Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The history of autonomous vehicle development and its likely futures and consequences
(USC Thesis Other)
The history of autonomous vehicle development and its likely futures and consequences
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE HISTORY OF AUTONOMOUS VEHICLE DEVELOPMENT AND ITS LIKELY
FUTURES AND CONSEQUENCES
by
Alanna Coombes
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(URBAN PLANNING AND DEVELOPMENT)
December 2024
Copyright 2024 Alanna Coombes
ii
Dedication
For Lewis de Luca and Gus Coombes Rawson
iii
Acknowledgements
I will be forever grateful to my husband who has enabled me to pursue my ambition and
passions while sacrificing many of his own along the way. And it’s been quite a journey!
I am incredibly grateful to my adviser at USC, Professor Genevieve Giuliano, for providing
me with the time to learn from her in thinking, writing and teaching. Her guidance and patience
have been extremely helpful. I’m very grateful too to Professors Peter Norton (University of
Virginia) and David Sloane (USC), my other committee members, for their strong interest,
guidance and support. I have been extremely fortunate to have such a strong committee to guide
and strengthen my dissertation.
I owe much gratitude too to Price School at USC and the professors there notably
Professors Dowell Myers and Marlon Boarnet. I’m grateful too to the Annenberg School at USC
and especially to Professors Peter Clarke and Sheila Murphy who inspired me and grounded me in
communications learning.
I’m grateful too to the City of London Corporation and my manager there, Paul Beckett,
who encouraged and enabled me to pursue my interests. Without his support I may never have
embarked on what has been an incredible adventure.
iv
Table of Contents
Page
Dedication ………………………………………………………………………………………………………… ii
Acknowledgements …………………………………………………………………………………………… iii
List of figures ………………………………………………………………………………………………….… vi
Abstract ……………………………………………………………………………….………………………..… vii
Chapter 1: Introduction ………………………………………………………………………………..….… 1
Chapter 2: Methodology ……………………………………………………..……………………….….… 8
Chapter 3: Theoretical Framework ………………………………………………………………..…… 15
Chapter 4: The history of transportation technology development: a literature
review
22
The Automobile ………………………………………………………….…………………….…....… 23
Highway-building ………………………………………………………..…………………………... 32
TNCs ……..………………………………………………………………………………………………... 47
The gap between hopes, predictions, expectations and reality ….………………..… 67
Synthesis and implications ……………………………..…………………………………………. 87
Chapter 5: The Path to AVs ……………………………………………………….……………………..… 90
A brief history of AV development ………………………………………….………………….. 90
Actor involvement ……………………………………………………….……………………….…… 99
Promotion of AVs ………………………………………………………..…….……………………… 117
Summary - echoes of the past …………………………………………………………………… 129
v
Chapter 6: Case Studies ………………………………………………………………………………….…. 131
Case study methodology ………………………………………………………………….……….. 132
San Francisco, CA …………………………………………………………………………………….. 133
Phoenix, AZ ……………………………………………………………………………………………… 189
Chapter 7: Analysis …………………………………………………………………………………………… 218
Chapter 8: Conclusion ……………………………………………………………………………………….. 245
Appendix 1: Survey Instrument ……………………………………………………………………….….. 255
Bibliography ……………………………………………………………………………………………………… 257
vi
List of Figures
Figure 1 Mobility confidence index attributes – level of comfort with AVs ………. 116
Figure 2 San Francisco coverage area for Waymo …………………………………………. 153
Figure 3 Cruise’s expanded service area May 2022 ……………………………………….. 154
Figure 4 Human drivers vs Cruise ………………………………………………………………… 162
Figure 5 Updated human ride-hail benchmark vs Cruise AVs in 1m ………………… 163
Figure 6 Comfort Level Sharing the Road And Interacting with Fully Automated,
Self-Driving Robotaxis: Non Rider ………………………………………………….. 176
Figure 7 Comfort level with AV technology being tested on streets and
highways near you ………………………………………………………………………… 177
Figure 8 Income of highest and lowest earning households in the Bay Area,
California and the United States as a whole between 2010 and 2019 … 178
Figure 9 Increasing ration between the highest and lowest 20% of earners in San
Francisco ……………………………………………………………………………………… 178
Figure 10 Waymo’s operating area in Phoenix Metro ………………………………………. 203
Figure 11 Cruise’s operating area in Phoenix Metro in 2023 …………………………….. 204
Figure 12: Cruise advertisement April 2023 …………………………………………………….. 229
vii
Abstract
Transportation technologies of the past have been presented as solutions to one or more
of the continuous challenges facing cities, especially improving road safety, reducing congestion
or the environmental impact of vehicles. Such technologies though have failed to meet the hopes
and expectations placed upon them. We have seen continued rising numbers of road deaths,
rising congestion and continued environmental costs. This dissertation explores why
transportation technologies have largely failed to solve these problems and how lessons of the
past may explain and predict outcomes for the current “new” technology, automated vehicles.
