of the major reasons why free- Atkins shows clearly how in every respect, ways around the world have failed to cope at least up until that time, there were ma- with demand is that transport infrastruc- jor deficiencies and flaws in conventional ture has a profound feedback effect on transport modelling studies. These prob- land use, encouraging and promoting new lems cover the models’ performance and development wherever the best facilities accuracy, structural deficiencies or speci- are provided (or are planned). Most of the fication errors in the models and amis- major US cities such as Chicago, New York match between the capabilities of the and Detroit, which built extensive freeway models and the purposes for which they systems as proposed by their grand trans- are used. Most importantly, this last as- portation studies, found that the freeways pect shows how they are not good pre- spread land use and generated more and dictive tools, and are of little use in ex- more traffic until very soon after com- amining genuine policy options designed pletion the freeways were already badly to effect change in cities. The specification congested. Sometimes this happened at errors are inherent in the data collection opening because urban sprawl had inten- process, the development of the zones sified to such an extent during the plan- and networks, the trip generation predic- ning and construction phase that the road tions, the trip distribution, modal split and facilities were already out-of-date. Many traffic assignment stages, calibration and studies now refer to these issues under validation and finally in forecasting ability. the rubric of “induced demand” (Goodwin, 1997; Zeibots, 2007; Siegel, 2007) and Such problems can still be found today in the USEPA now requires that environmen- terms of the way problems are framed and tal assessments of the transport emissions the solutions offered. This is because most impacts of new highways and freeways models are still premised more on a sup- formally take into account that more ve- ply-side approach of greater road infra- hicle miles of travel (VMT) are generated structure to solve circulation problems and by these projects than the traffic models other perceived transport inadequacies in typically predict (Newman and Kenworthy, cities, rather than on an approach which, 1999). for example, asks ‘how do we change the existing dependence on cars by providing Once locked into a primarily road-based more sustainable and cost-effective alter- system a momentum develops which is natives?’ very hard to stop. The response to the failure of freeways to cope with traffic con- 9 World Transport Policy and Practice Volume 18.4 Sept 2012 gestion is to suggest that still further roads their analyses were partly vindicated and are urgently needed. The new roads are the patterns of change towards more au- then justified again on technical grounds tomobile orientated growth are particular- in terms of time, fuel and other perceived ly evident in data from 1960 to 1980 (see savings to the community from eliminat- Kenworthy, 1990; Newman and Kenwor- ing the congestion. This sets in motion a thy, 1989). However, the full implications vicious circle or self-fulfilling prophecy of of their prescriptions for cities, particularly congestion, road building, sprawl, conges- in continental Europe, were soon found to tion and more road building. This is not be unacceptable. The rich fabric and ar- only favourable to the vested interests of chitectural history of the cities started to the road lobby and some land developers, be threatened by destructive road plans. but it also builds large and powerful gov- By the end of the 1960’s many city cen- ernment road bureaucracies whose pro- tres were drowning in traffic, inner city fessional actors see their future as contin- residential qualities were being eroded, air gent upon being able to justify large sums pollution and noise were becoming insuf- of money for road building. This commit- ferable in the tightly packed urban layouts ment often translates into direct political and a political momentum was developing activity where policy makers and politi- to do something about “the traffic”. cians are influenced by what is narrow or biased technical advice. In this way road What happened, particularly in West Ger- authorities can become de facto planning many around the late 1960’s, was a deter- agencies directly shaping land use in a city mination not to allow further deterioration and having a large vested interest in road- in the quality of cities and to peg back the based solutions to the transport problem. motorisation trend through a commitment to public transport and pedestrianisation. 4. The European experience This was the period when German cities began to plan and build their U-Bahn In the post-war period European cities be- and S-Bahn systems and create networks gan to increase in automobile ownership of pedestrian streets with festive market and use. The pace of change and impact on places and human scale city spaces. In the old cities became particularly evident Munich the process was accelerated by during the 1960’s as the transport plan- the 1972 Olympic Games. It involved a ning and traffic engineering professions large injection of Federal money into rapid in conjunction with decision-makers set transit and a curtailment of road expen- about attempting to cater for this growth ditures (Hall and Hass-Klau, 1985). Had in automobiles by expanding existing road they been fully implemented, the formu- capacities, building new radial roads into lations of conventional transport planning, the city centre and increasing parking fa- being based more on following established cilities to cope with the influx of vehicles. trends than creating new ones, would Transport planning solutions were clearly probably have seen European cities follow being strongly influenced by the mathe- a similar, but not so extreme pattern, as matically modelled results of the time. The that of the US and Australian cities due to typical 20-year graphs showed increasing land constraints. income and wealth, growing car ownership and use, declining public transport, walk- However, quite widespread popular rejec- ing and bicycling, projections of large road tion of this option in favour of protection and parking capacity increases to prevent and regeneration of the cities into people- the cities drowning in traffic and a sub- orientated places, and strong political and stantial mobilisation of public funds into financial support for public transport saw road infrastructure. the continental European cities at least partially break the self-fulfilling prophecy Transport planners used their land use- of more and more road building and in- transport planning models and overtly ob- tense automobile dependence. Cities in jective forecasting techniques to show an Britain followed the road-based solutions inevitable trend of accelerating motorisa- to a greater extent, as evidenced for ex- tion and how to cope with it. But what hap- ample by the removal of the tram systems pened? To some extent they were right, in all UK cities except Blackpool and the 10 World Transport Policy and Practice Volume 18.4 Sept 2012 relative scarcity of new urban rail sys- prophet to the statistician, the visionary, tems since then (Barry, 1991). It is only goal oriented element and the accompa- comparatively recently that new systems nying moral exhortation have atrophied, have been planned and implemented (e.g. while the analytical, number crunching light rail systems in Manchester, Sheffield, has hypertrophied. (Daly, 1978) Birmingham, Bristol and Edinburgh - see Simpson, 1989). Evidence of the battle fought in Europe between these two different approaches to The solution to the problem continental transport planning is partly seen in com- European cities were facing did not come ments by the Mayor of München and the out of relying on the guidance of technical Mayor of Vienna. transport planning methodologies. It came out of a political process involving a sense “With every million we spend on roads of vision which had to be fought as hard we will be closer to murdering our as anywhere in the world, and it had to city.” Mayor of München (1975) go largely against the mechanistically determined view of the future as prescribed “…unlimited individual mobility … is by conventional transport planning. Mon- an illusion … the future belongs to the heim (1988) relates, for example, the dif- means of public transportation” (and ficulties faced by those wishing to close off this will be) a driving force of city re- streets in the centre of Nürnberg to cre- newal”. Mayor of Vienna (Gratz, 1981) ate a pedestrian network. The transport planners claimed that it could not be done 5. The failure of model-based trans- because of the traffic volumes using the port planning streets. However, it was done and their worst fears did not eventuate because High level failures between 71% and 80% of the traffic sim- In the broad sweep of cities around the ply dissolved in each of the four stages as world, conventional transport planning the pedestrianisation progressed between practices and wisdom cannot claim to have 1972 and 1988 (Museumbrucke and Fleis- left behind a proud legacy. Nowhere is chbrucke – 1972/3; Karolinenstrasse and this more evident than in the traffic chaos Kaiserstrasse 1972/3; Bankgasse and Al- characteristic of many US cities such as derstrasse - 1982; Rathausplatz – 1988: Atlanta, Houston and Los Angeles. A simi- personal communication Rolf Monheim). lar situation is evident in Australia though Nürnberg now has one of the world’s most it hasn’t reached the proportions found in beautiful central cities and an underground the US. After decades of following the ad- railway to service it. vice of practitioners using transport-land use models as one of their basic modus This process which many European cen- operandi, cities have been left with few tres went through of deciding between a apparent solutions to their traffic problems mechanistically determined future for their and in many cases few options other than cities and one which involved a strong de- to endure the traffic chaos on the roads or gree of self-determination is neatly sum- provide token gestures such as HOV lanes. marised by Herman Daly. In referring to This problem which escalated in the US in the self-fulfilling nature of many energy the 1980s was seen partly in the multi- consumption predictions, he also effec- plication of articles and books at the time tively sums up the basic choices involved about congestion, with titles like: Metro- in setting a city’s transport agenda: politan Congestion: Towards a Tolerable Accommodation (Larson, 1988); Resolv- “We can make a collective social decision ing Gridlock in Southern California (Poole, regarding energy use and attempt to plan 1988); “Managing” Suburban Traffic Con- or shape the future under the guidance of gestion: A Strategy for Suburban Mobility moral will; or we can treat it as a prob- (Orski, 1987). Other articles and books on lem in predicting other peoples’ aggregate the subject abounded during this period behaviour and seek to outguess a mecha- (e.g., Cervero, 1986; Pratsch, 1986; Cer- nistically determined future. As the art of vero, 1984; Gleick, 1988). foretelling the future has shifted from the 11 World Transport Policy and Practice Volume 18.4 Sept 2012 The overwhelming impression is of cities A small digression into the issue of new that can only hope to throw palliatives at technologies and fuels for propulsion sys- a problem that has much deeper causes. tems demonstrates this point further. Policies that are frequently suggested such Many still believe that alternative fuels as extended or early work hours, carpools and new types of cars will be the pana- and van pools, computerised traffic lights cea to the problem of “peak oil”, which will or discounts on transit are hardly go- see the world’s supply of conventional oil ing to alleviate or significantly arrest the become increasingly problematic and ex- problem. Technological wizardry was also pensive (Campbell and Laherrere, 1995). sought through the Intelligent Highway To realise how persistent such thinking can System launched in the US in 1995, rep- be, one only has to consider the current resenting a conglomeration of traditional hype over electro-mobility in Europe which highway lobby interests and the IT indus- seems intent on replacing 1 km of petrol/ try to create a system that would essen- diesel driving with 1 km of electrically tially micro-manage traffic flows on the powered driving, as though there are not road system during peak hours in order already very good social, economic and to provide a better relationship between environmental reasons for fundamentally supply and demand (http://www.sti. reducing car use (Kenworthy, 2011). nasa.gov/tto/spinoff1996/36.html). This programme involves research into what Obviously, technological innovation is a causes congestion on a micro-level so that crucial element in progress and problem drivers can relate to the street system in solving and will always be sought, but there a totally interactive way, being told which is a constant need to weigh such innova- routes to take to avoid snares, when to tion against other issues and a more holis- enter freeways from ramps to get the best tic vision of a future society. In the case of run and at what speed to travel (Gleick, transport energy, the reality is that large- 1988). One of the key ideas is that some- scale fuel production from biomass, coal how, if congestion is understood in a more and oil shale has overwhelming econom- detailed way, roads and people can be ma- ic and environmental, as well as climatic nipulated through electronic surveillance and human adaptation problems, which to keep traffic flowing. It is also some- makes widespread use of these alterna- times suggested that California, for exam- tive fossil fuels for transport very unlikely.1 ple, should double-deck all its freeways, New automobiles such as electric cars, al- but perhaps as ‘tollways’, to encourage though dramatically improved technologi- those who can afford it to pay the correct cally from years ago and increasingly pi- economic price for the privilege of moving loted in cities today, are still decades away around the freeway system at peak hour from widespread market penetration, due (Gleick, 1988; Poole, 1988). mainly to the intense capital investment requirements in changing an entire auto- The disturbing part about all these ap- mobile industry, fuel production and dis- proaches is that they are seeking to treat tribution system and technical support only the symptoms of an ailing transport network over to electricity. The frenzied system, albeit in ever more technically so- search for alternative fuels in the 70’s and phisticated ways. This overlooks the root early 80’s after the first two oil crises in causes of the problems, which lie largely 1973 and 1979 and the on-going manifes- in inefficient land use patterns and trans- tation of this technological ‘silver bullet’ port policies that prioritise road capacity approach in electro-mobility today, delay increases over serious transport demand the search for deeper transport and plan- management (TDM), including proper road ning solutions, which will produce a better pricing (Whitelegg, 2011) and provision of overall quality of life in cities. higher quality public transport, cycling and walking infrastructure. The process of de- The above analogy in the field of transport veloping these technological solutions can energy should not be lost for its relevance create an unreal expectation that technol- to the central argument of this paper. ogy alone will solve the problems of the Transport planning as a whole seems slow city. to respond to the new imperatives in cities and the failures of the past. There is a 12 World Transport Policy and Practice Volume 18.4 Sept 2012 tendency to hang onto entrenched beliefs, Recent UK government pronouncements which have been shown to be false. For also herald a continuation of a “predict example, roads are still partly justified on and provide” approach, with a staggering the basis of simple cost-benefit analyses estimate of between 34% to 55% more involving savings in fuel, time and some- vehicle miles of travel in England between times emissions, time being the key item, 2010 and 2035 and rejection of the “peak which usually constitutes 70% to 80% car use” hypothesis (Local Transport To- of all monetized economic benefits. This day 592: http://www.transportxtra. occurs notwithstanding the widely docu- com/magazines/local_transport_today/ mented Marchetti Constant of a 65 to 70 news/?iid=467, accessed April 11, 2012). minute overall travel time budget in cities through the millennia regardless of the As if all of the foregoing evidence were not dominant transport mode and showing sufficient evidence of the inability of free- that time savings due to speed increases ways to really provide solutions to traffic in fact do not occur, but are rather just problems, we have further evidence of a used to travel further (Marchetti, 1994; reverse process of pulling freeways down Newman and Kenworthy, 2006). being highly successful in actually reducing traffic. This evidence supports one of It is very clear that from an urban sys- the key high-level as well as the technical tems perspective the analyses are wrong failures of transport planning and mode- and that fuel, time and emissions are re- ling: the failure to recognise traffic as be- ally costs of new urban road projects in having fundamentally like a gas and not a cities particularly in those already highly liquid. dependent on cars (Newman and Kenwor- thy, 1984 and 1988; Newman, Kenwor- Bringing together various sources, there thy and Lyons, 1989). The fact that major are innumerable historical and more renew roads are sometimes still touted as cent examples of ‘trip-degeneration’, as solutions to congestion seems to suggest it was termed by the late John Roberts something of an inability to learn from (TEST, 1992): past events. Almost fifty years of experience has demonstrated the futility of • By 1998 there were already at least sixty building more extensive road systems to documented cases worldwide where roads relieve congestion and the environmental were closed or traffic capacity was reduced implications for the city of that approach and 20% to 60% of traffic disappeared. are dramatic and widespread. For exam- • Tower Bridge, London closed in 1994 due ple, the push for radical solutions to trans- to structural problems: after 3 years traffic port problems in Los Angeles in the 1980s was still not back to original levels. came from government environmental or- • Part of London’s ring road, the Ring of ganisations faced with some of the world’s Steel was closed in 1993: Traffic fell by worst air pollution which was estimated 40%. to cost the community at that time some • London’s Hammersmith Bridge (30,000 $US9.4 billion per year (The West Austral- vehicles per day) was closed to all traf- ian September 20, 1989 p80/1). fic except buses and cyclists due to structural problems. A survey of commuters And yet the ‘roading’ approach still per- a few days before closure and the same sists. Bremen, a city in northern Germany people after showed 21% no longer drove renowned for its progress in car sharing to work. They switched to transit, walking and non-motorised mode use, still strug- and cycling and congestion in surrounding gles with the issue of a major bypass (see areas did not markedly increase. WTPP 18.1/18.2 editorial), as does Aber- • West Side Highway, New York City: deen in Scotland with its extremely expen- 1973 one section collapsed and most of sive Aberdeen Western Peripheral Route. the route was closed. A 1976 study of the Lancaster’s Northern Bypass, which cost remaining portion based on traffic counts £130 million for 4.5kms or £29 million per three years prior to closure and two years kilometer, is another recent example of ex- after showed 53% of trips disappeared cursions into this less than fruitful, indeed and of those trips, 93% did not reappear destructive transport strategy approach. elsewhere. 13 World Transport Policy and Practice Volume 18.4 Sept 2012 • In 1989 an Earthquake destroyed the April 10). Embarcadero Freeway in San Francisco and it was not rebuilt and the predicted These projects are on top of the almost chaos never materialised. The whole wa- fifty years of successful pedestrianisation terfront area of San Francisco was revital- schemes in European cities such as Mu- ised. nich, Copenhagen, Köln and so on, which • In 1996 San Francisco’s Central Freeway have also showed that significant amounts upper deck was torn down due to instabil- of traffic simply disappear following road ity following earthquake damage and traf- closures and the cities become more liv- fic chaos did not materialise. able and sustainable (e.g. see earlier ex- • Melbourne, Swanston Street Transit ample from Nürnberg). Mall: Street carrying 30,000 vehicles per day closed to regular traffic. Traffic chaos The extent to which transport planning in surrounding streets was vehemently has lost its way since the Second World projected during the lead up period. After War and still so often fails to provide de- closure there was no chaos. Some increas- cision-makers with the answers to urban es in volumes were found on surrounding transport problems, is partly seen in the streets, but it was well within the capacity following statement from the late 1980s of the street to handle it. reviewing Transportation Planning in a • Portland: Harbor Drive Freeway along Changing World (Nijkamp and Reichman, the Willamette River waterfront was closed 1987). Excerpts from the review read as in 1974 and then removed and a linear follows: was park created (Tom McCall Waterfront Park). Traffic chaos did mot materialise, “This book clearly illustrates the con- but the whole downtown was revitalised fused and contradictory world of trans- with LRT, people places and markets and portation planning. It is a collection of sound urban design for pedestrians and papers from a series of three inter- transit users. national workshops on transportation • Seoul, South Korea: The Cheonggye ele- sponsored by the European Science vated expressway of 6 km in length running Foundation, and has no solid context through central Seoul was torn down be- or clear message that could be utilised tween 2003 and 2005, along with Cheong- to improve the world of transporta- gye Road beneath it, together carrying tion planning…the book fails to get to 168,000 vehicles per day, and no traffic grips with the enormous failures of chaos ensued. In fact the traffic engineer technical and model-based transpor- interviewed in the documentary film made tation planning ideology and is very about the project reveals that the overall thin on societal and cultural impacts… average traffic speed in the City of Seoul There is no clear discussion of the … actually rose by 1.2 km/h, contrary to the multiple implications of increasingly more normal expectation of gridlock. This higher levels of motorisation. Walk- project has led to a more general “road ing and cycling and road safety and diet” approach in Seoul, emphasising new urban design do not figure in this vol- bus lanes, improved subway operations ume, and these are major issues of and more walkable environments. The city importance to transportation policy in is being greened (Schiller et al, 2010). the age of car dominance…The book is part of the problem it would claim (Sources: Surface Transportation Policy to be examining…contributing to the Project, March 1998 issue of Progress; mass of ‘received wisdom’ which de- Seattle Urban Mobility Plan: Case Stud- lays innovation, social awareness and ies in Urban Freeway Removal (found at: genuinely critical transportation policy http://www.cityofseattle.net/transpor- analysis.” (Whitelegg, 1988) tation/docs/ump/06%20SEATTLE%20 Case%20studies%20in%20urban%20 There is a sense that a significant part of freeway%20removal.pdf, accessed April the malaise into which transport planning 10, 2012; Siegel, 2007; http://www.pbs. has fallen is a preoccupation with the nar- org/e2/episodes/310_seoul_the_stream_ row cost-benefit analyses for new roads, of_consciousness_trailer.html. accessed which form the ‘punch line’ of the detailed 14 World Transport Policy and Practice Volume 18.4 Sept 2012 technical and mathematical modelling ap- sions, beloved traditions, human will, proaches. Such approaches go deeper into and it is these which are really impor- the microscopic aspects of, for example, tant. While cost-benefit analysis and travel behaviour or modal choice, without its many successors have been dis- much of an appreciation for their broader credited, basic understanding of the context or how the work can be effectively destructive effects of relying primarily applied to produce tangibly better cities on numbers in the formation of pub- for people. There is a danger that a lot of lic policy still has not penetrated the transport planning modelling and predic- consciousness of the planning profes- tions of the future are done without much sion in the United States, and that of a feeling for what might be called the profession is gradually committing “soul” of the city and what will need to be suicide in consequence, persisting in done in practical terms if it is to become the delusion that it is a science which a better place to live. It seems that there it never was and never can be” (Ba- is little point in becoming increasingly incon, 1988; 2) volved with “tools” and “means” if there is no clear direction or leadership from within the discipline as a whole about how the This tendency to see transportation some- transport planner should really be trying what in isolation from the broader prob- to contribute to the betterment of the city. lems and issues in the city and a general lack of vision, leads to a growing isolation Edmund Bacon, the famous American ur- between transport planners and decision ban planner and architect from Philadelphia makers. From the technical viewpoint and outspoken critic of automobile-based it is safer for many to stay more or less planning, provides a useful comment on within the framework and methods de- the consequences of overzealous and nar- veloped within the period of the “grand row economic number crunching in the plans”, which basically view increasing formulation of urban transport policies: motorisation as almost inevitable and the ‘rightness’ of more roads as a God-given “The sad thing is how often the plan- truth, than to provide policy direction and ners in the United States seize mind- vision–a contribution which might help cit- lessly upon the latest fashionable ies find a way of breaking the automobile planning gimmick. The cost-benefit planning treadmill. ratio was one of the first of these, a” scientific” method for determin- Technical failures ing where a highway should be placed by adding up the costs of alternative Even when judged on the basis of whether highway routes and comparing these the technical procedures are producing ac- with a quantification of the value curate predictions of, for example, future in dollars of the time saved by the traffic flows and relationships between highway user. This was adopted uni- public and private transport, the over- versally as the only right way to do whelming weight of evidence has been on things until its continued use imposed the negative side (Atkins, 1986). It is one such outrageous consequences that it thing for a transport-land use model to ac- dawned on someone that saving the curately reproduce the present situation. highway users a few seconds of time It is quite another for it to accurately re- would be less socially and economical- flect what may happen in the future under ly desirable than destroying irreplace- a complex array of urban pressures and able landscapes, historical sections forces, or what the result might be where of cities, coherent neighbourhoods, a city is given a glimpse of a future quite or networks of human relationships. different to what exists today. There is a Underlying it all was the failure to re- danger that the modelling process is so alize that the development of policy shaped by existing patterns that it is un- through the manipulation of numbers able to respond correctly or creatively to is always bound to be wrong because significantly different circumstances (e.g. numbers by definition leave out the markedly higher localised TOD densities) unquantifiable variables: Human pas- or to meaningfully incorporate significant 15 World Transport Policy and Practice Volume 18.4 Sept 2012 factors outside its usual outlook (complex significant break in them. Instead- Port climate change, social changes or qualita- landers built a new light rail line (MAX, tive changes in the city environment that opened in 1986) in place of a freeway and demand new approaches). Conventional embraced it and their revitalised central transport modelling is simply too geared city in a way that was not predicted by to extrapolating and magnifying existing any transport studies. The story of how patterns to be of significant use in guiding this change occurred so that conventional cities towards an alternative future. The transport plans were rejected and a new future being demanded today in all cities is vision enacted is a valuable case study for a low carbon, regenerative approach to (Edner and Arrington, 1985). the urban environment, which at the same time delivers a high quality of life. Other cities too have shown surprisingly rapid land use adaptation and success In particular, the techniques of transport with their new rail systems. For example, planning are not well-suited to predicting Washington DC and in particular some Ca- human responses to qualitative changes in nadian cities such as Vancouver have nu- the character of a city or the way people merous examples of integration between may respond to a new transport option. new rail systems and high density, mixed The models may suggest little if any re- use development (Newman and Kenwor- sponse and yet the changes may be quite thy, 1999; Schiller et al, 2010). Los An- rapid and marked. For example, a city may geles’ rail development was linked in its make a major effort to humanise its cen- planning phases to proposed major new tral city through urban design improve- commercial and mixed-use developments ments, city art and festive market places, (Keefer, 1986) and this has occurred both pedestrianisation schemes, and other traf- around its new Metro stations on the Red fic limitation strategies. At the same time and Purple lines (e.g. at Wilshire/Vermont it may decide to install or upgrade rail and Hollywood and Vine) and around the services with the result that people may light rail system (e.g. at Del Mar station on discover new ways of experiencing their the Gold line). city. This can begin to set new relationships between transport and land use in- The retreat into ever-increasing sophisti- cluding reductions in parking, greater de- cation and microscopic detail in the tech- mand for central and inner city housing, niques of transport planning without a joint development of high density, mixed clear vision, aim or goal, seems to lead use TOD complexes around stations, bet- to an increasingly deadened sense of pur- ter pedestrian and bicycle links and facili- pose within the profession and an inability ties and still more public transport. Travel to provide policy makers with sound guid- behaviour and housing options can change ance. Decision-makers, who must cope quite rapidly under these circumstances, with an increasingly complex set of de- in ways that transport modelling does not mands related to local, regional, national anticipate or incorporate well. Certainly, it and global sustainability needs, provision would be most unlikely that a convention- of more diverse housing options, social al transport planning study would recom- and community needs and depressed fi- mend that such changes be made to a city. nancial situations, often find many of the prescriptions from traditional transport To a great extent all the changes just men- analyses blinkered and unworkable. tioned have happened in Portland, Oregon over the last 25 to 30 years, at least in In summary, the technical world of trans- the central and inner city areas. Twenty to port planning finds it difficult to get be- thirty years is a rather typical time horizon yond a view that the city’s future can be for a transport planning study. Portland’s predicted and provided for by mathemati- transportation plans of the 1970s were for cal equations based on often-debatable more freeways, which would have added transport economic and behavioural the- more traffic pressure to downtown - Port ories. Relative transport costs, resource land and made it a less hospitable place. efficiency measures, narrow cost-benefit Transport patterns would have been a con- analyses and other abstractions from the tinuation of existing trends rather than a world of transport modelling are not of 16 World Transport Policy and Practice Volume 18.4 Sept 2012 themselves adequate to the task of guid- to leave the inner city if they are still ing decision making and fulfilling diverse there, and the stronger will become community expectations about the future the desire to live just there, leading of a city and its quality of life. Without a a modern life in an old town full of wider social and environmental as well as atmosphere”(p. 36). broader economic context, transport planning often loses sight of other important With the benefit of hindsight, reurbanisa- forces and is always at risk of working in tion is in fact what many cities have been a vacuum, producing answers that are of embracing since that time. In the case of little or no use to politicians, the commu- Australian cities, this started as “urban nity and business leaders, or the long term consolidation”, the US cities have their sustainability and liveability of cities. “smart growth” programmes and count- less European cities have been and still are 6. Planning with vision regenerating former industrial and port areas into vibrant new communities (e.g. It is very rare to find a transport planning Hafen City, Hamburg to name just one). treatise which makes a clear statement about the broader intent of the work, a Experience in Perth, Western Australia statement which sets a clear human context or vision and gives the mathematical Perth’s urban history strengthens the view and modelling work substance, direction that the seemingly objective and techni- and meaning. Transport and Reurbanisa- cally “correct” or safe view from the world tion (Klaassen et al, 1981), historically of transport planning fails to respond to was one of the first works to break this the realities and needs of the changing mould. While clearly translating its trans- city. For years Perth had numerous transport prescriptions for cities into mathemat- port studies, performed using methods ical modelling terms, transport was clearly and techniques respectable in the best of directed towards encouraging a process transport planning circles. They all pre- of “reurbanisation” which the authors saw dicted growing car orientation, declining as crucial to the total life and meaning of public transport and a need for bigger and the city. They make a very clear statement better roads. And this is what was experi- early in the work that sets a human con- enced, certainly until the late 1970’s and text for their transport planning expertise early 1980’s. and makes the book readable and mean- ingful. During the late 1970’s and into the 80’s, there was a sense in some quarters that They describe ‘reurbanisation’, their ulti- the city was being cast along a path of self- mate goal, in the following way: fulfilling prophecies. The old diesel suburban rail lines were being systematically re- “The process thus set going is one of moved or run down (the 19 km Fremantle once more turning degenerated urban line, one of only three suburban rail lines, patches into city quarters with living was closed to secure land for a freeway) cores, fulfilling a real economic, social and new land for more freeways was being and cultural function, a process of re- reserved on increasingly tenuous grounds urbanisation. Its ultimate fascinating for roads with no known date of construc- objective is the revival of the old core tion exposing government to enormous cities, fascinating to many individuals claims for injurious affection and result- who have learnt in hard practice that ing in the Road Reserves Review (Govern- living near to nature means mowing ment of Western Australia, 1991). Led by the lawn every week, driving down- a civil society action group called “Friends town in long queues every morn- of The Railways” established in early 1979 ing and driving out of town in long to oppose the Fremantle rail line closure, queues in the evening; that a subur- a change in political direction provided a ban home means buying a second car window of opportunity to redress Perth’s for their wives so that they may flee lopsided transport planning and this be- the periphery, etc. The more people gan a fundamental and far-reaching mo- realise all this, the less they will want mentum in favour of public transport and 17 World Transport Policy and Practice Volume 18.4 Sept 2012 more compact urban development pat- A rapid transit study was conducted in the terns, which could help reshape the car 1980s which concluded that an O-Bahn dependent urban environment. A peti- busway or a train service could be intro- tion with 110,000 signatures against the duced in the middle of the freeway - an Fremantle line closure was presented to option made possible at the insistence of the State Parliament in 1979, at a time a former Labor Premier of WA in the ear- when the population of Perth was about ly 1970s, not by transport planners who, 860,000. Real pressures for such changes on the contrary, had in previous decades had been creeping up continually through removed two rail reserves from the met- the 1970’s, however it wasn’t until the ropolitan planning scheme. The transport mid-80’s that a real consensus started to consultants involved the community in develop at the political level that some- the assessment of the options and a clear thing needed to be done to balance Perth’s preference for rail was given. The consult- urban sprawl and road orientation. ants however concluded, with encourage- ment from Perth’s transport planners, that Little help was provided from the urban a busway was preferred on cost grounds. transport planning fraternity who contin- A further study was done which evaluat- ued to provide road-oriented solutions and ed this report and brought in more of the unimaginative public transport studies, land use options created by rail, which a which invariably favoured busways over busway does not provide (as well as high- railways. At the political level, amongst lighting a technical issue about where all some legacy-driven politicians, a vision the buses entering the central city from began to develop of the way the city need- the busway would actually end up). This ed to change and how this might occur. new report preferred rail and showed it to This constituted a realisation that the only be a viable economic option, so a govern- way the city was going to generate inno- ment decision was made to build the first vative initiatives was to seriously question major extension of a suburban rail system the car and bus-based solutions. This re- in modern Australian history (Newman, quired embracing ideas from community, 1992). academic and professional sources whose experience and perspective outside tradi- A $400m commitment to electrifying the tional transport planning suggested that three old rail lines and building a new line other directions were possible (Newman, to the northern suburbs, was the biggest 2011). single capital investment by the State Aus- tralian Labor Party government in its term After the reintroduction of the old Freman- of office during that time. The popular- tle diesel rail service in 1983 and during ity of the decision and the shifts in think- the subsequent electrification of the whole ing about land use since then (containing rail system from 1988 to 1991, achieve- urban sprawl, focusing on development ments that involved the community and near rail stations etc), have confirmed the politicians being forced to lead their trans- game-changing nature of this decision. port professionals into a more balanced and visionary approach, the next focus In order to maximise the potential of the was on the sprawling northern suburbs. connection between land use and rail serv- The northern corridor has taken a huge ices in the way Toronto, Vancouver, Singa- share of Perth’s urban growth over many pore and other cities have done, it became decades, based firmly on the automobile necessary for Perth to gain experience in and low-density suburbs. A freeway was areas such as joint public-private devel- provided in the centre of the corridor, but opment and value capture around railway after great popularity in the beginning, it stations. These approaches have been rapidly became congested causing many common practice in many cities for some people in the late 1980s to demand better years, particularly those that have installed transport options for the corridor, the most new rail systems (e.g. Los Angeles’ Spe- popular of which was a rail line (Newman, cial Benefit Assessment Districts around 1992). rail stations). This shift to a rail orientation set a new context for public transport planning in Perth and offered the opportu- 18 World Transport Policy and Practice Volume 18.4 Sept 2012 nity to better link land use planning to the age has exploded six-fold from 1988 to public transport system in projects, such 2011. In Adelaide, which had a similar and as Perth’s “Subi Centro” TOD at Subiaco comparable suburban diesel rail service railway station a few kilometres west of to Perth in the 1980s (and still has, apart the CBD. This is something that would not from an extension of its one tram line), have happened while road and bus plan- but has experienced no politically or com- ning predominated. munity-led rail revival until now, rail use It has now become accepted that it is not has stagnated. sufficient for the Government to just build Experience in Portland, Oregon and operate a new rail system. Rather, much of the community, the development Similar experiences occurred in Portland, industry as well as the public transport op- Oregon through the 1970’s and 1980’s. erator, acknowledge that they can all ben- The long fight in Portland away from the efit together by ensuring that the potential early technical transport studies and their of the rail system to help reshape the city overwhelming road orientation is summa- is fully realised. These new forces are now rised by Edner and Arrington (1985) in the changing the nature of the city, incremen- following way: tally and in small, slow steps, away from more car dependence and dispersed land “…initial political stirrings for a transit use, into a more focused and transit-ori- option were substantially unsupport- ented city. ed by comprehensive technical studies. The thrust was to wean Portland Figure 1 shows the growth in total rail us- away from a highway-based system age in Perth and Adelaide over the last 23 and buy time to develop a balanced years. In Perth, where improvements and alternative using transit and limited extension of the system has occurred, us- highway improvements.” (p. 14)