Past technologies have failed in part because those who developed and promoted them
prioritized goals other than mobility – their profits, their careers, or to find a fix for other social or
economic problems – all to the end of selling the technology. Transportation technology has been
promised to make life easier and more comfortable, generate wealth, create jobs, strengthen
economies, along with making mobility safer, more efficient, and more environmentally friendly.
Despite a century of failure, the public continues to believe that a world of safe, congestion-free,
easy, and clean automobility is continually just around the corner. An ecosystem of people and
organizations that benefit from the sale of cars continue to provide an illusion that it is.
This dissertation describes the history of the development of autonomous vehicles (AVs)
and conducts two case studies, Phoenix, AZ and San Francisco CA. Using the lenses of theories of
technology adoption and the history of prior transportation technologies, I find that the same
dynamics and motivations of the past are playing out today, and they are coming from a similar
coalition of interests. Claims for autonomous vehicles echo those used previously and are equally
misleading. Negative impacts are overlooked, and losers are likely to be those who have lost out
in the past. I conclude with some suggested ways forward. If developed, autonomous vehicle
technology has the potential to reduce road crashes and save lives but unless we learn the lessons
viii
of the past and have an open and honest appraisal of AVs’ likely impact in the future, we will find
ourselves driving in circles.
1
Chapter 1: Introduction
1
“I saw in their eyes something I was to see over and over in every part of the nation- a burning
desire to go, to move, to get under way, anyplace, away from any Here. .. Nearly every American
hungers to move.”
John Steinbeck in Travels with Charley
Motivation for this dissertation
My first car was a 1973 Triumph Spitfire IV in ‘saffron yellow’ – a convertible with a long
hood out front and a rear nicely in balance in a Jaguar E Type Roadster kind of way. I loved that
car but there were elements of an abusive relationship; it was indifferent to my feelings and the
care I gave it - and wasn’t capable of loving me back. Although I looked after its health and wellbeing, it was negligent towards mine; the alternator failed when I was alone late at night outside
Chesterfield; the petrol tank ripped outside Newport Pagnell; and the steering wheel detached
from the steering column outside Huddersfield.
My love for classic cars – and avoidance of premature death - has stayed with me and
has more recently been combined with an interest in smart city technologies. This developed when
I led the smart city program for the City of London Corporation during which it seemed parts of
the planning profession were quite giddy with the prospect of new technology and Doing Things
with all the data they hoped to collect. I learnt about autonomous vehicles and my interest
deepened. In part my interest was in the technology, but in part it was in planners’ and
politicians’ response to it.
1 Steinbeck, J. (2004) Travels with Charley; in search of America. The Folio Society, London. p9
2
Having worked directly for politicians in political advice and public affairs roles I
recognized there can be a wide gap between the stated public reason for supporting something
and the complexity of real reasons. As I looked more at the history of transportation technologies
there seemed to be repeated cycles of claims and disappointments – very well described in
Gartner’s hype cycle2
. With some dismay I reflected on previous transportation technologies that
were thought to be the Holy Grail of mobility – such as Transportation Network Companies (TNCs)
such as Uber - and I wondered if this is how it had to be. Why were we so often disappointed?
Were there warning signs that should have alerted us that we were heading down the wrong
road? Are autonomous vehicles likely to be just another of a long line of over-hyped
transportation technologies that render city mobility in the same sorry state so many cities now
experience?.
Embarking on my PhD and this dissertation I hoped to find a way to address these
questions and provide a positive, proactive way forward for cities. There seemed to be a time
imperative too because of the importance of scrutinizing technologies at an early stage before
they become just another fact of life - there are limits to what can be done once technologies
become embedded3
’. My hope is that it is still possible to shape autonomous vehicle (AV)
technology so that it helps to improve mobility and city livability and that is why I think this work
is so important right now.