Figure 1. Perth and Adelaide rail use from 1988 to 2011. Source: Constructed from public transport operator data from Perth and Adelaide 19 World Transport Policy and Practice Volume 18.4 Sept 2012 This involved doing some new technical In both Perth and Portland transport plan- studies with an emphasis on transit, but ning was really forced through a political this time under the auspices of Oregon’s process, especially involving the commu- Governor who was convinced that change nity, to take on new directions. There was was needed. little evidence that the transport profession was about to provide the initiatives “The key question facing the Gover- itself. nor’s Task Force (GTF) was whether transit was a viable alternative to 7. Transport planning–the crucial link freeway investment. Sixty-eight sys- to land use tem configurations for the region were ultimately evaluated. These configu- Ultimately the transport problem is a land rations were identified as alternatives use problem. Porter (1987) in discussing to PVMATS (Portland-Vancouver Met- America’s rapid slide into national gridlock ropolitan Area Transportation Study) in the 1980s stated that: and its highway emphasis. That study, initiated in 1959 but not formally “Most communities are trying to over- adopted until 1971, assumed that come the traffic crisis in ways that transit ridership and operation would actually perpetuate it. Most projects stabilize and, at worst, continue a being planned and developed in fast trend of decline into the future.” (p. growing areas build in automobile de- 14) pendency, which leads to congested arteries which results in cries to re- Like nearly every ‘grand plan’ from the duce densities of development, which 1950’s, Portland’s transportation study in turn creates greater dependency on was roads-orientated and virtually as- automobiles.” (p. 34) sumed the demise of public transport. However, the GTF’s report provided the Pucher (1988) described and compared first technical justification for transit based the success of urban public transport sub- on a range of factors including positive en- sidies in the US in relation to other coun- vironmental effects, but more importantly tries and found that in virtually no other according to Edner and Arrington: country have subsidies been as ineffective as they have been in the US. His final con- “…it set the stage for developing the clusion makes the point that without: technical and political decision-making capability for regional transit plan- “…policies to increase the cost of auto ning. The GTF report was a crucial el- travel and to promote a more com- ement in the decision to withdraw the pact land-use pattern, it seems un- Mt Hood Freeway…This technical jus- likely that any significant changes will tification initiated a linked technical/ be possible in the US urban transpor- political decision-making process… tation system.” (p. 402) Freeways were de-emphasized to the benefit of transit and a CBD focus.” It would still appear today that if trans- (p. 15) port planning is to provide any clear policy guidance to decision-makers confronted They summarise the decision to build a rail with how to respond to car dependency line instead of a freeway as: and congestion, the planners’ approach must incorporate some more radical vi- “…a major shift in the functional and sions of compact land use patterns both in philosophic role of transit in the re- developing areas and through selective in- gion… (which) …ruptured the political fill and redevelopment in older areas. The fabric of transportation decision-mak- way technical transport studies are con- ing, realigning the roles and responsi- ceived and the policies and prescriptions bilities of many political and technical that result from them, can then be geared actors.” (p. 2) towards achieving those visions. Even if some of the solutions seem unachievable within prevailing social, economic and po- 20 World Transport Policy and Practice Volume 18.4 Sept 2012 litical realities, the fact that an attempt 8. Making transport planning a better has been made to provide an alternative tool in reshaping the auto-city vision is a forward step, which develops momentum. In particular, it can begin to There is nothing inherent in the actual give substance to the idea that a better techniques of the land use-transport mod- balance between cars and other modes is elling process or the other technical pro- possible. Once started, this momentum cedures of transport planning which will can gradually filter its way through com- inevitably produce road-biased results. It munities, bureaucracies and the political is more the way decisions are made about arena and finally into concrete change, as how to use the techniques. Historically has been the case in Portland and Perth. road planners have dominated this exercise. If a genuine attempt is made to It is interesting that in Portland the proc- consider alternatives to urban sprawl and ess of planning and building MAX was more freeways, which is accompanied by characterised by a lot of cynicism particu- a community engagement process, such larly in the press. A competition was run as Perth’s Dialogue With The City (see to find the best name for the system. A Schiller et al, 2010), it is possible to give Street Car Named Expire finally won. MAX new directions to the transport planning however is a big success, both as a trans- process. It is also possible to build in more port system and as a focal point for new sophisticated feedback mechanisms where development. Howard (1988) reported transport and land use are dealt with in an shortly after the opening of the first line iterative manner, one progressively affect- in 1986, that benefit assessment districts ing the other. This would in all likelihood established in downtown to return to the be an improvement over existing practic- community some of the private land val- es, though the results of such models are ue increases of the system “ have proven still subject to considerable debate and in- very successful, partly because the sys- consistency and are by no means guaran- tem is so appealing to the public that the teed to come up with prescriptions, which development community is jumping on the will assist cities in transitioning to lower bandwagon to expand the scope of that car dependence (Webster, Bly and Paulley, program” (p172). Having made the big 1988). break and built a railway instead of a freeway, the next steps were easier. The only Ultimately it comes back to the first stage arguments in Portland about MAX became of the process - the formation of goals and who should get the next extension. objectives. In the past, and unfortunately still today in many places, land use-trans- Circumstances also change, which make port modelling has chased something of a what is achievable a constantly changing fairy-tale world where transport demand thing. For example, in Perth 27 years ago and supply are meant to be kept in equi- it would not have been conceivable that librium by planning road systems to cope the city would now have electrified three with projected traffic volumes–a sort of existing rail lines, constructed a 33 km line “transport utopia”. The pressing require- to the north, a 71 km line to the south ment was, and often still is, to try to keep and be actively pursuing the develop- ahead of congestion. Interestingly, even ment of a LRT system and further heavy after decades of experience in US and UK rail extensions and new lines. Nor would it cities of building freeways while conges- have been thinkable that all these changes tion relief remained an elusive goal, large would also be allied to efforts at rail sta- new roads are still today called for and tions to provide a focus for land use devel- justified for their ability to relieve conges- opment. These changes did not come out tion. Fundamentally, in too many places of traditional transport planning analyses, a genuine alternative, such as minimis- but there is a chance that they could have, ing unnecessary private motorised travel, given the right issues, contexts and goals has not really been embraced. Too often, to shape the transport planning process. walking and bicycling are not considered in any serious way, while public transport is often still seen as an addendum to the main game of catering for private mobility. 21 World Transport Policy and Practice Volume 18.4 Sept 2012 In many cases there is very little mean- judgements… transportation models ingful variation in the different scenarios cannot directly give answers to policy provided by transport studies – just minor questions, nor can they derive trans- variations on a fundamentally road-orient- portation system alternatives. Final ed theme. decisions cannot be reduced to a set of mathematical equations.” (p II-1). A change to more comprehensive planning requires better specification of the goals 9. Conclusion and objectives of transport studies and more emphasis on public transport in the The key to making transport planning a models. Since the 1980s there has been better contributor to policy development a rapid growth in new light rail systems does not lie in giving what amounts to throughout the USA and Canada (com- open-ended briefs, such as asking com- mencing in Edmonton in 1978 and then puter-modelling studies to predict traffic Calgary and San Diego in 1981). These 20 years into the future and how many new systems have been introduced largely new roads are going to be necessary to as broader community or politically led ini- cope with it. History shows that under this tiatives in response to a dire need for mo- kind of direction it is almost assured that bility alternatives, rather than as techni- within the logic of the models’ own analy- cal decisions from conventional transport sis enough traffic will be foreseen to jus- analyses. This political and community- tify any number of new roads. This is how driven response reflects widespread dis- the grand transportation studies of the enchantment with the problems created 1950’s, 60’s and 70’s universally recom- by automobile dependence such as con- mended elaborate freeway networks and gestion, local, regional and global environ- presided over a massive decline in urban mental impacts and social and economic public transport. inequity in transport. Rather than promoting greater freedom, unbridled personal Even asking traffic models to assess mobility derived from mass prosperity and whether traffic projections alone justify automobile use, have created high levels the construction now of a particular road, of individual frustration where a car is the tends to take the decision out of context only alternative for most trips (Eno Foun- with other important values, visions and dation, 1988). directions in the city, which are often in direct opposition to the idea of building Urban and transport planners everywhere more road capacity (e.g. the aforemen- can assert their role in the development tioned conflict in Bremen between build- of cities through new goals and objec- ing a new ring road and all its sustainable tives for transport-land use modelling transport achievements to date). The in- based around balancing the roles of vari- ternal workings of traffic models tend to ous modes of transport, minimising total generate self-fulfilling prophecies of traffic motorised travel in the urban system and and future road needs without considera- reducing the costs of urban land develop- tion of the broader implications. Such an ment through reducing urban sprawl – in approach is an invitation to perpetuate or short creating more sustainable and in- strengthen dependence on the car and a deed regenerative cities. There is no com- sure way of generating a sense of power- pulsive reason why transport planning lessness within public planning about in- should favour roads and suburban sprawl fluencing the future land use and transport to the exclusion of other modes and more directions of any city. compact patterns of development. As stated by the last major Australian urban What appears to be needed is a strong vi- “grand plan”, the Sydney Area Transporta- sion of what overall directions are most de- tion Study (SATS, 1974) concerning land sirable for the city and then, if necessary, use-transport modelling: to seek guidance from transport modelling about how to get there. Questions about “Some of the inputs into the models the need for new roads should not be put are based on assumptions of a politi- in terms of “Is the new road justified on cal nature or those containing value traffic grounds?” Based on historical- ex 22 World Transport Policy and Practice Volume 18.4 Sept 2012 perience, traffic models are very likely to • What goals could be set for extending conclude that it is. Rather we should be or introducing new rail systems to help re- asking questions such as: duce car dependence?