2 Gartner’s hype cycle is commonly used to describe and explain the pattern of rapid rush of enthusiasm for a
technology’s potential, followed by disillusionment in the face of real-world challenges. The cycle shows
technologies moving through stages from when the technology is launched (‘innovation trigger’) to when the hype
has dissipated and it’s clear how useful the technology has proved to be (‘Plateau of Productivity’). It also
estimates how long the innovation will take to reach that end point.
3
Stilgoe, J. (2020) Who’s Driving Innovation? Palgrave Macmillan, Switzerland
3
Definitions
‘Autonomous vehicle’ is a vehicle that can reach a destination without the need for human
input. It is a vehicle that can drive anywhere, under any conditions, without the need to have
traveled there before, and without any human input either from within the car or elsewhere.
Strictly speaking, ‘autonomous’ also means a vehicle that can think for itself (as well as drive
itself) and as one Nissan engineer has pointed out, “A truly autonomous car would be one where
you request it to take you to work and it decides to go to the beach instead4
.” But a shared
common language is important and the commonly accepted term of ‘autonomous’ tends to be a
vehicle that simply can drive itself with or without a safety driver – and won’t get you into trouble
with your boss.
‘Automated vehicles’ is a term sometimes used interchangeably with ‘autonomous
vehicles’ and in some ways is a better term because it enables gradation of the technology using
a well-recognized scale between 0 and 5. The Society of Automotive Engineers level indicates
Level 0 as having no features that operate outside the control of the driver; Level 1 which
automatically works low level functions such as the wipers or deploys anti-lock brakes - through
to level 5, which is full autonomy and frees-up all vehicle occupants from any contribution to the
driving task whatever the weather or environment presents. Most new cars have at least Level 1
automation; robotaxis – those vehicles that provide passenger services without a driver at the
wheel – operate at Level 4. Auto and tech companies – or AC companies (companies working to
develop AV technology) are striving to reach Level 5 but none yet exist5
.
4 Autotrader (2018, January 26) Automated vs. Autonomous Vehicles: Is There a Difference? Autotrader
https://www.autotrader.com/car-news/automated-vs-autonomous-vehicles-is-there-a-difference-273139
Retrieved July 23, 2024
5 NHTSA (n.d.) Levels of automation. NHTSA.gov https://www.nhtsa.gov/sites/nhtsa.gov/files/2022-05/Level-ofAutomation-052522-tag.pdf
4
Recently some auto and tech companies have withdrawn from the pursuit of Level 5
automation and instead focused on increasing automated features in conventional cars – Argo AI,
backed by the Ford Motor Company and Volkswagen Group, is a notable example but more
prominently, Cruise in December 2024. The focus on developing automated features, rather than
Level 5 (ie full) autonomy, could be seen as a reflection of AVs’ place in a hype cycle in the ‘Slope
of Enlightenment’: the dreamy vision has evaporated but some practical tools or benefits are
emerging from all that work.
Some auto or tech companies of course continue to work towards achieving the AV dream
and promise to turn the autonomous vision into a reality. Tesla’s ‘Full Self-Driving’ system, while
only operating at Level 2 automation6 has been the focus of court battles over mis-selling of its
technology, giving the impression it is more advanced than it is. This has presented dangers to
drivers facing complex and unanticipated situations with an unproven technology7 with
consequent crashes. More successful is Waymo, the Google-backed robotaxi company that
operates at Level 4 and has been more quietly and more modestly amassing many millions of
miles of autonomous driving experience.
The Significance of Autonomous Vehicles.