• Is the new road in keeping with the vi- Working within this type of visionary sion of the future city and its sustainability framework, transport planning can make and will building it contribute to or detract a constructive contribution to urban policy from this? development. The last thing that is needed is self-fulfilling road prophecies from com- • Is it desirable to see the projected traffic puter models telling communities what the growth fulfilled or should planning policy city will look like and what they are going actively seek to prevent or modify, rather to have to do to cope with such scenari- than facilitate that growth? os. In simple terms, “predict and provide” planning, which treats traffic as a kind of • What are the alternative options for a immutable liquid that will simply flow over city in assessing a road proposal and how everything if not catered for, needs to be do these alternatives relate to a range of replaced with a “debate and decide” ap- other objectives (e.g. environmental pro- proach which treats traffic as an expand- tection, urban regeneration, reduced re- able or compressible gas and allows cities source and energy consumption, livabil- to shape a regenerative future for them- ity of neighbourhoods, accessibility for all selves based on a decline in automobile groups in the community, transport safety dependence. etc)? Author details: Collectively we need to be able to say what we would like the city to look like in Jeff Kenworthy 20 years time: Curtin University Sustainability Policy In- stitute (CUSP) • What sort of improvements in the urban Curtin University of Technology environment do we want? Perth, Western Australia

• What qualities and diversity do we want [email protected] in urban lifestyles? (e.g. communal spaces where once only streets existed; mixed 10. References land uses as opposed to rigid zoning; more mixture in dwelling styles and density, Atkins, S.T. (1986) Transportation plan- greater sociability versus increased priva- ning models - what the papers say. Traffic tism). Engineering and Control, September, 460- 467. • What do we want the central city and sub-centres to look like? Bacon, E.N. (1988) Planning and planners in the post-petroleum age: Fundamental • Do we want strong centres and what issues facing metropolitan development should the balance be in modes of access and conservation. Regional Development to and within these centres? Dialogue 9 (3), 1-6.

• What kind of overall population and job Barry, M. (1991) Through the cities: The densities should the city aim for to mini- revolution in light rail. Frankfort Press, mise car dependence and where should Dublin, Ireland (pp 57-81). higher densities be concentrated? Ben Bouanah, J. and Stein, M.M. (1978) • What goals could be set in terms of re- Urban transportation models: A general- ducing car dependence? (e.g. parking lev- ized process for international application. els in the CBD and other centres, modal Traffic Quarterly, 32, 449-470. split for various types of trips including goals for walking and cycling and targets Black, J. (1981) Urban transport planning. for reducing overall per capita car use). Croom Helm, London, 221 pp. 23 World Transport Policy and Practice Volume 18.4 Sept 2012 Cervero, R. (1986) Suburban gridlock. Goodwin, P. (1997) Solving congestion Centre for Urban Policy Research, Rutgers (when we must not build roads, increase University, New Jersey, 188pp. spending, lose votes, damage the economy, harm the environment and will never Brown, H.J., Ginn, J.R., and James, F.J. find equilibrium. Inaugural Lecture for The (1972) Land use-transportation planning Professorship of Transport Policy, Univer- studies. In: Brown, H.J., Ginn, J.R., and sity College London, October 23, 2012. James, F.J., Kain, J.F. and Strazsheim, M.R. (http://www2.cege.ucl.ac.uk/cts/tsu/pb- (1972) Empirical models of urban land ginau.htm, accessed, April 5, 2012). use: Suggestions on research objectives and organisation. National Bureau of Eco- Gratz, L. (1981) The Vienna Underground nomic Research, Washington DC, pp 6-16 construction. Stadtbaudirektion, Wien, (available: http://www.nber.org/chapters/ February. c3975.pdf, accessed April 5, 2012). Gunnarsson, O. and Leleur, S. (1989) Campbell, C. J. and Laherrere, J. H. (1995) Trends in urban transport planning - The The World’s Oil Supply 1930-2050. Petro- current shift in solving transport problems. consultants, Geneva. Prospect 2, 2-6 (International Federation of Housing and Planning). Cervero, R. (1984) Managing the traffic impacts of suburban office growth. Trans- Hall, P. and Hass-Klau, C. (1985) Can rail portation Quarterly 38(4), 533-550. save the city?: The impacts of rail transit and pedestrianisation on British and Ger- Daly, H. (1978) On thinking about energy man cities. Gower, England, 241pp. in the future. Natural Resources Forum 3, 19-26. Howard, J. (1988) Value capture and benefit sharing for public transit systems. In: Drake, J.W. (1973) The administration of Attoe, W. (ed) Transit, Land Use and Urban transportation modelling projects. D. C. Form. Center for the Study of American Heath, Lexington, Massachussetts. Architecture, School of Architecture, Uni- versity of Texas at Austin, 171-178. Edner, S.M. and Arrington, G.B., Jr. (1985) Urban decision making for transportation Keefer, L E (1985) Joint development at investment: Portland’s light rail transit transit stations in the United States. Trans- line. US Department of Transportation, portation 12, 333-342. Technology Sharing Program, Report Nº DOT-I-85-03, US Government Printing of- Kenworthy, J.R. (1990) Insights into the fice, Washington DC. growth of automobile dependence from 1960 to 1980 in thirty-two international Eno Foundation (1988) Report of the 20th cities. NERDDC Project No 1050 Transport Annual Joint Conference Eno Foundation Energy Conservation Policies for Australian Board of Directors and Board of Consult- Cities. Final Report. ants. Transportation Quarterly XLII (1), 141-154.

Gleick, J. (1988) National gridlock: Scien- tists tackle the traffic jam. The New York Times Magazine, May 8.

Government of Western Australia (1991) Road Reserves Review: Final Report. A Joint Study for the Department of Plan- ning and Urban Development, Department of Transport, Main Roads Department, and Transperth, Perth.

24 World Transport Policy and Practice Volume 18.4 Sept 2012 Kenworthy, J. (2011) International Bench- Newman, P.W.G. (1992) The rebirth of the marking and Best Practice in Adapting to Perth suburban railways. In: Urban and a Future of Electric Mobility in Germany: regional planning in WA: Historical and Sustainable Transport or Just Electric critical perspectives. Hedgcock, D. and Cars? Report to Hessen State Government Yiftachel, O. (Eds), Paradigm Press, Perth. through the University of Applied Scienc- es, Frankfurt am Main, February. Newman, P. (2011) The Perth rail transformation: Some political lessons learned. Kenworthy, J.R. and Newman, P.W.G. http://sustainability.curtin.edu.au/local/ (1986a) The potential of ethanol as a docs/The_Perth_Rail_Transformation.pdf transport fuel: A review based on techno- (accessed April 7, 2012). logical, economic and environmental criteria. Issues in Energy Policy in Western Newman, P.W.G. and Kenworthy, J.R. Australia, Discussion Paper Nº 6/86, En- (1984) The use and abuse of driving cy- vironmental Science, Murdoch University, cle research: Clarifying the relationship 40 pp. between traffic congestion, energy and emissions. Transportation Quarterly 38 Kenworthy, J.R. and Newman, P.W.G. (4), 615-635. (1986b) From hype to mothballs: An assessment of synthetic crude oils from oil Newman, P.W.G. and Kenworthy, J.R. shale, coal and oil sands. Issues in Energy (1988) The transport energy trade-off: Policy in Western Australia, Discussion Pa- Fuel-efficient traffic versus fuel-efficient per Nº 7/86, Environmental Science, Mur- cities. Transportation Research - A 22A(3), doch University, 90pp. 163 -174.

Klaassen, L.H., Bourdrez, J.A. and Volmull- Newman, P.W.G. and Kenworthy, J.R. er, J. (1981) Transport and reurbanisation. (1989) Cities and automobile depend- Gower, England, 214 pp. ence: An international sourcebook. Gower, Aldershot, England. Larson T.D. (1988) Metropolitan congestion: Towards a tolerable accommodation. Newman, P. and Kenworthy, J. (1999) Sus- Transportation Quarterly 42 (4), 489-498. tainability and cities: Overcoming automobile dependence. Island Press, Washington Marchetti C. (1994) Anthropological invar- DC. iants in travel behaviour. Technical Fore- casting and Social Change, 47(1), 75-78. Newman, P. and Kenworthy, J. (2006) Ur- ban design to reduce automobile depend- Millard-Ball, A. and Schipper, L. (2010) Are ence. Opolis 2 (1), 35-52. we reaching peak travel? Trends in passenger transport in eight industrialized coun- Newman, P.W.G., Kenworthy, J.R. and Ly- tries. Transport Reviews, 2010, 1-22. First ons, T.J. (1988) Does free flowing traffic published on 18 November 2010 (iFirst). save energy and lower emissions in cities? Search 19, 267-272. Mitchell, R.B. and Rapkin, C. (1954) Urban traffic: A function of land use. Columbia Nijkamp, P. and Reichman, S. (eds) (1987) University Press, New York. Transportation planning in a changing world. Gower and the European Science Monheim, R. (1988) Pedestrian zones in Foundation, 340pp. West Germany - the dynamic development of an effective instrument to enliven the Newman, P. and Kenworthy, J. (2011) city centre. In New life for city centres: ‘Peak Car Use’: Understanding the Demise Planning, transport and conservation in of Automobile Dependence. World Trans- British and German cities. Carmen Hass- port Policy and Practice 17 (2), 31-42. Klau (ed), Anglo German Foundation, Lon- don, p. 107-155.

25 World Transport Policy and Practice Volume 18.4 Sept 2012 Orski, C.K. (1987) “Managing” suburban Whitelegg, J. (1988) Transport planning. traffic congestion: A strategy for suburban Cities, August, 309-310 (a book review of mobility. Transportation Quarterly 41 (4), Nijkamp and Reichman, 1987). 457-476. Whitelegg, J. (2011) Pay as you go: Man- Puentes, R. and Tomer, A. (2009) The aging traffic impacts in a world class city. Road Less Travelled: An Analysis of Vehi- Eco-Logica Ltd, Lancaster. cle Miles Traveled Trends in the U.S. Met- ropolitan Infrastructure Initiatives Series, Zeibots, M. (2007) Space, time, econom- Brookings Institution, Washington DC. ics and asphalt. Doctor of Philosophy Dissertation, University of Technology, Poole, R.W. (Jr) Resolving gridlock in Sydney. Southern California. Transportation Quar- terly 42 (4), 499-527.

Porter, D.R. (1987) The future doesn’t work. Urban Land, June 34-35.

Pratsch, L.W. (1986) Reducing commuter traffic congestion. Transportation Quar- terly 40 (4), 591-600.

Pucher, J. (1988) Urban public transport subsidies in Western Europe and North America. Transportation Quarterly 42 (3), 377-402.

SATS (1974) Sydney area transport study 1974. Vol 2, Minister for Transport, NSW.

Siegel, C. (2007) Removing freeways, restoring cities: From induced demand to reduced demand. Preservation Institute: http://www.preservenet.com/freeways/ FreewaysInducedReduced.html (accessed April 4, 2012).

Simpson, B.J. (1989) Urban rail transit: An appraisal. Contractor Report 140, TRRL, Crowthorne, Berkshire, England.

Stopher, P.R. and Meyburg, A.H. (1975) Urban transportation modeling and planning. Lexington Books, Lexington, Mas- sachusetts.

TEST (1992) Trip degeneration: A literature review. Transport and Environment Studies, TEST Report 99, London.

Webster, F.V., Bly, P.H. and Paulley, N.J. (eds) (1988) Urban land use and transport interaction: Policies and models. Re- port of the International Study Group on Land-Use/Transport Interaction, Avebury, Aldershot, England.