Cohen et al note, “Autonomous vehicles (AVs) have the potential to cause profound shifts
across a wide range of areas of human life, including economic structures, land use, lifestyles and
personal well-being8
”. This ‘potential’ is in part dependent on whether the technology can be
6 Nordhoff, S. (2024). A conceptual framework for automation disengagements. Scientific Reports, 14(1), 8654–
8654. https://doi.org/10.1038/s41598-024-57882-6
7
Linja, A., Mamun, T. I., & Mueller, S. T. (2022). When Self-Driving Fails: Evaluating Social Media Posts Regarding
Problems and Misconceptions about Tesla’s FSD Mode. Multimodal Technologies and Interaction, 6(10), 86-.
https://doi.org/10.3390/mti6100086
8
Cohen, T., Stilgoe, J., Stares, S., Akyelken, N., Cavoli, C., Day, J., Dickinson, J., Fors, V., Hopkins, D., Lyons, G.,
Marres, N., Newman, J., Reardon, L., Sipe, N., Tennant, C., Wadud, Z., & Wigley, E. (2020). A constructive role for
5
developed fully – which – as of December 2024 - it is not. Sure, robotaxis are providing driverless
rides to Angelenos and others right now but the technology is constrained by place, dependent on
remote intervention and not proven.
An alternative viewpoint is that once the novelty wears off AVs will normalize and things
may not change so very much. Because there will still be a market for conventional vehicles for a
long time we can expect no real changes in road layout or other perceived changes; by reducing
the cost of driving for some though (the AV users) we can expect more vehicles, more VMT and
more competition for road space.
Current state of autonomous vehicle technology
As of October 2024, there were 84 AV testing sites across the US hosted within eighteen
states. Thirty four of these testing sites were on streets and twelve on highways; the rest on
business campuses, parking lots or driveways, university campuses and other unspecified
locations9
.
The autonomous vehicles tested include cars, trucks, and small shuttles carrying eight to
ten people. Many include a human safety driver on board but some do not. Of the smaller
vehicles, Waymo and Cruise have been the biggest car AV companies and they have operated
with and without safety drivers. They have steering wheels and from the inside appear to be
regular cars. Shuttles, such as those tested by Zoox, tend to be pods without steering wheels or
pedals and travel at a slower speed.
social science in the development of automated vehicles. Transportation Research Interdisciplinary Perspectives, 6,
100133
9 NHTSA (n.d.) AV Tracking Tool. NHTSA.gov Retrieved December 15, 2024 https://www.nhtsa.gov/automatedvehicle-test-tracking-tool
6
None of the vehicles operating on roads today operate at Level 5 – that is without any
human input and able to travel anywhere, in any weather, without the need for prior mapping of
the territory. The most advanced are those of Waymo and (to a lesser extent, Cruise) that have
offered paid-for rides to the public with ‘robotaxi’ services.
Robotaxis are however not as autonomous as they may appear. They appear to need no
human intervention save for those requested by passengers who are able to connect with an
operator when they are in the vehicle should they need assistance (such as because the vehicle
won’t move). Evidence suggests however that a human intervenes in the driving of these Level 4
‘robotaxis’ regularly. Following an incident in San Francisco in October 2023, it was revealed that
Cruise was dependent on a team of remote operators to intervene regularly. There were 1.5
workers dedicated to each Cruise vehicle and they were intervening every 2.5-5 miles10
.
The AV Vision
The vision of autonomous vehicles has been of a safe, clean, easy, congestion-free, driving
experience enjoyed by relaxed passengers gliding along sparsely filled roads. It’s a vision that’s
been revisited regularly for around 100 years with increasing levels of detail, technological knowhow and plausibility. Each time, autonomous vehicles (AVs) are presented as a ‘silver bullet’ - a
simple solution to the seemingly intractable problems of road accidents, congestion and the
environmental externalities of transportation. Whether such a ‘silver bullet’ – a seemingly magical
solution to a complex problem – is capable of being developed fully is the nub of on-going work
of AVs.
10 Mickle, T., Metz, C., & Lu, Y. (2023). G.M.’s cruise moved fast in the driverless race. it got ugly. New York: New
York Times Company. https://www.nytimes.com/2023/11/03/technology/cruise-general-motors-self-drivingcars.html
7
Summary explanation of chapters
Chapter 1, this chapter, has set out the motivation for this dissertation and the
significance of the work. Chapter 2 describes the research methodology, the research approach,
the problem the dissertation seeks to address and the questions that underlie it. It also sets out
the methodology used and the reasons for the choices made within that.
Chapter 3 is a literature review, examining the significant writings on the transportation
technologies that have paved the way for AVs: cars, highway-building, and Transportation
Network Companies (TNCs). It uses the theoretical lenses of technological determinism, social
constructivism and coevolution of technology and society to provide a context for the
understanding of AV development.
The literature review finds that the theory of coevolution of technology and society most
closely aligns with the development of transportation technologies of the past and goes on to
consider the implications of this on cities. The review assesses the processes, actors and
stakeholders involved to see if they hold clues as to why outcomes are not as good as hoped so
that lessons might be learnt for AVs.
Chapter 4 turns to the development of autonomous vehicles and finds the same similar
stakeholders, motives, promotional messages, and concentrations of influence at play. The AV
case study cities (San Francisco, CA and Phoenix, AZ) in Chapter 5 examines whether what is
happening on the ground reflects the national picture and the more historic transportation
technology development. Chapter 6 provides an analysis of the findings and Chapter 7 considers
the implications of the research and what now needs to happen with AVs to prevent the same
negative consequences of the past.
8
Chapter 2: Methodology
“Moving about the planet will be faster, safer, easier, comfier, greener, cheaper, and whooshier.
Best of all, there will be no traffic.11
”
Alex Davies, Wired Magazine
Research objectives: the research questions/ objectives
This dissertation considers the history of transportation technology development and uses
it to reflect on likely futures, developments and consequences of AVs.
There is much to learn from transportation initiatives of the past if we are looking for a
better future. That learning could include: how we tend to think about technology; patterns of
persuasion and outcomes; the role of different tiers of government; the contribution of engineers
and urban planners and the impact this has; who can effect change and who cannot – and the
perceptions of this; the short and long-term costs and benefits of a transportation innovation; the
unintended consequences – and which of these could be predicted and which could not. This
dissertation will examine and assess these issues in relation to AVs and use the findings to reflect
on the current development of the technology in a way that may assist the public and current and
future policymakers.
11 Davies, A. (2017, 31 December) The World's Fair Future of 1939 and the Quest for Our Next Utopia. Wired.
https://www.wired.com/story/worlds-fair-1939-futurama-utopia/
9
Research questions include:
● What lessons in technology promotion and outcome can we draw from the history of
transportation technology?
● Does the current development of AVs follow the pattern of previous transportation
technologies?
● How are AVs currently being developed and promoted? Who is engaged in the process
and how do they do so?
● How can theories of technology development inform our understanding of the future?
● Who has historically been involved and who has benefitted?
In order to respond to these questions, this dissertation will take a number of approaches. To
learn from the past other major technological innovations are identified and examined to see how
they developed and to examine their consequences. I will consider alternative theories of
technological development, exploring the relationship between the technology and the actors
involved in its development in order to understand patterns of influence and outcomes.
The technologies assessed will be of the 20th and 21st centuries and will include the
introduction of the automobile a hundred years ago, freeway-building and Transportation
Network Companies. These historic examples have been chosen because they have had significant
transformational claims made about them and have helped to lay the foundation for current AVs
and their use case.
Why case studies?
To learn from the present, case studies have been designed to reveal whether there is a
risk of repeating the mistakes of the past and to inform the future.
10
Evidence has been gathered through: analysis of public documents including public agency
reports and meeting agendas and minutes; public agency strategy documents; publicly available
agency letters; news articles and media coverage; websites of AV companies and public agencies;
public meeting videos; media and social media coverage; and historical records. In-depth
personal interviews have been conducted with key players in case study cities, including local
journalists, business group representatives and government and public sector officials.
Robert K Yan has noted, case studies are helpful in examining the issues of social and
political complexity and when ‘how’ or ‘why’ questions are being asked; he notes that “the case
studies unique strength is its ability to deal with a full variety of evidence - documents, artifacts,
interviews, and observations”12
. The case studies provide strong detail at a more micro level of
analysis and enable an understanding of what is happening on the ground and whether this aligns
with a national narrative or reflects historic patterns.
Two case studies have been chosen that are sufficiently different to draw lessons: San
Francisco, CA and the Phoenix metro area, AZ. Both places have been early adopters of AV
technology – and so have more accumulated knowledge and experience from which to draw.
The two case study areas have had a very different experience of AVs: one has had a
somewhat turbulent experience marked by protest and political opposition; the other has had a
relatively smooth experience within a context of a welcoming civic environment. They contrast
politically, with San Francisco dominated by Democrats at city and state level and Phoenix,
traditionally a swing city, has now moved closer to the Republican politics of the state
government. Importantly they differ strongly in regulatory approach – with strong regulation in
California and very little regulation in Phoenix. San Francisco has a single city-county status,
12 Yin RK. (1994) Case Study Research : Design and Methods. 2nd ed. Sage Publications
11
which in theory should make for easier decision-making without reference to an additional tier of
governance.