26 World Transport Policy and Practice Volume 18.4 Sept 2012 Traffic Growth: Modelling a Global the Netherlands, New Zealand, Norway, Phenomenon Spain, Sweden, Switzerland, Turkey and the United States. David Gargett The main result of all this data preparation Introduction has been the ability to document the con- In countries around the globe, traffic sistent and yet varied patterns over time growth has been a feature of the post in vehicle kilometres per person in many World War Two experience. The automo- countries around the globe. After rapid bile and commercial vehicles have multi- growth in the sixties and seventies, growth plied, as living has increasingly been inter- in traffic (all vehicle types) per capita has twined with mobility. consistently slowed, with many countries approaching saturation. In order to examine the similarities and differences in trends around the world, a The main results of the study are models recent report by the Bureau of Infrastruc- of vkt per capita as a function of real pet- ture, Transport and Regional Economics rol prices, fluctuations in the economy and in Australia (BITRE 2012a) describes the of a saturating effect of time. Each country different patterns of traffic growth in 25 is different, but the patterns of the models countries, and their determinants. The are amazingly similar. The models explain current article is an extract of that report, the common finding around the devel- focusing on three of the 25 countries cov- oped world of a fairly linear trend in total ered in BITRE (2012a). vkt over the past four decades – slowing growth in population has been matched There is a remarkable commonality be- by a declining rate of growth in traffic per tween countries in the determinants of person. Lately, there has been a signifi- traffic growth, but the variety of their op- cant effect of the global financial crisis in erations has generated an amazing variety lowering traffic levels per capital around of traffic growth patterns. Understanding the world. the determinants of past and likely future traffic growth is important to understand The models were used to provide base- the needs for infrastructure investment, case and scenario forecasts of future for congestion amelioration, for the road trends in traffic growth in the 25 countries. safety task and many other trends that The base-case forecast assumed that over concern governments and citizens in all the next decade unemployment would fall the 25 countries. as the negative effect of the Global Finan- cial Crisis wore off and that the petrol price Study Outline would remain unchanged in real terms. In The study and report that are the basis for scenarios, the unemployment rate and the this article (BITRE 2012a) examined the negative GFC effect were assumed to not trends in the growth of road traffic (vehicle decline. The scenarios for the petrol price kilometres travelled or vkt) in 25 countries in each country were based on an anal- around the world. ysis that linked world oil production and world GDP with world oil prices, and then Much of that report was technical in na- showed how corresponding petrol prices in ture, dealing with the sources of data, and each country could be calculated. details of variable construction and modelling. For example, data back to at least In sections that follow, the model con- 1963 was been assembled for 25 countries struction will be described, and models on vehicle kilometres travelled by vehicle for Australia, Germany and the United type, numbers of vehicles by vehicle type, Kingdom presented. Then follows the de- population, petrol prices, consumer prices scription of the petrol price modelling and and unemployment. Models were derived models for the three counties. Finally, the for Australia, Austria, Belgium, Britain, derivation of the scenarios forecasts for Canada, the Czech Republic, Denmark, the three countries will be described and Finland, France, Germany, Greece, Hun- results presented. gary, Ireland, Israel, Italy, Japan, Korea, 27 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 1 Australian trend in vehicle kilometres travelled per person Modelling World-Wide Trends in Traf- traffic growth. There are common patterns fic per Person in VKT per person around the world.

Traffic level (vehicle kilometres travelled In the extremely long term, VKT per per- or VKT) is modelled according to the fol- son has followed a traditional S-shaped lowing formula: adoption curve, as can be seen in the Aus- Traffic Level (VKT) = VKT per person tralian data in Figure 1. * number of people Demograpvhers can provide an under- But in the period from 1963 (the period standing of the dynamics of the population of most modelling done in the analysis), (number of people). many of the developed countries’ saturat- That leaves VKT per person, which is much ing trends can be modelled using a curve more tractable to model than aggregate convex in time, using positive time and

Figure 2 A saturating curve fitted using time and time squared 28 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 3 Patterns of traffic per person in 25 countries negative time squared variables, as illus- The other major finding of the study was trated in Figure 2 (for Germany). that downward deviations from these saturating trends occur when petrol prices or These saturating trends (or saturating S- unemployment levels rise or when eco- shaped logistic trends for long, or more nomic crises spook consumers. quickly developing, or earlier stage country datasets) represent the basic finding Real petrol prices in the various coun- from this research on 25 countries around tries, given the huge swings in world oil the world. People can only fit so much prices, have had similar large swings over travel into their days. Figure 3 shows the time (see Figure 4). The traffic modelling patterns of traffic per person for the 25 used ‘real home-currency’ petrol prices. countries cover by the research. The level of fuel tax is an important in-

Figure 4 Real petrol prices, 25 countries 29 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 5 Components of predicted levels of Australian traffic per person fluence on moderating the influence of 0.5 in 2008 and then 1.0 until 2011, after swings in world oil prices on domestic fuel which time the effect was assumed to de- prices. For example, European countries, cline to zero by 2021. with their high fuel taxes, are paradoxi- cally more sheltered in percentage change Australia terms from swings in world oil prices. The effect of this on the modelling is that the Figure 5 shows the growth of Australian ‘real petrol price’ variable had a lesser ef- traffic per person since 1963, and the fect on traffic per person in high fuel tax saturating trend fit to it, together with the countries. downward shifts from this curve caused by higher unemployment, higher petrol pric- The effect of the global financial crisis on es and the GFC. traffic levels is apparent in most countries (refer back to Figure 3). This effect is in- Figure 6 shows that the modelling also dependent of the effect of associated in- produced a fairly accurate prediction of creases in unemployment, and was mod- aggregate Australian traffic levels over the elled by dummy variables, generally set at period.

Figure 6 Actual/predicted aggregate traffic levels in Australia 30 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 7 Components of predicted levels of German traffic per person by higher unemployment, higher petrol Germany prices and the GFC.

Figure 7 shows the growth of German traf- Figure 8 shows that the modelling also fic per person since 1963, and the satu- produced a fairly accurate prediction of rating trend fit to it, together with the aggregate German traffic levels over the downward shifts from this curve caused period.

Figure 8 Actual/predicted aggregate traffic levels in Germany GFC does not seem to have affected Brit- The United Kingdom ain).

Figure 9 shows the growth of British traffic per person since 1963 and the saturating trend fit to it, together with the downward shifts from this curve caused by higher unemployment and higher petrol prices (the 31 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 9 Components of predicted levels of British traffic per person Figure 10 shows that the modelling also produces a fairly accurate prediction of aggregate British traffic levels over the period.

Figure 10 Actual/predicted aggregate traffic levels in Britain Petrol Prices The basic mechanism of the fuel price chain is depicted in Figure 11. In most countries around the world, one of the important influences on traffic growth Potential supply is a measure of the long- has been fluctuations in petrol prices. So run, business-as-usual level of possible an understanding and modelling of how world oil supply, balancing the depletion individual countries’ petrol prices are set of older fields by new field development is important in understanding the outlook and non-conventional sources of total liq- for traffic growth given different scenarios uids (where total liquids equals the sum about world oil prices. of conventional crude oil, non-conventional crude, and other liquids fuel sources). Actual supply (equal to actual demand) 32 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 11 The Fuel Price Chain is determined by the interaction of OPEC ated by various agencies, for example the surplus decisions and the oil price. International Energy Agency (IEA 2011). Actual supply in the short-run (equal to Once the world oil price is set, putting the demand) is determined by the additional price through the domestic exchange rate interaction of levels of OPEC surplus with gives a landed oil price. This determines the price of oil. the energy component of domestic petrol prices. Then adding wholesale/retail mar- Figure 12 shows the close relationship be- gins, excise tax and goods and services tween changes in the 3-year average of tax, results in the retail petrol price (price the ratio of world oil demand to world GDP, at the pump). and the real oil price expressed in real Spe- cial Drawing Rights (SDRs) per barrel of oil The potential world oil supply is equivalent (West Texas Intermediate (WTI) to 1998 to the long-run supply estimates gener- and OPEC reference basket thereafter).

Figure 12 World oil demand/price relationship 33 World Transport Policy and Practice Volume 18.4 Sept 2012 Table 1 shows a regression on the change year’s change less that of 3 years ago) and in world oil/GDP ratio using the log of the dummies. real SDR oil price, an ‘echo’ variable (last

Table 1 Change in 3-year average of world oil/GDP ratio as a function of the real SDR oil price The fit of the prediction from the equation to the actual ratio changes is shown in Fig- ure 13.

Figure 13 Actual and predicted annual percentage change in the world oil/GDP ratio

34 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 14 Actual and predicted world oil/GDP ratio Figure 14 shows the fit of the resulting surplus). But first it is necessary to specify predicted oil/GDP ratio versus the actual the potential oil supply. ratio. Using the equation in Table 1 and world Figure 15 shows three different scenarios GDP forecasts, a price of oil can be pre- for potential world oil supply (defined as dicted that will balance demand and ac- total liquids) to 2030. tual supply (potential supply less OPEC

Figure 15 Three oil supply scenarios The highest supply scenario comes from a in the next two decades from non-conven- peak oil optimist (Odell 2003). tional sources and from fields discovered and developed in that time ( e.g. Aleklett The IEA scenario comes from the Interna- 2009). tional Energy Agency (IEA 2011). The ‘IEA adjusted’ potential supply sce- Table 2 shows the calculations behind the nario corrects what some believe are op- world oil price-setting mechanism for the timistic IEA assumptions regarding supply IEA scenario. 35 World Transport Policy and Practice Volume 18.4 Sept 2012 Table 2 World oil supply/demand/price framework, IEA scenario for potential supply 36 World Transport Policy and Practice Volume 18.4 Sept 2012 Australian Petrol Prices by the Consumer Price Index. A regression was performed using the energy content Subtracting total tax (federal and state price of petrol regressed against the land- excise, and Goods and Services Tax) from ed oil price and several dummy variables. nominal petrol price and dividing by the Figure 15 shows the fit of the regression Consumer Price Index, gives the energy model values to the actual energy content content of the Australian petrol price. The price data. Adding tax to the model values landed price of oil in Australia is calculated for energy content gives a model value as simply the price oil in US dollars per for total petrol price. The fit between this barrel times the Australian exchange rate model value and the actual Australian pet- in Australian dollars per US dollar divided rol price is shown in Figure 16.

Figure 15 Actual and predicted energy component price of Australian petrol

Figure 16 Actual and predicted price of Australian petrol

37 World Transport Policy and Practice Volume 18.4 Sept 2012 German Petrol Prices using the energy content price of petrol regressed against the landed oil price and Subtracting total tax from the nominal one dummy variable. Figure 17 shows the petrol price and dividing by the Consum- fit of the regression model values to the er Price Index, gives the energy content actual energy content price data. Adding price of German petrol in Euros. The land- tax to the model values for energy content ed price of oil in Germany is calculated as gives a model value for total petrol price. simply the price of oil in US dollars per The fit between this model value and the barrel times the exchange rate in Euros actual German petrol price is shown in Fig- per US dollar divided by the Consumer ure 18. Price Index. A regression was performed

Figure 17 Actual and predicted energy component price of German petrol

Figure 18 Actual and predicted price of German petrol

38 World Transport Policy and Practice Volume 18.4 Sept 2012 United Kingdom Petrol Prices content price of petrol regressed against the landed oil price and several dummy Subtracting total tax from the nominal variables. Figure 19 shows the fit of the petrol price and dividing by the Consumer regression model values to the actual en- Price Index, gives the energy content price ergy content price data. Adding tax to the of British petrol in Pence. The landed price model values for energy content gives a of oil in Britain is calculated as simply the model value for total petrol price. The fit price of oil in US dollars per barrel times between this model value and the actual the exchange rate in Pounds per US dollar British petrol price is shown in Figure 20. divided by the Consumer Price Index. A regression was performed using the energy

Figure 19 Actual and predicted energy component price of British petrol

Figure 20 Actual and predicted price of British petrol

39 World Transport Policy and Practice Volume 18.4 Sept 2012 Scenario Testing sociated with IEA adjusted, IEA and Odell supply projections respectively). The three The base-case projections in Chapter 2 oil price scenarios are shown in Figure 21. were based on assumed values for the In all three oil price scenarios, exchange three main explanatory variables in the rate projections were derived from USDA models. Petrol prices were assumed to forecasts of real exchange rates (USDA- remain constant at their 2011 estimated ERA 2011). Excise was assumed to remain real values. Unemployment was assumed constant in real terms at the 2011 esti- to decline to 2015 and then remain con- mated amount. Sales tax rates were held stant. Finally, any effect from the Global constant at estimated 2011 values. Financial Crisis was assumed to disappear over the first decade. In a fourth scenario, unemployment was assumed to remain unchanged after 2011, The research examined five scenario pro- making it higher than in the base case. jections of traffic growth in addition to the base case. Finally, the GFC effect was assumed to persist for the whole forecast period, con- Three petrol price scenarios were generat- tinuing its negative effect on traffic levels. ed – high, medium and low. The base-case held real petrol prices constant at their es- The following sections examine the effects timated 2011 values. The high, medium of the five scenarios on the base-case pro- and low scenario petrol prices were gener- jection of traffic growth in example each ated from the three oil price scenarios (as- country.

Figure 21 World Oil Price Scenarios

40 World Transport Policy and Practice Volume 18.4 Sept 2012 Australian scenario projections der the base case line in Figure 21. This is because unemployment in Australia was Australian aggregate vkt projections range low in 2011, and the base case decline is around an upward base case trend. minor.

Looking at Figure 22, it can be seen that High, medium and low petrol price sce- the faster upward trend in the base case narios are spread around the base case. to 2020 is due to the negative effect of For Australia, both unemployment and the the GFC coming off. For Australia, the GFC petrol price have not in the past produced effect is the one generating the most sig- marked deviations of per person vkt from nificant downward deviation from the base the saturating trend, and this is reflected case. in the scenarios.

Unemployment being unchanged is the least significant, being almost hidden un-

Figure 22 Australian vehicle kilometre scenarios

41 World Transport Policy and Practice Volume 18.4 Sept 2012 German Scenario Projections the projection period.

German aggregate vkt projections range Unemployment being unchanged is less around a slowing upward base case trend. significant. This is because unemployment This is the combination of slightly declin- in Germany is moderate in 2011, and the ing forecast population with a trend in per base case decline is also moderate. person vkt that will be approaching saturation by the end of the forecast period High and low petrol price scenarios are (with the former starting to outweigh the spread around the base case, with the latter by the mid 2020s). medium scenario being almost identical with the base case. The high petrol Looking at Figure 23, it can be seen that price scenario causes the most significant for Germany, the GFC effect is the one downward deviation from the base-case generating the most significant downward traffic growth trend. deviation from the base case for most of

Figure 23 German vehicle kilometre scenarios

42 World Transport Policy and Practice Volume 18.4 Sept 2012 British Scenario Projections ployment in Britain was high in 2011, and the base case decline is significant. British aggregate vkt projections range around an upward base case trend. This High and low petrol price scenarios are is the combination of an almost stable spread around the base case, with the me- trend in per person vkt (that will be ap- dium scenario being almost identical with proaching saturation by 2016), with a the base case. moderate population growth trend. The high petrol price scenario generates Looking at Figure 24, it can be seen that the most significant downward influence for Britain, the unemployment effect gen- on the base-case projection. erates a significant downward deviation from the base case. This is because unem-

Figure 24 British vehicle kilometre scenarios

Conclusions The research described here has provided an overview of the different patterns of In countries around the world, traffic traffic growth in 25 other countries around growth has been a feature of the post the world, three of which have been exam- World War Two experience. The automo- ined here. In addition it has modelled the bile and commercial vehicles have multi- determinants of this growth. plied, as living has increasingly been inter- twined with mobility. The main determinant of the pattern of growth in traffic per person has been a And yet in spite of its obvious benefits, trend toward saturation in per capita traf- traffic growth presents one of the key chal- fic levels. After rapid growth in the- six lenges to those tasked with assuring the ties and seventies, growth in traffic per continued benefits of mobility. It is there- capita (all vehicle types) has consistently fore important to understand the nature, slowed, with many countries already at or causes and outlooks for traffic growth. approaching saturation.

43 World Transport Policy and Practice Volume 18.4 Sept 2012 The main contribution of this research are References models of vkt per capita as a function of this saturating effect over time, of petrol Aleklett, K. 2009, Future transportation prices, and of fluctuations in the econo- fuels without ‘business as usual’ as an op- my. Each country is different, but the pat- tion terns of the models are amazingly similar. http://www.aspo-australia.org.au/Refer- The models explain the common finding ences/Aleklett/20090610Sydney1Sma around the world of falling growth rates in rt2009.pdf aggregate traffic levels over the past four decades – a falling growth rate in popula- Bureau of Infrastructure, Transport and tion has been being reinforced by a de- Regional Economics (2007) Estimating ur- clining rate of growth in traffic per person. ban traffic and congestion cost trends for Lately, there has been a significant effect Australian cities, Working Paper 71, BI- from the global financial crisis in lowering TRE, Canberra, Australia traffic levels per capita around the world. http://www.bitre.gov.au/info. aspx?ResourceId=249&NodeId=59 But there are also many differences in the patterns of traffic growth uncovered in the Bureau of Infrastructure, Transport and 25 countries. Levels of traffic per capita Regional Economics (2010) Effectiveness differ substantially across the sample, of measures to reduce road fatality rates, from high levels in the United States to Information Sheet 39, BITRE, Canberra, low levels in Turkey (the latter being still in Australia the exponential growth stage). The trend http://www.bitre.gov.au/publica - toward saturation also varies substantially, tions/2010/is_039.aspx with some trends being highly curved and already reached saturation, while other Bureau of Infrastructure, Transport and trends are not even approaching satura- Regional Economics (2011) Road vehi- tion. Some jurisdictions respond much cle kilometres travelled: estimation from more substantially to changes in petrol state and territory fuel sales, Report 124, prices than others, while some show sub- BITRE, Canberra, Australia stantial effects from changes in the econ- http://www.bitre.gov.au/publica - omy. tions/2011/report_124.aspx

But however much they are alike or differ, Bureau of Infrastructure, Transport and Re- the models can be used to provide base- gional Economics (2012a) Traffic growth: case forecasts of future trends in traffic Modelling a Global Phenomenon, Report growth. These are useful in a variety of 128, BITRE, Canberra, Australia. contexts, for instance, in forecasting road http://www.bitre.gov.au/publica - fatalities from fatality rates (BITRE 2010), tions/2012/report_128.aspx examining needs for infrastructure, examining traffic growth in cities (Gargett and Bureau of Infrastructure, Transport and Cosgrove 2003, Gargett and Gafney 2003, Regional Economics (2012b) Traffic growth BITRE 2007, BITRE 2012b), and many in Australia, Report 127, BITRE, Canberra, other uses. Australia. http://www.bitre.gov.au/publica - Author details: tions/2012/report_127.aspx

Dr David Gargett Gargett, D. and Gafney, J. (2003) Traffic Research Leader growth in Australian cities: causes, pre- Bureau of Infrastructure, Transport and vention and cure, BITRE Staff Paper, Can- Regional Economics berra, Australia. GPO Box 501 http://www.bitre.gov.au/publica - Canberra ACT 2601 tions/2003/sp_002.aspx Australia

Email [email protected] 44 World Transport Policy and Practice Volume 18.4 Sept 2012 Gargett, D. and Cosgrove, D. (2003) Pre- dicting traffic growth in Australian cities, BITRE Staff Paper, Canberra, Australia http://www.bitre.gov.au/publica - tions/2003/sp_001.aspx

Gargett, D. (2010) Petrol prices in Austral- ia, paper given at the Australian Transport Research Forum, Canberra, Australia http://www.bitre.gov.au/publica - tions/2010/sp_005.aspx

IEA (2011) Energy prices and taxes, Inter- national Energy Agency, Paris, France.