The case studies also contrast geographically – west coast and southwest – with differing
climates. They also differ in street layout and geography – with Phoenix inland, and having wide
streets and San Francisco, hilly, at the northern tip of a peninsula with a more historic, smaller
grid street pattern. The cities contrast significantly in terms of population density with 3,105
people per square mile in Phoenix compared to 18,629 in San Francisco13. The areas have a very
different number of daily vehicle miles travelled - 91m in the Phoenix area contrasting with 56m in
San Francisco14. Phoenix is also much less congested than San Francisco with the average driver
having lost 45 hours in congestion in 2023 compared to Phoenix’s 14, according to Inrix15
.
San Francisco also contrasts with Phoenix in the prosperity of its people – it is a global
city with high income earners in an economy dominated by professional services with banking and
high tech. Unlike Phoenix, it has a strong and well established and used public transit system and
a history that included freeway revolts which prevented its downtown being dominated by
freeways.
The case studies include an examination of the various documentation associated with the
AV programs, including state and local regulations, the long-range planning strategy, council
committee reports including any annual reviews of the programs, executive orders, regulations
and policies associated with the programs, the staffing structure and the governance
arrangements (and their agenda papers and minutes), program plans and reviews by the various
13 US Census Bureau (2024) Quick Facts Phoenix City, Arizona; San Francisco city, California. Census.gov
https://www.census.gov/quickfacts/fact/table/phoenixcityarizona,sanfranciscocitycalifornia/PST045224
14 Federal Highway Administration: Highway Statistics (2020) Federal Highway Administration.
https://www.fhwa.dot.gov/policyinformation/statistics/2020/hm71.cfm
15 Inrix (n.d.) Inrix 2023 Global Traffic Scorecard. Inrix. https://inrix.com/scorecard/#city-ranking-list
12
companies and organizations involved, local news reports, webpages, and any local opinion
surveys.
Interviews
Interviews were carried out with a range of people, representing key organizations
involved in city decision-making or the community. The interviews provide insight into the
perceptions and motivations of the actors involved. Before the interviews, desk research identified
the main issues and positions held and the interviews then conducted to verify them. The
interviews were conducted with the acceptance that comments would be attributed to individuals
on condition that they were given the opportunity to review and amend anything attributed to
them. The interview schedule was approved by the Institutional Review Board of the University of
Southern California in March 2024.
The interviews have been carried out by Zoom in a semi-structured way, with a standard
list of questions for all and with follow-ups that reflect issues more unique to the individual’s
experience. The semi-structured approach aimed to provide a consistency of questions across
case study areas, to encourage openness and enable a tailoring of questions based on the
particular nature of the city and its program and the role that person played within it.
The interviews sought to establish views and facts that have been flagged as significant in
the literature review. In particular they sought to establish the city’s motivations; stakeholder
involvement; expectations of the testbeds; whether people thought AVs inevitable; and whether
they’re optimistic about the technology in the future.
Between nine and eleven months following the interviews I went back to each interviewee
to invite them to review the quotes and views I attributed to them. All quotes were approved by
the interviewees. The interview questions are set out in Appendix 1.
13
Analysis: The methods used to analyze the findings.
The findings from the case studies were then analyzed and used to reflect on the likely
futures and consequences of further AV development. The processes of the past have been
assessed to see if similar dynamics are at work. In particular an assessment has been made
around whether lessons have been learnt around how technology is sold, why the outcomes of
that technology so often fail to meet the hopes and expectations laid out when launched, and how
the development of AVs reflect the process of previous transportation technologies.
Learning, knowledge gap-filling and contribution to urban planning
The learning from this work will be an understanding of the mistakes made from the past
and whether the lessons from them have been learnt today. Establishing the nature of the
dynamic associated with transportation technologies – what drives it – is key to assessing the
future with AVs. There are risks that if the same people, organizations and processes are involved
with the same motives then we may find the outcomes of AVs are the same limited mobility
options and decreased livability that have embattled cities – while believing we’re tackling the
safety, environmental, accessibility problems we have faced for over a hundred years.
Much of the research on autonomous vehicles is from an engineering perspective – this
dissertation takes a more holistic approach for a more rounded view. It draws on the disciplines of
history, political science, policy-making, behavior science, and urban planning. For the latter
discipline it has a particular relevance: it aims to demonstrate the importance of the involvement
of urban planners in the development of AVs and urge their contribution to consequent policymaking. The starting point to the assessment of AVs should be how we want our cities to be, then
decide who should be shaping decisions.
14
Langdon Winner stated in an interview in 2021 that the 20th century approach to
technology was, ‘innovate first. Ponder the implications later’; this dissertation aims to help avoid
that. He continued, ‘In that light there is a consistent disposition to encourage potentially world
changing developments to unfold and to offer erudite, retrospective (but likely irrelevant)
commentaries as the fascinating prospects emerge. …. The idea that one might announce a firm
‘No’ to any attractive, innovative pathway is simply out of the question16’. My aim is to make it
clear that this technology can be shaped by society and encourage anyone interested to take an
active role in recommending a way forward.
16 Winner, L. (2020, November 21) Interview with Langdon Winner: Autonomous Technology: Then and Still Now.
Langdon Winner. https://www.langdonwinner.com/other-writings/2020/11/21/interview-with-langdon-winnerautonomous-technology-then-and-still-now
15
Chapter 3: Theoretical Framework
A central theme of this dissertation is what or who has driven AV development. There are
two contrasting perspectives: a technological determinist perspective which stresses technology
as a force that shapes society; and a social constructionist perspective which sees technology as
shaped by societal choices. In addition there is a third perspective, the coevolution of technology
and society in which each influences and shapes each other17
.
Technological Determinism
Technological determinists see technology as having ‘a more or less autonomous force,
that cannot, or at least not easily, be resisted’18. Sally Wyatt describes it as having two dynamics:
first, that technological developments take place outside society - by inventors, engineers, or
designers - independent of social, economic and political forces; and second, that technological
change causes or determines social change19
. This power is played out in how technology is
developed and its consequent impacts.
Since the 1980s deterministic views have been challenged by sociologists and historians of
technology20. Deterministic claims have since been condemned as, ‘intellectually impoverished
17 Rip, A. & Kemp, R. (1998). Technological change. In S. Rayner & E.L. Malone (Eds.), Human choice and climate
change (pp. 327–399). Columbus, Ohio: Battelle Press.
18 Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and society, and
how they affect the current debate about artificial intelligence. Human Affairs (Bratislava, Slovakia), 30(4), p501.
19 Wyatt, S. (2007) in Hackett, Amsterdamska, O., Lynch, M. E., Wajcman, J., Sismondo, S., Bijker, W. E., Turner, S.,
Thorpe, C., Latour, B., & Clarke, A. E. (2007). Technological Determinism Is Dead; Long Live Technological
Determinism. In The Handbook of Science and Technology Studies. MIT Press.
20 Dafoe. (2015). On Technological Determinism: A Typology, Scope Conditions, and a Mechanism. Science,
Technology, & Human Values, 40(6), 1047–1076.
16
and politically debilitating’21: ‘intellectually impoverished’ because of a perceived linear, one
dimensional view of technological development; and ‘politically debilitating’ because it suggests
social and political interventions are futile. Succinctly, Jasanoff has described it as, ‘historically
flawed, politically dangerous, and ethically questionable’ and that ‘Insisting that technological
advances are inevitable keeps us from acknowledging the disparities of wealth and power that
drive innovation for good or ill.’22. These sociologists and historians of technology have pointed
instead to a very different dynamic – that of social constructivism.
Social Constructivism
A social constructivist perspective sees technology development shaped by human
interest, values and action. In contrast to technological determinism, social constructivists see
technology as subordinate to the power of society; that the development and adoption of
technologies are influenced by social factors such as cultural values, power dynamics, and
negotiation among various stakeholders.
For social constructivists, the general public are part of the construction of technologies.
They create demands – such as for cheaper air travel or easier parking payments – and
technology responds to that need. This is simple market forces at work. Technology developers
aim to anticipate consumer needs or – a more neat trick - create a need that consumers were not
aware they had. Car maker, Henry Ford, has been quoted as saying (although there’s no evidence
he did so), “If I had asked people what they wanted they would have said faster horses.”