IEA (2011) World energy outlook, Interna- tional Energy Agency, Paris, France.

Odell, P. 2003, Why carbon fuels will dominate the 21st century’s global energy economy, Multi-Science Publishing, Essex, U.K.

OPEC (2011) Reference basket oil price, available at http://www.opec.org/opec_web/en/

USDA-ERA (2011) Real exchange rate projections, United States Department of Ag- riculture, Economic Research Service, database available at http://www.ers.usda.gov/data/macroeco- nomics/

45 World Transport Policy and Practice Volume 18.4 Sept 2012 Congestion Offsets: routes, with hardly so much as a peak Transforming Cities by Letting Buses period bus service in operation, the Her- Compete culean task of trying to match the capacity of road infrastructure with demand simply Matthew Bradley and Jeff Kenworthy became too much.5 During earlier times, much academic focus had been on how we Better Cities are Possible were going to achieve a car-based mobility future. By contrast, since the 1980s Transport planners used to dream big we have, to quote the well known trans- dreams. Gleaming ribbons of steel and port academic Professor Phillip Goodwin, concrete were going to take us, uninter- been trying to solve congestion “when we rupted, anywhere we wanted to go. The must not build roads, increase spending, automobile would allow people to have ac- lose votes, damage the economy, or harm cess to all the good things their cities had the environment, and we will never find to offer, while also enjoying all the ben- equilibrium.”6 Although the mobility dream efits of a freestanding house on a spacious may have been at best an overly enthusi- suburban block. Sufficient roadway space astic venture and at worst a fool’s errand, would be provided so that everyone could it was at least a project embarked upon share in this bright mobility-based future.1 with a positive outlook, not the grim task outlined by Professor Goodwin. Not that Although this dream has had its share of he is the most downbeat example. Bleak critics in recent times, it was not without would be the only word to describe the as- merit.2 The dream of mobility, what might tonishingly titled Transportation Quarterly be called the great freeway project, ran into article “Metropolitan Congestion: Towards problems not because of any lack of good A Tolerable Accommodation”.7 Feel the joie ideas or intentions, but because inherently de vivre. it could not be made to work. In pursuit of this dream many cities drew up detailed This paper rejects the proposition that the transportation plans replete with highly best we can do with our road networks is ambitious road building projects. Indeed muddle through, or learn to tolerate con- this was the era of the grand transpor- gestion. Congestion is a solvable problem. tation plan, with transportation planning Better cities are possible. Most urban net- methods initially developed in Detroit and works, such as the telephone or electricity Chicago being rapidly exported not only networks, function smoothly and efficiently to the rest of the USA, but also to much in many cities around the world; making a of the world.3 Following such grand trans- telephone call is no more difficult than di- portation plans many cities built extensive alling a number and electricity is available road systems, at great expense in terms at the flick of a switch. Roads are unusual of both money and damage to urban form in that they are the only urban network (bisected neighbourhoods and widespread that can regularly be observed functioning demolition of inner city housing and busi- very poorly, even in wealthy cities where nesses), only to see traffic congestion turn all other urban networks are running well. the dream of mobility into something more In cities where the water, sewage, power, like one long traffic jam.4 gas, internet and telephone networks are all running relatively smoothly, it is not un- Not only has this outcome considerably common to see road networks completely dampened public enthusiasm for grand overwhelmed by congestion. road building projects, it has also had its effect academically. In the 1980s a pleth- All urban networks other than roads use ora of downbeat journal articles started to one of two techniques to manage ‘traffic’ appear as numerous countries, but most on their systems; they overwhelm demand notably the USA, descended into suburban with capacity or they manage demand to gridlock; an ‘honour’ previously reserved keep it below capacity. The classic exam- for CBD-bound arterial roads. With the ex- ple of overwhelming demand with capac- treme decentralisation of work into ‘salt ity is an electricity network designed with and pepper’ patterns across the landscape, sufficient capacity to continue functioning served primarily by cars along non-radial on the hottest day of the year. Demand 46 World Transport Policy and Practice Volume 18.4 Sept 2012 management is also used on many urban these weaknesses, an alternative must networks, if only inasmuch as networks be offered. The following section, “Sim- have a price on their use, thus dampening ple and Fair Congestion Control”, will use demand. The great freeway project was an an examination of both the strengths and attempt to overwhelm demand with capac- weaknesses of road pricing as a basis from ity, an attempt which, given widespread which to develop an alternative conges- congestion, it would have to be said has tion control method, namely congestion failed. The alternative to overwhelming offsets. demand with capacity is demand control, but this approach is only presently used With a key objective for congestion offsets on a small number of roads, such as toll being to retain the strengths of road pric- roads. With one traffic control approach, ing, some readers may, in due course, end overwhelming demand with capacity, hav- up seeing sufficient similarities between ing failed, and the other, demand control, road pricing and congestion offsets that not in general use, widespread congestion they view congestion offsets as just a vari- has resulted. ant of road pricing, particularly in cases where road pricing has been packaged So if the attempt to overwhelm demand along with the subsidising of public trans- with capacity has failed, why not wide- port, such as in London. Although this pa- spread demand control? Why do we run per acknowledges that road pricing has one, and only one, of our urban networks significantly informed the development of without some protection against conges- congestion offsets, it rejects the propo- tion? If congestion-proofing all other net- sition that congestion offsets are just a works makes sense, why does it not make variant of road pricing, and, in the section sense for roads? “Simple and Fair Congestion Control”, will argue that congestion offsets are a distinct The answer given here will be two fold. In method of congestion control. the section “How Cars Monopolise Roads and Prevent Buses from Competing” it will As will be discussed, a key difference be- be argued that the damage being done by tween congestion offsets and road pric- not congestion-proofing roads is greatly ing is the mechanism each approach uses underestimated, in that the result is not to achieve congestion control. With road just a poorly performing road system, but pricing congestion control is achieved by a structurally inefficient urban transport constructing a market for road space, and solution. Secondly, in the following sec- then only selling the quantity of road space tion “Simple and Fair Congestion Control”, that is actually available. By contrast, with it will be argued that road pricing, the only congestion offsets congestion control is congestion-proofing method under gener- achieved by altering the regulations under al discussion, has serious weaknesses. To which roadways are managed; sufficiently summarise, this paper will argue that we punishing roadway overuse and sufficient- do not congestion-proof roads because the ly rewarding roadway underuse so as to damage being done is greatly underesti- reduce the quantity of road space used to mated and because the only congestion- the quantity of road space that is actually proofing method under general discussion available. Congestion offsets retain road- has serious weaknesses. ways as a commons, with the objective being simply to better regulate that com- To address the first part of this problem, mons. Regulating a commons is something that we underestimate the damage being that everyone accepts, even in transport, done by not congestion-proofing roads, it as for example when one receives a fine is enough to simply show the greater dam- for parking on a street beyond the desig- age; something which will be done in the nated time allowance, or for parking on a next section, “How Cars Monopolise Roads street in a location where parking is not and Prevent Buses from Competing”. To permitted. By changing the regulations address the second part of this problem, under which roadways are managed, so as that the only congestion-proofing meth- to eliminate congestion, it will be argued od under general discussion has serious that our cities could be not just improved, weaknesses, it is not enough to just show but transformed. In short, this paper will 47 World Transport Policy and Practice Volume 18.4 Sept 2012 argue that we should not accept a ‘toler- Figure 1.8 This model is sufficient to show able accommodation’ with congestion. the lower speed performance of buses, as compared to cars. How Cars Monopolise Roads and Pre- vent Buses from Competing To be able to determine the comparative speed of the car and BRT services in Fig- All else being equal, it is not possible for ure 1, it is necessary to know the speed a bus service to provide a faster journey of the roadway and BRT line, the distance time than a car. Along bus routes, bus between the two points, the bus boarding services incur two unavoidable time over- and alighting times and the bus accelera- heads which cars do not; namely slowing tion and deceleration times. A major up- down and speeding up from bus stops and date of the Millennium Cities Database for waiting for passengers to board and alight. Sustainable Transport shows that in for- The impact of these factors can be dem- ty-three cities covering the US, Canada, onstrated using a somewhat simplified Australia, Europe and selected Asian cit- model that compares a bus service to a ies, the average roadway speed in 2006 car service, in the same setting. To give was 40.9 km/h.9 For the model presented buses their best chance in such a compari- here, this figure will be rounded down, and son, to see buses at their best, cars will 40 km/h will be used as the roadway speed be compared to a Bus Rapid Transit (BRT) for cars and for the speed of the BRT line. service, rather than to a traditional bus The distance between points A and B will

Figure 1: Comparison of Car and BRT Services Between Two Points service. BRT systems maximise the speed be 800 metres, which is approximately the performance of buses by using exclusive average inter-stop distance for BRT sys- rights-of-way and by measures such as tems surveyed in the Bus Rapid Transit level boarding and pre-boarding ticket- Planning Guide.10 This 800 metre value is ing. For cars, the model presented will, for also very close to the often-stated value of now, assume roadways free of congestion, half a mile between stations or stops for thereby also showing cars at their best. In the transit-oriented development planning the real world, cars and buses are often approach.11 not seen at their best; buses normally operate without the benefit of an exclusive For boarding and alighting times, each will right-of-way, and cars are, in many cases, be 12 seconds, approximately half the av- far from free of the effects of congestion. erage dwell time of the BRT systems sur- The impact of these real world factors on veyed in the Bus Rapid Transit Planning the service speed of cars and buses will be Guide.12 Bus acceleration and deceleration addressed later in this section. For now, will both be 1 m/s2; typical average per- a model of a simplified notional transport formance numbers.13 system between just two stops is shown in 48 World Transport Policy and Practice Volume 18.4 Sept 2012 The roadway speed number is sufficient tabase for Sustainable Transport shows, to calculate the speed of the car service, using the forty cities for which data are which is 40 km/h, the same as the road- available, that the average ratio between way speed. Calculating the speed of the bus speed and car speed is 54%.14 The BRT service is somewhat more complex. worst performing city is Chicago at 30% With the distance of 800 metres between and the best performing city is Hamburg stops being known, the average speed at 89%. To compare this real world data of the BRT service can be determined by to the theoretical outcome produced by calculating how long a bus would take to the model, roadway speed must be taken travel between the stops. The total time into account, as the model is based on a is the sum of the boarding, alighting, ac- speed of 40 km/h whereas the forty real celeration and deceleration times, plus the world cities have average roadway speeds time spent at the roadway speed. At the ranging from 24.0 km/h for Berlin to 57.5 previously given acceleration and decel- km/h for Chicago. Slower roadway speeds eration figure, a bus would reach the BRT make it easier for buses to compete with line speed of 40 km/h in approximately 11 cars, as the buses have to spend less time seconds over approximately 62 metres, getting up to and down from the roadway and decelerate from the line speed over speed, and also because board and alight the same time and distance. With accel- times become a smaller proportion of the eration and deceleration both taking 62 total trip time. Figure 2 shows, across the metres, this leaves 676 metres of the 800 speed range 0 km/h to 60 km/h, the ratio metre distance between stops occurring at between bus and car speeds produced by the BRT line speed of 40 km/h, a distance the model. that would be covered in approximately 61 seconds. Add to this time the combined As can be seen in Figure 2, when all other acceleration and deceleration time of 22 factors are kept constant, slower roadway seconds and the combined board and speeds result in better performance ratios alight time of 24 seconds, then the total for buses. If the cities from the Millennium time taken to travel between the stops is Cities Database are compared to the mod- 107 seconds. A distance of 800 metres el, at the average roadway speed for each covered in 107 seconds gives a speed of city, an idea can be gained of how well the approximately 27 km/h, so about 67% as bus systems in these cities are perform- fast as a car. ing. Figure 3 shows the forty cities from the Millennium Cities Database and the The update of the Millennium Cities Da- percentage by which they underperform

Figure 2: Ratio of Bus Speed to Car Speed Across a Range of Roadway Speeds 49 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 3: Real World Performance Compared to the Model or outperform the model. Negative values ly in Europe, with eight of the ten best per- show cities that are underperforming the forming cities being European. It was in model and positive values ones that are such European cities that the grand trans- outperforming it. portation plan era was at its most limited, and where serious efforts have been made With the average road speed for each city to have transit as a competitive travel op- used as the speed for the model, only four tion. The cities with the worst performing cities outperform the modelled BRT sys- bus systems are predominately in the USA, tem, namely Berlin, Frankfurt, Brussels with seven of the ten worst performing cit- and Hamburg. Three of those four cities, ies being US cities. The USA is where the Berlin, Frankfurt, and Brussels, only out- great freeway project achieved its zenith perform the model slightly, with ratios that and where efforts to accommodate and are three, five and eight percent higher provide for competitive transit systems, than the model respectively. The only city for example through the provision of bus to significantly outperform the model is lanes, reached their nadir. Hamburg, which has a ratio 15% higher than the model. An assumption of the For both the modelled case and the em- model is that cars and buses operate at pirical data, the data represents only the the same speed, which is not the case in transit line speed; the ‘in-vehicle time’ Hamburg. Due to the total removal of its from the perspective of a passenger. Nei- light rail system, Hamburg now has most ther the average ratio of 54% bus speed major bus routes operating in the middle versus car speed for the real world data, of wide roads on dedicated bus lanes, and or the 67% ratio for the model include it also has pre-emptive green-wave traf- time spent waiting for a bus to arrive, or fic light technology in operation. Although the speed difference between walking to a some caution is warranted when compar- main road to catch a bus, as compared to ing real world and theoretical data, the driving there. Given the figures used in the fact that only 10% of the real world cities model, the best a BRT system can deliver outperform the theoretical model suggests is a speed 67% as fast as a car. To achieve that the presented BRT system is a higher this outcome, a passenger would have to than average performance bus system. be travelling from a location right on top This is as intended; the model presented of a BRT stop to a location right on top of here aims to show buses at their best, not another BRT stop and arrive at their de- to show an average bus system. parture stop exactly as their bus is about to depart. This is far from the typical case. In terms of general trends, in Figure 3 it Normally, passengers have to walk some can be seen that the cities with the best distance to catch their bus service, wait for performing bus systems are predominate- that service, and then, after alighting their 50 World Transport Policy and Practice Volume 18.4 Sept 2012 bus, walk to reach their final destination. seconds, yielding a door-to-door travel Normal bus use is door-to-door, not stop- time of 1,440 seconds and, over the 15.6 to-stop (or, in the case of traditional bus kilometre total distance, an average trip services, door-to-door, not kerb-to-kerb). speed of 39 km/h. As before, calculating the speed of the BRT service is somewhat To calculate the speed impact of these two more complicated. Also in a similar fashion additional factors, getting to and from BRT as before, with the door-to-door trip dis- stops and waiting for services to arrive, the tance of 15.6 kilometres being known, the model presented must be expanded to be average speed of the BRT service can be a model of a full urban transport system. calculated by calculating the total time a Based on the 800 metre distance between person would take to make this trip using stops, the model will be expanded to con- the service. The total time is the sum of sist of an 800 metre grid of major roads the time passengers spend walking to and along which BRT lines also run, with a BRT from their departure and destination BRT interchange stop at every intersection; a stops, the time spent waiting for bus serv- model that is similar to the Squaresville ices to arrive, plus the time spent traveling model of Dr. Paul Mees.15 From the US on the BRT system itself. With regards to 2009 National Household Travel Survey, wait times, a reasonable assumption for a the average trip length for this model ur- grid of BRT lines is that a person would ban transport system will be, as near as have to wait twice for a bus, once on a possible, 15.7 kilometres.16 On an 800 horizontal line and once on a vertical line. metre grid, the average distance travelled At the previously mentioned figure of an to get to a stop, assuming uniform popula- average wait of three minutes for a bus, tion density, would be 400 metres at both the total wait time would be six minutes. the departure and arrival end, so 800 me- As for the 200 metre walking trip at each tres in total. To, again, give buses their end of the journey, at a brisk pace of five best chance in this comparison, a tran- km/h this distance of 400 metres would be sit-oriented development style of urban covered in 288 seconds. Finally, the time planning will be assumed, with sufficient spent on the BRT system itself needs to be focusing of development around each calculated. From previous calculations, the transit stop to reduce total walking dis- time between stops is 107 seconds, which tance to 200 metres at both the departure yields, for the 19 stops being travelled, a and arrival end, so 400 metres in total, or BRT travel time of 2,033 seconds. Add all half the distance at uniform density. This these values together, and the total 15.6 leaves 15.3 kilometres of the 15.7 kilome- kilometre journey takes 2,681 seconds, at tre distance to be travelled on the BRT sys- an average door-to-door speed of approxi- tem, which, at 800 metres between stops, mately 20.9 km/h. is 19.125 stops. For this model, 19 stops will be used, yielding a total trip distance So, in a congestion free environment, a BRT of 15.6 kilometres. As for passenger wait system, on a grid of BRT lines spaced at a time; this is determined by bus service typical transit-oriented development level frequency. The service frequency for BRT of 800 metres with development focused systems surveyed in the Bus Rapid Tran- around stops, is able to provide a door- sit Planning Guide, averaging the available to-door travel speed approximately 54% peak and non-peak figures, is one bus on as fast as a car. Slight further improve- average every six minutes, which gives an ments might be possible to this theoretical average wait time of three minutes.17 outcome via the running of express buses, though their speed impact tends to be cur- The car speed on non-arterial roads will, tailed by the concomitant increase in wait for this model, be assumed to be 20 km/h, times at stops, as passengers watch other an aggressively traffic calmed level. At this people’s express buses whoosh by. Anal- speed, the 400 metre trip to and from an ysis showing, in an uncongested setting, arterial road, 200 metres at both the start BRT delivering door-to-door travel speeds and end of the journey, would take a to- only a little better than half as fast as a tal time of 72 seconds. The remaining dis- car, is not the slightest bit controversial. tance of 15,200 metres on arterial roads There is no analysis anywhere showing running at 40 km/h would take 1,368 BRT systems being even close to competi- 51 World Transport Policy and Practice Volume 18.4 Sept 2012 tive with cars in terms of speed, in an un- in Urban Transit Systems and Technology congested setting, because there cannot for private autos on a street (as opposed be such analysis. With regards to door- to on a freeway) which is “heavily loaded to-door trip times, there are four areas in but somewhat below capacity.”19 At an av- which buses incur time costs that cars do erage occupancy of 1.67 people per car, not; namely time accelerating from and the average level recorded by the US 2009 decelerating to stops, time boarding and National Household Travel Survey,20 one alighting, time waiting for buses to arrive, lane with a capacity of 700 passenger cars and the extra time needed to walk to and per hour (pcph) can service a load up to from bus stops. 1,169 people per hour using cars. For the sake of simplicity, the measure “people There are no areas in which cars incur per hour” will be referred to, from here on, time costs that buses do not. As a con- as passengers per hour (pph), the metric sequence, regardless of how the figures used for public transport. When reporting are ‘tweaked’ in a model such as the one values for cars, all passengers per hour given, no model whose method and data levels will, of course, include the driver of has any grounding in the real world is go- a car. The occupancy level being used here ing to show anything except even very is somewhat generous since it is an overall good BRT systems only providing a door- average occupancy for cars, which includes to-door speed around half that provided more heavily loaded cars on weekends by a car, in an uncongested setting. In- and other off-peak times when people go deed, the overall speed outcome from the out together for social purposes and roads model, which has BRT providing a service are often less busy. Roads are most heav- speed 54% as fast as a car, is based on a ily loaded during peak or near peak times, transit line speed that is 67% as fast as a when average occupancies tend to be only car. Only seven out of the previously dis- a little over one person per car, due pri- cussed forty cities from the Millennium Cit- marily to the widely acknowledged issue ies Database update had a bus perform- of single occupancy driving when traveling ance outcome greater than 67%.18 All of to and from work. these cities are European, and all have an average roadway speed lower than the 40 By contrast to the 1,169 pph capacity us- km/h speed used for the model, with the ing cars, a BRT service could support far average roadway speed for this group of higher loads in a single lane. For exam- seven cities being 29.4 km/h. As previous- ple, a simple BRT service might consist of ly discussed, lower roadway speeds result one all-stop and one express bus service, in better bus performance ratios. both running at three minute intervals. If such a service were run using typical It is the above type of analysis that under- twelve metre buses, then the crush-load pinned the great freeway project; the post- capacity of each bus would be of the order war attempt to provide automobile mobil- of 85 passengers.21 Unlike car services, ity to all via vast road building projects. the normal mode of operation for buses After all, if it can be demonstrated in a few is for strangers to travel together and so pages that buses cannot possibly compete all of the capacity of these buses could with the speed of cars, why not deliver the be utilised. The maximum capacity level ‘best’ option to all. It is, though, an as- of the example BRT service would only be sumption of the above analysis that there touched for one moment and at one point are no capacity constraints and therefore along its route, with other times and areas there is no congestion. If analysis such as running at lower levels due to passenger that above is conducted with capacity con- load variation, but this is also what occurs straints, a very different picture emerges. with at-capacity car services. Two services each running at three minute intervals, Consider, for example, a typical urban gives a total service level of 40 buses an road. The capacity of the road lanes that hour. With each bus having a maximum make up such roads depends on many fac- capacity of 85 passengers, the overall BRT tors. Here, a capacity figure of 700 pas- service would have a maximum capacity senger cars per hour per lane will be used; of 3,400 pph. the mid-point of the capacity range given 52 World Transport Policy and Practice Volume 18.4 Sept 2012 The 40 buses an hour needed to run this typically run independently as is shown in 3,400 pph bus service would only use a Figure 4; in the general case, buses share very small proportion of the capacity of a roadway with cars. By running buses a BRT lane. By comparison, any attempt in an exclusive lane, BRT in many ways to service this 3,400 pph load using cars has more in common with heavy rail, as on the example 700 pcph roadway lane, those who conceived this mode intended, which is heavily loaded when servicing than it does with buses on normal shared- 1,169 pph, would very quickly reduce the use roads. Indeed, a strong case could be lane from full speed to a zero or near-zero made that the running of buses in an ex- speed. Using the previous door-to-door clusive lane, rather than in traffic, is the journey speed calculation method, Figure defining feature of BRT, and the reason 4 shows the speed of a car service oper- we have a separate name for this transit ating from full roadway speed, the previ- mode. Figure 5 shows what happens when ously used 40 km/h figure, to the roadway the car and bus services are run on the being jammed, at 0 km/h. The roadway same roadway; in other words what hap- speed scale in Figure 4 is shown, left to pens when the BRT buses are taken off right, from highest speed to lowest speed, their exclusive right-of-way and mixed in reversing the typical presentation, so as with car traffic. to show the impact on service speed as congestion increases. For comparison pur- With cars and buses sharing the roadway, poses, the speed of the example BRT serv- there are no points at which buses outper- ice is also shown, unaffected by the car- form cars. As can be seen in Figure 5, al- related congestion degrading the speed of though it is the car users who are causing the roadway that the cars are running on. the congestion, the bus users are bearing the cost of congestion along with the car Presented separately, the superior speed users. Operating in a congested environ- performance of cars at higher car roadway ment imposes two costs on bus services, speeds, which would occur at lower load the time cost to passengers of longer levels, is clear. Also clear, though, is the journeys and the extra cost to operators superior speed performance of buses at of running a slower service, with more higher loads, once congestion has taken its buses and drivers being needed to serv- toll and the speed of the car roadway has ice routes at lower speeds and, of course, dropped. In this example, once the road- higher generalised operating costs such as way speed has dropped to below approxi- fuel and wear and tear. By bearing a part mately 21 km/h, the example BRT service of the burden of the congestion caused is providing a faster door-to-door journey by cars, buses provide a worse service time than a car. But cars and buses do not and cost more to run. Instead of cars be-

Figure 4: Car and BRT Service Speed on Separate Roadways 53 World Transport Policy and Practice Volume 18.4 Sept 2012 Figure 5: Car and Bus Service Speed on a Shared Roadway ing faster at lower loads and buses being The capacity range that buses on normal faster at higher loads, cars become faster roads can comfortably service, and the at all loads by offloading part of their con- percentage of roadway used, is a critical gestion cost onto buses. In effect, the car issue. If you reduce the above described mode has been allowed to ‘poison’ the bus bus service from 14 buses an hour to sev- mode. en buses an hour, one bus approximately every nine minutes, you reduce the load The role of the bus mode, the mode crip- serviced to 595 pph, and the percent- pled by cars, is not one that the car mode age of roadway used to 2%. At this nine can effectively substitute for. At a level of minute frequency, you are close to the 1,169 pph, where cars are heavily load- border of where people stop just turning ing a single lane, there is just enough load up and waiting for a bus, and start looking to run a high frequency bus service. For at timetables;24 a characteristic that will example, a load level of 1,169 pph would be used here to mark the low end of the fill approximately 14 standard twelve me- load range which buses on normal roads tre buses to capacity every hour, yielding can be said to comfortably service. At the a service frequency of one bus approxi- high end of the load range, if you increase mately every four minutes.22 To show the frequency of the all-stop service to one the amount of road space that would be bus every three minutes and run three used by a 14 buses per hour service, it is parallel express services at the same serv- necessary to know how much road space ice frequency, you now have a 6,800 pph a bus takes up compared to a car. From bus service using approximately 23% of the Transport Research Board’s Highway the available roadway slots. Use 18 me- Capacity Manual, buses will be calculated tre, 110 passenger buses,25 which will use here as taking up twice the road space of slightly more roadway space, then capac- cars, the mid-point between the figures ity increases to 8,800 pph. This 8,800 pph for level and rolling terrain.23 Using this level is in the same general capacity range conversion rate, a 14 buses per hour serv- as a three lane freeway running just cars, ice would require 28 of the 700 pcph ‘slots’ reflective of the massively better roadway available on the example road lane. In utilisation of buses in comparison to cars. other words, the described 14 buses per Although, in principle, the load that buses hour service would use up 4% of the avail- on normal roads are able to service could able road space, one-twenty-fifth of the be pushed higher than this level, it is at road space required by cars to service the these higher load levels that BRT and rail same load. Another way of looking at this solutions begin to come into their own. is to say that to provide the same service using cars would require twenty-five times In terms of infrastructure, buses on nor- more roadway space than using buses. mal roads are able to comfortably service mid-level loads using only a small slice of 54 World Transport Policy and Practice Volume 18.4 Sept 2012 the roadway capacity of one lane. In the Perhaps the most important of these dif- above examples, 595 to 8,800 pph loads ferences is that light rail vehicles have a were serviced using approximately 2 to maximum passenger capacity far higher 23% of the capacity of a heavily loaded than that of buses. Urban Transit Systems roadway lane. If the approximate mid- and Technology lists light rail vehicles as points of these values are taken, then having a maximum passenger capacity buses on normal roads can be character- more than six times that of buses, with ised as a transport mode that can serv- the maximum capacity of regular buses ice loads of the order of 5,000 pph, and being listed as 120 passengers, whereas which require approximately one-eighth multi-carriage light rail vehicles are listed of the capacity of a heavily loaded road as having a maximum capacity of 750 pas- lane to do so. Attempt to service these sengers.26 Not only does this greater ca- mid-level loads with cars, and the road- pacity mean that light rail can service the way resources required are of the order of same passenger load as buses using less twenty-five times higher, due to the car’s drivers, it also reduces all costs that in- inefficient use of roadway space; hence crease with fleet size, such as vehicle and the chronic congestion problems of the car driver scheduling costs. The ubiquitous use mode. Attempt to service these loads us- of either pre-boarding ticketing or conduc- ing BRT and, with the defining character- tors to issue tickets is a further advantage istic of BRT being the use of an exclusive of light rail, though, in this case, one that right-of-way, you incur 100% of the road could be adopted more widely by tradition- lane infrastructure cost, eight times that al bus services. For all of the differences used by buses on normal roads. Attempt between buses and light rail, there is one to service these loads with heavy rail and stark similarity; when light rail is run in you again incur 100% of the right-of-way traffic it has exactly the same car-caused infrastructure cost, plus you are trying congestion problems that buses on normal to service mid-level loads with a mode roads have. In a situation where conges- that excels at high-loads. As with cars, in tion was controlled, the performance of servicing mid-level passenger loads, BRT light rail systems running in traffic would and heavy rail also have massively high- be enhanced, and much of the discussion er right-of-way infrastructure needs than in this paper, which applies to buses, may buses on normal roads, due, perversely, also apply to such light rail systems. to their passenger carrying efficiency. BRT and heavy rail take 100% of a lane and To summarise, currently all of the public then, at mid-level passenger loads, use transport options for servicing mid-level that lane so efficiently that most of it is left loads either have their speed performance unused. Cars on normal road lanes, and crippled by the congestion caused by cars, buses and trains in exclusive lanes, both, or have their financial position seriously for divergent reasons, use vastly more degraded by having to pick up 100% of right-of-way infrastructure when servicing their right-of-way cost, eight times the av- mid-level passenger loads than buses on erage amount required by buses on nor- normal roads. From the previous calcula- mal roads to service mid-level loads. Due tions, cars use twenty-five times the right- to the congestion caused by cars, cities of-way infrastructure of buses on normal lack a really effective mid-level transport roads, and both BRT and heavy rail use mode, one that is both reasonably fast and eight times as much. reasonably cheap.

Heavy rail, the running of trains in an ex- Cars are normally occupied by one or two clusive right-of-way, is not, of course, the people, buses by many tens of people only way to operate trains; trains can also and heavy rail trains by many hundreds be run on normal roads. Such services are of people. The proper role for cars, buses known variously as trams or light rail, and and heavy rail trains as, respectively, low, although some such systems run in sec- mid, and high load services could not be tions at least on exclusive lanes, others more apparent. It is a simplification to say run on normal roads, mixed in with traf- that cars work best between 0 and 1,000 fic. Light rail services differ from bus serv- pph, buses work best between 1,000 and ices in a number of important respects. 10,000 pph and heavy rail trains work best 55 World Transport Policy and Practice Volume 18.4 Sept 2012 between 10,000 and 100,000 pph; but it in a dysfunctional manner. Nevertheless, is not much of a simplification. Although, such operation is the norm. It is, after all, with multiple lanes, cars can provide many dysfunctional to run a network, any net- times the previously calculated 1,169 pph work, where the service provided regularly capacity of a single lane, such multi-lane collapses due to load exceeding capacity, roads can be seen, in many situations, as without instituting some form of conges- merely a desperate attempt to keep the tion control. Furthermore, in both the pop- low capacity car mode functioning, an is- ular and academic literature, there seems sue that will be discussed in greater detail to be a distinct absence of either lay peo- in the next section. ple or planning experts rushing to defend congestion as the sensible outcome; when The 1,000 to 10,000 pph mid-level load congestion is talked about it is normally in range is not a minor one; vast swathes tones of condemnation not praise. So if, as of the urban transport task occur in this a matter of logic, it would seem running range. Cars, though, via the mechanism road networks without congestion control of congestion, have stopped buses on nor- makes no sense, and if congestion is wide- mal roads, the obvious transport mode for ly disparaged, why do the overwhelming servicing these mid-level loads, from com- majority of cities continue to run their peting on its merits. Consequently, a ca- road networks without congestion control? pability chasm has opened up between low and high loads, with no mode being avail- The answer that will be given here is that able to effectively service mid-level loads. there is only really one idea, road pricing, The result has been disastrous. The story in the ‘marketplace of ideas’ about roadway of many modern cities, particularly low to congestion control, and that road pricing mid-density cities, is the car mode running has two serious weaknesses, which will be well over capacity and the heavy rail mode discussed.28 There are many ideas in the running well under capacity, as transport related disciplines of urban planning and planners desperately try to make each transportation planning — transit-oriented mode service mid-level passenger loads development, balanced transport, sustain- that neither is suited to. A train line op- able planning — but, other than road pric- erating at a fraction of its potential capac- ing, none of them are directly aimed at ity, running alongside a heavily congested congestion. None can show, as road pric- arterial road, is the signature image of the ing can, in a short, clear and compelling modern transport planning dilemma. This manner how chronic congestion could be dilemma will not be resolved until cars eliminated. Ironically, the free market idea are no longer permitted to offload their as to what to do about congestion lacks congestion costs onto buses, or, in other competition. words, until buses are allowed to compete. That the economics profession has some- Simple and Fair Congestion Control thing significant to say about congestion is not surprising; the management of scar- What would lead a person to believe — city is commonly said to be the central even for a moment — that a sensible ap- focus of economic thought.29 Roads are proach to urban transport was possible a scarce resource and so, as a matter of without some form of roadway congestion inevitable economic logic, if you do not control? For cities without railways, roads implement a rationing scheme then queu- are the entire motorised transportation so- ing (i.e. congestion) becomes your default lution, and even for those cities that have rationing scheme. Not having a formal significant rail systems, roads still play a rationing scheme in place does not, after critical role. For example, London, home of all, change the scarce nature of the road- the well-known London Underground, as way resource. Economists highlight the well as a substantial surface rail system, fact that the use of queuing as a ration- still had 67% of its motorised travel hap- ing scheme leads to a wildly inefficient al- pening in private cars as of 2006.27 Few location of roadway resources. For exam- if any cities have such widespread alter- ple, an eye surgeon, whose time might be natives to road transport that they can billed out at hundreds of dollars an hour afford to let their road networks operate sits in traffic the same as, for example, a 56 World Transport Policy and Practice Volume 18.4 Sept 2012 junior office clerk, whose time might only local services, not for the right to reside be billed out at fifteen dollars an hour. at a house, as such. In a similar manner, The economic solution to this problem is to pay to use road space is to implicitly road pricing,30 a market based approach, acknowledge that you do not own that which is really what you would expect any road space. As with local council rates, mainstream economist to say, regardless road taxes are widely accepted as these of the road context. Present to an econo- payments are used to maintain the roads, mist a general problem where the prima- not for the right to use the roads as such. ry context was resource scarcity, and the Currently roads are owned by the people primary symptom was queuing, and then and held in trust by the government; they ask for a ‘diagnosis’ and they would likely are a commons. Road pricing involves say that the correct course of treatment is transferring ownership of the roads from to establish a market and may also com- the populous to either the government or ment that you have a common, but readily private interests, who then charge people treatable ailment, called socialism. for access to what was previously theirs. People’s dislike of being dispossessed of The basic mechanism of road pricing is what is theirs, and therefore of road pric- that people pay for the road space they ing, would be considered by some to be use. If the objective is simply congestion proper, and really should be considered by control, then these payments may be lim- most to be expected. Land is not a second- ited to times and places where there is a tier issue, and a strong case can be made congestion problem, with free access to that public roads and streets represent the the roads allowed otherwise. This mecha- last great area of common land. Whether nism has one key strength, namely that ownership of this land is transferred to it works. Set a price for being on a road, the government or to private interests, with higher prices during times and places as required by road pricing, both options of higher demand for road space, and traf- involve people’s disenfranchisement from fic levels will be reduced. Make the price arguably this last great commons. This sufficiently high and congestion will be requirement for a loss of ownership has eliminated. No one needs to be convinced made road pricing all but impossible to sell of this, it is both intellectually obvious and politically, and is a key political weakness can be seen on a number of tollways which of road pricing as a congestion control have been built above free access road- method. ways, where the tollway is running below capacity and the free access roadway is An alternative to the loss of ownership re- hopelessly congested. Along with the key quired by road pricing is to keep the roads strength of road pricing — that it works as common property, held in trust by the — it will be argued here that road pricing government, but to change the regula- has two key weaknesses, one political and tions under which that commons is man- one organisational. The objective of the aged. The current roadway regulations, rest of this section will be to outline these most everywhere, distill down to saying: two weaknesses and thereby develop an “you may take more than your fair share alternative approach, namely congestion of available road space.” Given that not offsets, which retains the key strength of taking more than your fair share is such road pricing — that it works — but which a basic moral principle that it tends to be does not have its weaknesses. Both the one of the first taught to young children, weaknesses of road pricing relate to the it is little short of bizarre that we run our mechanism of its operation; that money is roads this way. What might the roadway paid to buy access to road space. commons look like if it were instead managed under the principle: “you may, on Paying for something is an implicit ac- average, take only your fair share of avail- knowledgement that you do not own what able road space?” you are paying for. People do not pay rent on a house that they own; as to do so At the moment people claim road space by would be to accept that they do not own it. taking a vehicle onto a roadway. If that ve- Local council rates are generally accepted, hicle is a car, one road slot is claimed; for but these payments are used to provide larger vehicles, such as buses and trucks 57 World Transport Policy and Practice Volume 18.4 Sept 2012 a greater number of slots are claimed, and travelling in buses, using the same occu- for smaller vehicles, such as motorcycles, pancy levels and other values as the previ- a lesser number of slots are claimed. Con- ous section, how much road space is used gestion occurs when the number of road by each group? To load the example single slots claimed is in excess of the number of lane to capacity equally with car and bus road slots that are available. The number users requires approximately 2,250 road- of people inside a vehicle has, in the gen- way users per hour, approximately 1,125 eral case, no effect on the amount of road users for each mode. With the users so space that can be claimed. It is difficult split, Figure 6 shows the road space utili- to see how an approach based on vehicle sation levels for these 2,250 roadway us- size could even begin to meet any criteria ers on a road lane that has a capacity of of fairness; it is just by default what hap- 700 pcph. pens. Instead, consider the simplest notion of a fair share; that a fair share is an As shown by the average bar in Figure even share. In the classic example, if there 6, when a roadway lane with 700 avail- is a cake to be divided up between peo- able slots is loaded with 2,250 users, the ple then it is ‘fair’ that everyone gets an fair share of road space is approximately even slice. This is a simple approach, and 0.311 slots per user. Figure 6 also shows it does have its weaknesses. For example, that car users are using far more of the what if one of the people who the cake is roadway than the average, at a level of being shared between is currently starving approximately 0.599 slots per user, and to death; should that person only get an that bus users are using far less, at a lev- even share? Similarly, in a transport con- el of approximately 0.024 slots per user. text, few people object to disabled drivers Even if the bus occupancy level is reduced receiving privileged parking access. Nev- to be the same as that for cars, based on ertheless, an even slice, an even share, is the assumption that a car has five seats, a widely used moral principle, and will be car users still come in well above the av- used here to examine what taking one’s erage, and bus user still well below. At fair share of road space might involve. an equal 33% occupancy, the car users’ How does road use look if “a fair share” is slot use remains at approximately 0.599 taken to mean that people are entitled to slots, and the bus users’ slot use increases an even slice of a roadway they are on? from approximately 0.024 to 0.071 slots. To examine this question, it is useful to re- So, even with equal occupancy, car users turn to the roadway capacity example of are still taking approximately eight times the previous section, where one lane had the road space of bus users. The overuse a capacity, when heavily loaded, of 700 of road space by car users can be seen pcph. If such a road lane is loaded equally more clearly in Figure 7 which, using the with people travelling in cars and those same occupancy levels as those for Figure

Figure 6: Use of Slots by Mode 58 World Transport Policy and Practice Volume 18.4 Sept 2012 6, shows the difference from the average, Road pricing gives a lead as to how the with positive values showing slot overuse above situation might be addressed, such and negative values showing slot under- that roadways are managed under the use. principle: “you may, on average, take only your fair share of available road space.” In Figure 7, it can be seen that, for this ex- With road pricing, roadway users are cus- ample, car users are taking approximately tomers, not owners, and so everyone us- 0.288 slots more than their fair share and ing roadways pays and there is a net flow bus users are taking approximately 0.288 of value from the customers to the ex- slots less than their fair share. At this ternal roadway owners, be they private point, with the roadway heavily loaded, interests or the government. If, though, each bus user is generously ‘donating’ ap- roadways are a commons, such transfers proximately 0.288 slots to a car user. Even should be made between the owners, from in the absence of congestion, as is the case those who are over-using a roadway to here, the above situation is highly unfair. those who are under-using it, with no net The car users’ congestion-free trip is only flow of value outside the group of own- being made possible by the bus passen- ers, namely the public. This could be done, gers’ decision to take a bus, thus freeing during times when road space is scarce, up road space which allows for the car us- by charging private vehicle users, who are ers roadway overuse. Tip over the line into on average overusing the roads, and by congestion, and not only are car users tak- paying public bus users, who are on aver- ing more than their fair share of roadway age under-using the roads. space, they begin dumping part of the cost

Figure 7: Overuse and Under-use of Slots by Mode of the congestion they are causing onto Paying public bus users directly would risk bus users. All statements about buses be- people chasing the payment as income, ing “unable to compete” with cars should but such payments could be made indi- be viewed in this light. Due to the way rectly, by reducing the price of bus tick- roads are regulated, or not regulated, bus ets via a discount. There is no real limit users are ‘punished’ with congestion while to how large such discounts could be, as also donating road space to those that ticket prices could be raised to accommo- are dishing out that punishment. Take the date higher discount levels, with the extra bus; no good deed will go unpunished. In funds being used to improve bus system this situation, of course buses “can’t com- infrastructure, or to fund other costs as- pete.” Saying that buses cannot compete sociated with providing a better bus serv- is like saying a boxer who has been given ice.31 These discounts are not a subsidy, tranquillisers cannot compete against an- they are a payment for the use of a scarce other who has been given amphetamines; resource from the people who are overus- technically true but highly misleading. ing that resource to the people who had a 59 World Transport Policy and Practice Volume 18.4 Sept 2012 right to that resource but gave it up. The The second key weakness of road pric- community could decide how much traf- ing is organisational, and again relates to fic would be allowed on a given roadway, the fact that, under road pricing, money econometric professionals determine what is paid to buy access to road space. Pay- transfer price would need to be charged ments come in many forms, such as pur- at given times to hit that target, and each chases, taxes and fines, and each - pay person overusing the roadway pays each ment type has a different character. With person under-using it. The higher the road pricing the payment type is a fee for transfer price from car users to bus users, a service, with each driver acting in the the less attractive the car option becomes role of a customer who is purchasing road and the more attractive the bus option be- space. When making a purchase, people comes. As more people use buses, overall expect to pay for what they use, not for vehicle load drops. If the outcome is road- more than they use. Few people would way under-use, the transfer price was set be happy if they went to a shop to buy a too high. If the outcome is roadway over- litre of milk and the shopkeeper tried to use, the transfer price was set too low. charge them for two litres. Because road pricing works via the payment of a fee for Such an approach would be financially a service, precise metering is needed. In neutral for a person who was not over-us- the few cities that have implemented road ing the roads. Take, for example, someone pricing, sophisticated toll gate systems are who travelled to work and back Monday to the norm, and, in significant part due to Friday, where there was only one-fifth of this complexity, such road pricing systems the required road space available to make tend to be limited to the centre of the city. that journey in a car everyday. To achieve a given traffic level, the cost of taking a If the objective is instead to regulate the car on this trip to work might be set at roadway commons, a fee for a service is $8 a day, charged in a manner to be dis- not the only payment method that might cussed shortly. Bus users would, for this be used. Another method, already widely example, receive a $2 discount on their used in other areas of roadway commons fare each day for their roadway under- regulation, is available; namely the penal- use. If a person took the bus to work four ising of misbehaviour. For roadways we al- days a week and their car one day a week ready have a wide spectrum of penalties they would pay $8 on the day they took for different types of misbehaviour, from, their car and, on the other four days, have for example, small fines for parking in a their bus ticket discounted by $2, for a clearway to jail time for causing death by total discount of $8. The requirement to the reckless use of a motor vehicle. There pay $8 and the total discount of $8 cancel is no reason that the offence of “Causing each other out. For those currently using Congestion” could not join the long list of a fair share of roadway space, the intro- all of our other traffic offences, with appro- duction of congestion offsets would have priate penalties attached. Tickets for caus- a negligible direct effect; the money lost ing congestion could be issued using the and the money gained would balance out. same general approach as parking tick- For those using more than their fair share ets; any private vehicle on a given stretch of roadway space, they would pay those of road, on a given side of the road, at a using less than their fair share. In both given time, incurs a penalty. In contrast cases, the public would remain owners of to parking tickets, though, such penal- the roads. Such a system would cost mon- ties would be incurred by moving vehicles, ey to administer, but as with local council rather than by parked vehicles. The moral taxes or road taxes, such payments would basis for issuing such penalties is straight- be used to provide a roadway congestion forward. Taking a car onto a congested control service, not for the right to use a roadway worsens the speed performance fair share of the roads, as such. Conges- of that roadway for all. It is contributing to tion offsets do not require the loss of pub- the despoiling of a public good. It is gener- lic ownership of the roadway commons, a ally accepted that it is immoral to despoil key political weakness of road pricing. a public good by taking more than your fair share of it. Immoral does not become moral just because everyone is doing it. 60 World Transport Policy and Practice Volume 18.4 Sept 2012 Congestion offsets treat people as citizens, more. A person might drive all month in not customers. Citizens are expected to a congestion-prone area and only be fined behave properly and to, amongst other once during that month, but if that fine things, not wantonly despoil public goods. is a significant percentage of the - aver For those who do not behave properly, age monthly wage, behaviour will soon socially agreed sanctions are imposed. A change. Congestion offsets do not require penalty approach is no different in prin- precise metering of roadway use, a key or- ciple to controlling speeding; we do not ganisational weakness of road pricing. have a market for speeding, we just pe- nalise it sufficiently to reduce it to a social- To bring the above two points together; ly acceptable level. It is currently the case congestion offsets punish people who that driving a car onto a congested road is cause congestion and, in equal measure, a widely accepted practice; a person who reward people who alleviate congestion. would be horrified to see someone drop Unlike road pricing, money is collected via litter in a public place, does not think twice fines for misbehaviour rather than a fee to about driving onto a road that they know purchase a service. Also unlike road pric- to be congested. Attitudes have changed ing, the money collected via fines is explic- in the past though; drink driving used to itly paid to those helping to reduce conges- be widely socially tolerated, smoking in tion, in the form of cheaper bus services. public buildings used to be legal in many Figure 8 shows mock-ups of how these places where it is now banned. Attitudes fines and payments might look. The traffic can and do change. infringement shown in Figure 8 would, on average, be issued once for every 50 of Penalising people for “Causing Congestion” the bus tickets issued, thus both penalis- has a major organisational advantage over ing and rewarding by the same amount, charging a fee for a service, namely that namely $100. people do not usually complain when society fails to punish them for their mis- The approach shown in Figure 8 is one of behaviour. Unlike prices, fines are fairly offsets, thus the congestion offsets name. randomly imposed, with ‘speed traps’ and Due to widespread public discussion of ‘booze buses’ catching people sometimes, climate change, the best known form of and not at other times. With fines, there is offsets are carbon offsets. Organisations no need for the precise metering of road who emit carbon dioxide as a result of pricing. As with parking fines, a city gov- their operations, who are said to have a ernment could just erect signs, citywide, certain ‘carbon footprint’, may choose to saying that in these places, at these times, eliminate that footprint by buying carbon congestion is a problem and so such-and- offsets, thus becoming ‘carbon neutral’. such a penalty will be imposed. These The money paid for carbon offsets is used penalties could be enforced sufficiently to plant trees, or to fund other activities to achieve the desired outcome, and no that remove carbon dioxide from the at-

Figure 8: Road Overuse Fines and Under-use Payments 61 World Transport Policy and Practice Volume 18.4 Sept 2012 mosphere, with the outcome being no net privacy, or as a fine, for those who value carbon dioxide entering the atmosphere. privacy more than determinacy. It would Congestion offsets are the same concept, be likely that most commercial vehicle op- except rather than planting trees to re- erators, including freight operators, would move carbon, people use buses to remove opt for the determinacy of a fee. Under congestion. Importantly, unlike the case congestion offsets, charging the fine as a of carbon offsets, where no single govern- fee would be an auxiliary measure, avail- ment controls the atmosphere, cities nor- able on a vehicle-by-vehicle basis, with mally operate as a single political area and fines as the default mechanism. so congestion offsets could be mandated for all by a single government. It should be noted that a determinis- tic outcome could be achieved without a The defining characteristic of an offset is high-tech, live-location metering solution. that it enables something negative to be We have a well-established technology to wholly undone. With carbon, if enough off- deal with outlier risks, namely insurance. sets are purchased, an organisation’s car- It comes with an administrative overhead, bon emissions can be wholly undone, and and the unpleasant but necessary need for they can become carbon neutral. Many fraud prevention measures, but it works. types of bad behaviour cannot be undone. In a similar vein, sophisticated technol- For example, when a person parks in a ogy such as that used in red-light cameras clear-way, causing two lanes of a road to would no doubt be very useful in issuing drop to being only one lane, great incon- congestion fines, but the task could also venience is caused. The fine received for be done by a police officer with an in- this type of bad behaviour does not undo fringement pad. Likewise, super-computer the damage; it is purely a punishment, a modelling to determine optimal congestion deterrent. By contrast, congestion offsets offset levels would no doubt be beneficial, are a way of ensuring, in effect, that peo- but crude try-and-adjust methods would ple only take their fair share of road space, achieve an acceptable outcome. Today’s by responding to any attempt to take more high-technology — live location tracking, than a fair share in a way that brings road automatic fine cameras, super-computer use back under an agreed traffic level. modelling — would be very useful in re- Under a system of congestion offsets, if ducing cost and increasing efficiency, but people wish to take what would be more such technologies are not necessary to than their fair share of road space they are deploy congestion offsets. We could have required to sufficiently fund bus services deployed congestion offsets at any time such that the unfair portion of road space since the invention of the automobile; pa- they are taking becomes available. Con- per fines, insurance and crude try-and-ad- gestion offsets would require all road us- just modelling methods are all technolo- ers to be congestion neutral. gies that pre-date the automobile.

With the damage caused by congestion Living for so long without congestion con- being wholly unwound, congestion offsets trol has led to the most perverse results, could be optionally charged as, in effect, something that can be best demonstrated a fee, instead of a fine. Take the example by considering the alternative. Imagine if previously shown in Figure 8 where a con- it had been normal practice for towns to gestion fine of $100 is levied, but where the introduce a system of congestion offsets fine is incurred, on average, only once out a generation or so before they had any of every fifty times that a given road sec- chronic congestion problems. Car drivers tion is driven on. In this example a $2 con- in such towns would, from time-to-time, gestion fine could alternatively be charged receive road space use notices, stating each time, with GPS and mobile communi- that on a given date and time they were cation technology now making this easily recorded using one road slot when, say, achievable. Indeed, many commercial ve- two road slots were available; giving a hicles are already fitted with live location roadway overuse level of zero slots. This tracking equipment. Congestion offsets notice would record that the fine for over- could be charged as, in effect, a fee, for using zero slots is zero dollars, and thank those who value determinacy more than the driver for staying within the available 62 World Transport Policy and Practice Volume 18.4 Sept 2012 road space. Similarly, on the small bus though, should private vehicles get to use services run by such towns, bus tickets up all this space when there are other us- would record that passengers had under- ers of street space? Why not have more used zero of the contested slots, due to a bike lanes or wider footpaths? The same surplus of one slot per vehicle, and that point could be made for space use within they were receiving a discount of zero dol- a single road lane. Throughout this paper, lars on their tickets. all calculations have been made on the assumption that all of the capacity of a road During the twenty or thirty years before lane was to be used by vehicles. If a road these towns had sufficient traffic to cause lane could take 700 cars, it did take 700 chronic congestion, no fines or discounts cars, and without congestion control this is would be issued, but both car and bus us- indeed what happens. A street with more ers would see, as their towns grew, the traffic, though, is a worse street for eve- excess road slots slowly being used up. ryone using it except the vehicle users.32 For car drivers this would draw them clos- Under a system of congestion offsets it er and closer to the day where they were would be possible to consciously run road going to start being fined for their -road lanes significantly below their maximum way overuse. By contrast, the diminishing capacity. In the absence of congestion number of free roadway slots would draw control, many of our cities have become bus passengers closer and closer to the awash with both traffic and bitumen. day where they were going to start being paid for their roadway under-use. One day, One option for congestion offsets would be which people would have seen coming for to implement them in a minimal way; to decades, car drivers would start receiving eliminate chronic congestion, but to still occasional small fines for their roadway provide as much mobility as possible using overuse, and bus passengers would start cars. This would be an improvement over receiving tickets which stated that they the current situation. Alternatively, con- had underused a tiny percentage of the gestion offsets could be used in a trans- contested road slots, and that they were formative way; to slowly squeeze a great therefore receiving a discount of one cent deal of both the traffic and the bitumen on their tickets. Over time, both the size out of our cities. Such an approach would of the fines being issued to drivers and the see chronic congestion become a thing of size of the bus ticket discount would in- the past, indeed many roads may be run crease. Buses would slide into greater use well below capacity, so as to balance the as people found their share of the roadway needs of vehicle users and other street us- commons dropping in line with urbanisa- ers. Many road widening projects would tion. There would be a slow increase in most likely be deemed no longer neces- public transport use, simply as a matter sary, and all existing road grades above of rational decision-making. Chronic con- two lanes, one in either direction, could be gestion would never be a feature of such reviewed to see if the extra lanes were re- towns’ transport planning landscape, as ally necessary, or whether the land could they grew from towns to cities. be put to better uses. Mass mobility could be provided by a grid of high frequency, Would a city with such a history ever talk BRT-style bus services running on normal about building large arterial roads simply congestion-free roads. Under such a sys- to carry car traffic, or heavy rail lines in tem most people would travel on buses places where the passenger load simply most of the time, with cars playing an im- did not justify the construction of such portant, but far smaller role than they do lines? It seems most unlikely. Far more now. In short, congestion offsets could be likely, would be that both heavy road and used to help rehabilitate our cities from heavy rail infrastructure would be built their current degraded state, where they sparingly, and only when circumstances are heavy with both traffic and bitumen. fully justified such construction. Not so Congestion offsets could be an important in our cities today; once two lanes, one new tool to help transform our cities. in each direction, are no longer sufficient to soak up traffic, four lane roads, or even heavier grade roads are built. Why, 63 World Transport Policy and Practice Volume 18.4 Sept 2012 With congestion offsets now described, Also, on the penalty side of the penalty/ the question raised in the introduction as reward equation, road users in London are to whether congestion offsets constitute not fined for their misbehaviour of overus- a distinct congestion control method can ing the roads; they are instead offered the be revisited: are congestion offsets just chance to purchase access to road space. a variant of road pricing? From a concep- The whole behavioural mechanism, the tual standpoint, road pricing and conges- reward/penalty system and therefore psy- tion offsets are clearly different congestion chology of what this paper has proposed control methods. The central concept be- is very different from road pricing, even in hind congestion offsets is to keep road- cases where road pricing has been coupled ways as a commons, but to regulate that with the funding of public transport. commons better, whereas the central concept behind road pricing is to enclose the Finally, there is the impact that using fines, roadway commons, with the road space instead of fees, has on ease of deployment. then being ‘parcelled up’ for sale. Even This fines-versus-fees difference between with congestion offsets being a conceptu- congestion offsets and road pricing stems ally distinct congestion control method it directly from the fact that, unlike road is nevertheless possible for two conceptu- pricing, congestion offsets have a built-in ally distinct methods to yield very similar notion of what constitutes a fair share of policy prescriptions. If there are no signifi- road space, and can therefore advance the cant differences between the policy pre- proposition that people who wantonly de- scriptions of congestion offsets and those spoil a public good by taking more than of road pricing, then it would be difficult to their fair share of it should be punished, maintain a position that congestion offsets via a fine. Fines can be randomly imposed, are a distinct form of congestion control. fees cannot. The fact that fees cannot be randomly imposed saddles road pric- In cases where road pricing does not come ing with the requirement for precise me- packaged along with policy measures to tering, which has drastically restricted its funnel raised funds towards public trans- deployment, usually to small central city port, congestion offsets are clearly a dis- areas, even when congestion is a citywide tinct form of congestion control; conges- problem. This different policy prescription, tion offsets take funds from road users the use of fines instead of fees, yields a and also give funds to road users, where- concrete and highly beneficial outcome; as road pricing only takes funds from road it allows congestion control to be easily users. This is a very significant difference deployed citywide. Congestion offsets are in terms of policy prescriptions between not a variant of road pricing; they are a congestion offsets and road pricing. But distinct method of congestion control. what about circumstances where road pricing does come packaged along with For readers who remain unconvinced, who policy measures to funnel raised funds to- continue to view congestion offsets as wards public transport, such as in London. simply an aggressive repackaging of road Are congestion offsets a clearly different pricing into a more ‘palatable’ form, these form of congestion control in comparison readers are invited to consider the pos- to “road pricing plus public transport fund- sibility that an aggressive repackaging of ing?” Putting aside the fact that we are no road pricing into a more ‘palatable’ form longer talking about road pricing, as such, is, in fact, desperately needed. In terms but road pricing with an auxiliary policy of the number and scope of deployments, bolted on the side, there are still key dif- road pricing has been a spectacularly un- ferences in how road pricing and conges- successful policy. The idea of road pric- tion offsets take funds from road users ing has been around for decades, during and give funds to road users. Although the which time congestion has grown to be money raised from road pricing in London an enormous worldwide problem, and yet is ploughed back into improving their pub- only a handful of cities have adopted road lic transport system, one is not reminded pricing schemes. In a continuation of this every time one gets on a bus, via an ex- trend, a road pricing scheme for New York plicit financial reward, that one was under- has been recently shelved by the New York using his or her fair share of road space. State Assembly.33 The normal response, 64 World Transport Policy and Practice Volume 18.4 Sept 2012 when a ‘product’ fails to ‘sell’ well over an The urban transport problems of cities, extended period of time, is to redesign the no matter how dire, are, of course, only product into something that people actual- one part of the bigger urban sustainability ly want to buy. Instead, those advocating picture. However effective congestion off- road pricing appear to be doggedly trying sets might be in addressing the particular to sell something that, the evidence sug- sustainability problems of urban transport, gests, not many people want to buy. The they would, on the whole, be an ineffective difficulties inherent in this course of action tool for addressing the broader sustaina- are highlighted by one of the questions to bility problems of cities. Tools are normally be discussed at an urban congestion con- only any good at solving the problem they ference scheduled to be held some months were designed to solve, and congestion after the finalisation of this paper. One offsets are a tool that has been designed panel at this conference, somewhat sur- to solve transportation problems, not prisingly, plan to, amongst other things, wider social, economic or environmental engage themselves with “delving deeper problems. Introduce a system of conges- into the road pricing scheme and answer- tion offsets in a socially unjust, economi- ing the ultimate question of whether it cally stagnant and environmentally ruined is a death wish or real solution?”t4 That city that also happened to have terrible congestion control through traditional traffic problems, and the likely outcome road pricing approaches could possibly be would be a socially unjust, economically equated with a “death wish”, gives con- stagnant and environmentally ruined city siderable pause to thought and highlights that no longer had terrible traffic prob- the difficulties faced in trying to introduce lems. Ways of addressing these wider so- road pricing in its present form on a large cial, economic and environmental issues scale in cities. The authors suggest that have been extensively described by oth- this course of action should be reconsid- ers, and the broader sustainability of cit- ered in light of the demonstrated history ies, their capacity to become regenerative of its results, and the lack of any solid evi- in a wider sense, rests on a broad array of dence suggesting that there might be a issues and strategies for action.35 significant improvement of those results in the future. Conclusion

Whether the approach presented in this A major part of the urban transport prob- paper is called congestion offsets, or road lem today is a failure from the very be- pricing with fairness built in, or social road ginning to acknowledge that congestion is pricing, or road pricing 2.0, is not a mat- fundamentally inequitable and unfair, im- ter of great consequence. As the previous practical to construct away, and therefore section of this paper detailed, the urban must be properly charged for and control- transport situation is far worse than gen- led to eliminate the transport system dys- erally acknowledged. Buses on normal function which is systemic in cities today. roads, an entire transportation mode, and By permitting free access to roadways one that cannot be effectively substituted as a matter of policy, governments have for, has been crippled by congestion, with allowed cars to cripple the natural com- disastrous results. This terrible situation is petitor for mid-level loads, namely bus a matter of great consequence, and one services on normal roads. Cars dominate that will only be resolved by the wide- urban transport in most of the world not spread deployment of an effective form of due to any inherent performance superior- congestion control. Consequently, there is ity, but because they have what amounts a desperate need for a congestion control to a government-granted monopoly. In the method that works, that is easy to deploy absence of an effective mid-level transport and that has the concept of fairness built mode, the next best option for transport into its fundamental design principles, so planners has been to try and make low that it has some hope of being acceptable and high load modes look like good mid- to the general public. Road pricing meets level modes. To this end, the load capac- one of those three requirements, namely ity of roads has been greatly increased via that it works; congestion offsets meet all road building projects and the load capac- three. ity of railways has been greatly reduced by 65 World Transport Policy and Practice Volume 18.4 Sept 2012 operating them well under capacity. Run- our cities, in the absence of such a mode, ning railways well under capacity at least could be slowly unwound. Lean roads on has the advantage of only costing cities multi-purpose streets could become the money. Greatly increasing the load capac- norm, as they should have always been. ity of roads via road building projects has Simply by deciding to let buses compete, blighted our cities, leaving many of them we could help to transform our cities in bloated with both traffic and bitumen. significant ways by addressing two key issues: the modal split between private mo- As a solution to the urban transport prob- torised and public motorised modes and lem, road pricing has been near univer- the endemic congestion that plagues cit- sally rejected, and for good reason. Road ies. Like the great freeway project this is a pricing requires precise metering, making big dream, but our cities are our collective it difficult to deploy, and disenfranchises homes and we should dream big dreams the public from the roadway commons, about them. Congestion offsets could be a making it all but impossible to sell politi- key tool in helping to shift cities towards a cally. Road pricing is, in short, a complex new, regenerative model. and unfair form of congestion control. Author details: Congestion offsets have been presented here as an alternative to road pricing. In brief, congestion offsets involve penalis- [email protected] ing private vehicle users for overusing the roads and rewarding public bus users Jeff Kenworthy for under-using them. The design princi- Curtin University Sustainability Policy In- ple behind this approach is that roadway stitute (CUSP) congestion should be addressed not by Curtin University of Technology constructing a market, but by regulating Perth, Western Australia a commons. Under congestion offsets, a person who, on average, uses only their [email protected] fair share of roadway space would not incur any direct net costs. Even in the case of 1 Jacobs, J. (1961) The Death and Life of overuse, there is no transfer of value out- Great American Cities. Random House. New York. side of the users of the roads, with value Schneider, K. (1979) On The Nature of Cit- ies: Toward Enduring and Creative Human Environ- simply being transferred from over-users ments. Jossey-Bass Publishers, San Francisco. to under-users. Unlike road pricing, con- 2 Kunstler, J.H. (1994) The Geography of No- gestion offsets do not disenfranchise the where: The Rise And Decline of America’s Man-Made Landscape. Touchstone. New York. public of the roadway commons. Also fol- Stretton, H. (1975) Ideas of For Australian lowing from the design principle that con- Cities. Georgian House. Melbourne. gestion should be addressed by regulating 3 Kenworthy, J. (2012) Don’t Shoot Me, I’m Only The Transport Planner (Apologies to Sir Elton a commons, congestion offsets ‘lean’ on John). World Transport Policy and Practice, Issue roadway over-users not by charging them 18.4. 4 Ibid. for roadway space, as such, but by punishing them for their misbehaviour via fines. 5 Poole, R.W. (Jr) Resolving Gridlock in Southern California. Transportation Quarterly 42 Consequently, unlike road pricing, conges- (4), 499-527. tion offsets do not require precise meter- Orski, C.K. (1987) “Managing” Suburban ing. Congestion offsets are, in short, a Traffic Congestion: A Strategy for Suburban Mobility. Transportation Quarterly 41 (4), 457-476. simple and fair form of congestion control. Cervero, R. (1984) Managing the Traffic Impacts of Suburban Office Growth. Transportation Quarterly 38 (4), 533-550. Better cities are possible. A tolerable ac- Gleick, J. (1988) National Gridlock: Sci- commodation with congestion is not the entists Tackle the Traffic Jam. The New York Times only option. Congestion offsets, a simple Magazine, May 8. Pratsch, L.W. (1986) Reducing Commuter and fair alternative to road pricing, would Traffic Congestion. Transportation Quarterly 40 (4), allow congestion to be squeezed out of our 591-600. cities, and for bus services to emerge as an efficient and effective mode for servicing mid-level transport loads. The massive distortions we have allowed to build up in 6 Goodwin, P. (1997) Solving Congestion 66 World Transport Policy and Practice Volume 18.4 Sept 2012 (when we must not build roads, increase spend- Table 3-16. ing, lose votes, damage the economy, or harm the 26 Vuchic, V. (2007) Urban Transit Systems environment, and we will never find equilibrium). and Technology. John Wiley & Sons, Inc. Hoboken. Inaugural lecture for the professorship of transport New Jersey. p. 76. policy. University College London, 23rd October 27 Kenworthy, J. (2012) Update of The Mil- 1997. lennium Cities Database for Sustainable Transport. 7 Larson, T. D. (1988) Metropolitan Conges- Unpublished. tion: Towards A Tolerable Accommodation. Transpor- 28 It could be argued that the current Certifi- tation Quarterly. 42 (4), 489-498. cates of Entitlement systems in operation in Singa- 8 The approach shown in Figure 1 draws pore and Shanghai, run as monthly auctions for the significantly on material from the Travel Time Equa- right to buy a car, with the number of certificates tions and Diagrams section of Professor Vukan Vu- issued being set roughly to the attrition rate of old chic’s Urban Transit Systems and Technology book. cars, are an indirect form of congestion control. The 9 Kenworthy, J. (2012) Update of The Mil- Singaporean system has been successful in keeping lennium Cities Database for Sustainable Transport. car ownership at about 100 cars per 1000 people, Unpublished. despite a GDP per capita that would imply much 10 Wright, L., Hook, W., (eds) (2007) Bus higher ownership. This clearly limits the growth of Rapid Transit Planning Guide. Institute for Transpor- congestion in the medium to long term. The city of tation & Development Policy. New York. pp. 767- Beijing has also introduced basically the same sys- 778. tem, but runs it as a lotto scheme each month, so 11 Dittmar, H., Ohland, G.(eds) (2003) The it is pure luck as to who gets a ticket to buy a car. New Transit Town: Best Practices In Transit-Orient- These approaches are, though, such blunt policy ined Development. Island Press. Washington, p. 20. struments that they will not be considered further 12 Wright, L., Hook, W. (eds) (2007) Bus here. Congestion could also be controlled by banning Rapid Transit Planning Guide. Institute for Transpor- cars from cities outright. tation & Development Policy. New York. pp. 767- 29 Wessels, W. J. (2006) Economics: Barron’s 778. Educational Series (4th Edition). Barron’s Education- 13 U.S. Department of Transit, Federal Transit al Books. New York. p. 2. Administration (1992) Characteristics of Urban 30 Santos, G. (ed) (2004). Road Pricing: The- Transportation Systems. Chapter 3: Bus Transit. ory and Evidence. Research in Transport Economics: Table 3-15. Vol. 9. Elsevier Press. The Netherlands. 14 Kenworthy, J. (2012) Update of The Mil- 31 Although road pricing often comes pack- lennium Cities Database for Sustainable Transport. aged along with policy measures to funnel some or Unpublished. all of the funds raised towards public transport, such 15 Mees, P. (2000) A Very Public Solution: measures are auxiliary to the central mechanism of Transport in the Dispersed City. Melbourne Univer- road pricing; the commodification and sale of road sity Press. Victoria, Australia. pp. 138-142. space. By contrast, the discounting of bus tickets is 16 U.S. Department of Transport, Federal an integral part of congestion offsets, with the issuing Highway Administration (2009) Summary of Travel of discounts for roadway under-use being the mirror Trends: 2009 National Household Travel Survey. p. policy to charging for roadway over-use. Congestion 10. Table 3. Average person trip length, 2009. offsets have a built-in concept of what constitutes a 17 Wright, L., Hook, W. (eds) (2007) Bus fair-share of road space, and consequently are able Rapid Transit Planning Guide. Institute for Transpor- to have built-in ways of dealing with (cont.) tation & Development Policy. New York. pp. 767- roadway use both above and below that fair amount. 778. Congestion offsets are ‘carrot’ and ‘stick’ in equal 18 Kenworthy, J. (2012) Update of The Mil- measure. As the name would suggest, road pric- lennium Cities Database for Sustainable Transport. ing is a ‘stick’ only mechanism, which is why it is so Unpublished. commonly found packaged with ‘carrot’ measures, 19 Vuchic, V. (2007) Urban Transit Systems including the funnelling of raised funds towards pub- and Technology. John Wiley & Sons, Inc. Hoboken. lic transport. The frequency with which road pricing New Jersey. p. 77. is found packaged along with the funding of public 20 U.S. Department of Transport, Federal transport, suggests the need for a congestion control Highway Administration (2009) Summary of Travel method, such as congestion offsets, whose funda- Trends: 2009 National Household Travel Survey. p. mental design principles naturally encompass both 33. Table 3-16. All Purposes, 2009. ‘carrot’ and ‘stick’ measures. 21 U.S. Department of Transit, Federal Transit 32 Appleyard, D. (1981) Livable Streets. Uni- Administration (1992). Characteristics of Urban versity of California Press. Berkeley. California. pp. Transportation Systems. Chapter 3: Bus Transit. 15-28. Table 3-16. 33 New York Assembly Shelves Manhattan 22 Even using an average car occupancy Congestion Charge, The Guardian Website, Avail- level of 1.2 people per car, a typical figure for peak able: http://www.guardian.co.uk/politics/2008/ hour use, which would result in a load level of 840 apr/08/congestioncharging.pollution (Accessed 20 pph, one could still all but fill ten standard buses to August 2012) capacity every hour, delivering a six minute service 34 Tonkin’s Urban Congestion Conference, The frequency, better than most cities provide today. Tonkin Corporation Website, Available: http://www. 23 Transport Research Board (2000) Highway tonkincorporation.com/images/Eureka/ENG40E.pdf Capacity Manual 2000. National Research Council. (Accessed 30 August 2012) Washington D.C. Exhibit 21-8: Passenger-Car Equiv- 35 Newman, P. and Kenworthy, J. (1999) alents on Extended General Highway Segments. Sustainability and Cities: Overcoming Automobile 24 Ceder, A. (2003) Public Transport Timeta- Dependence. Island Press, Washington, DC. bling and Vehicle Scheduling. In: Lam, W.H.K., Bell, Schiller, P. Bruun, E. and Kenworthy, J. G.H., Advanced Modeling for Transit Operations and (2010). An Introduction to Sustainable Transporta- Service Planning. Elsevier Science Ltd. Pergamon. p. tion: Policy, Planning and Implementation. Earths- 37. can, London. 25 U.S. Department of Transit, Federal Transit Newman, P. and Jennings, I. (2008) Cities Administration (1992) Characteristics of Urban as Sustainable Ecosystems: Principles and Practices, Transportation Systems. Chapter 3: Bus Transit. Island Press, Washington DC. 67 World Transport Policy and Practice Volume 18.4 Sept 2012 Advertisement

NEW BOOK

Whitelegg, J “Quality of life and public management: redefining development in the local environment” Routledge http://www.routledge.com/books/details/9780415509558/ Hardback: 978-0-415-50955-8:

Abstract This book is a no-excuses manual for a significant upgrade in quality of life for all citizens wherever they live. It takes a hard look at Britain’s record on child poverty, elderly poverty, public transport, walking and cycling, air pollution, road safety, renewable energy and resilience and concludes that things are bad. The book is hard hitting. There is no excuse for poor quality outcomes in the 6th rich- est country in the world. Each chapter in the book reveals that there are many places where outcomes are excellent (but not in Britain) and if high quality performance can be found in Sweden, Germany, Switzerland or Denmark then it can also happen in Britain. The reasons for Brit- ain’s failure are the same as in Australia or the USA and are embedded in an ideology and culture that worships market forces downgrades the importance of local government and underfunds public services. In Britain there are major failures that are in need of urgent correction. Brit- ain is too centralised and local government is weak, unambitious and largely content with treading water, numbing bureaucracy and lack of vision. The book sends out a strongly positive message and spells out in great detail just what has to happen to improve things. Everyone can live in a high quality environment with well funded public services. There are no technical, organisational or financial reasons why an excellent quality of life cannot be designed and delivered. The missing ingredient is political will and this will only change when we change our politicians. Contents: children, older people, renewable energy, air pollution, built environment, resilience, powers of local; gov- ernemnt, reform of local government.

68 World Transport Policy and Practice Volume 18.4 Sept 2012