21 Bijker. (2015). Social Construction of Technology. In International Encyclopaedia of the Social & Behavioral
Sciences (Second Edition, Vol. 24, pp. 136). Elsevier Ltd.
22 Jasanoff, Sheila. (July/August 2021): 16-17. The Dangerous Appeal of Tech. MIT Technology Review 124.4
17
Innovators can’t control consumers’ response to their product and social constructivists give
weight to the consumer response (in contrast to technological determinists) and to public opinion.
A technology can be perceived differently by different social groups. Pinch & Bijker talk of
“interpretive flexibility” when it’s possible for social groups, which includes consumers, to shape
an artifact before “closure” when a design is set23. The example of the ‘Ordinary’, an early bicycle
model, which some non-users saw as a bicycle that basically didn’t work as it should and was
unsafe, while users saw it as a ‘macho bicycle’ – the risk was part of the fun and there were
opportunities to show off one’s athleticism to the women in the park24 (who may have been less
impressed than was perhaps thought). For the latter, the Ordinary worked as intended in all its
mirthful risk – but for others it was essentially a nonworking contraption. But as Bijker points out,
‘working’ and ‘nonworking’ of an artifact are socially constructed assessments, rather than
intrinsic properties; ‘”working” is merely in the eye of the beholder’, machines ‘work’ because they
have been accepted by relevant social groups.
I elaborate on this point because it has important implications for the development of
other transportation technologies as automakers and producers seek consumer views and shape
their products accordingly. This, in turn, affects how the technology develops and the subsequent
impact on the environment and its communities.
The bicycle of course gained a wider market by appealing to a broader section of the
population who wanted something done about early bicycles that were perceived to be as unsafe
and nonworking; and so the bicycle was developed so that it became safer and more useful.
Bijker’s demonstration of interpretative flexibility of the same bicycle shows that there is “an
immediate entrance point for a sociological explanation of the development of technical
23 Bijker, W. E., Hughes, T. P., & Pinch, T. (2012). The Social Construction of Technological Systems: New Directions
in the Sociology and History of Technology (Anniversary edition.). The MIT Press. pp20-21
24 Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs : toward a theory of sociotechnical change. MIT Press.
18
artifacts… If no interpretive flexibility could be demonstrated, all properties of an artifact could be
argued to be immanent after all. Then there would be no social dimension to design; only
application and diffusion - or context for short - would form the social dimensions of technical
development”25
.
Having established the flexibility inherent in the interpretation the next step should be to
map the process by which artifacts attain or fail to attain a stable interpretation. An artifact
“does not suddenly appear as the result of a singular act of heroic intervention: instead it is
gradually constructed in the social interactions between and within relevant social groups”26
.
Bijker establishes this flexibility by deconstructing the bicycle as an artifact establishing what it
means to people and the purpose it serves. He shows that the design of the bicycle was socially
constructed because the eventual design that succeeded was not the best in some objective sense
but that it was the one that most relevant social groups found fitted their needs. The bicycle, as
something that approximates to what we today understand as a bicycle, took 18-year to
process27
.
Coevolution of Technology and Society
Another theory of technological development is the coevolution of technology and society,
which sees agency with both, emphasizing a reciprocal and interactive relationship; it lies,
“somewhere between the poles of technological determinism and social constructivism”28. Here
25 Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs : toward a theory of sociotechnical change. MIT Press. p76
26 Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs : toward a theory of sociotechnical change. MIT Press. p270
27 Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs : toward a theory of sociotechnical change. MIT Press.
28 Hughes, T. P. (1994) Technological Momentum in Smith, & Marx, L. (1994). Does technology drive history? : the
dilemma of technological determinism. MIT Press. p112
19
technology and society evolve together in a dynamic process, with each responding, influencing,
adapting, and shaping the other.
Proponents of coevolution say social constructivism overstates the degree to which society
can steer or direct technological developments. Van de Poel describes the coevolutionists
recognition of a ‘non-malleability’ of technology – it’s not as flexible as social constructionists
believe for a number of reasons. Chief amongst these reasons is that technology brings novelty –
something that did not exist before and so it is not possible to understand fully the implications
and consequences of the technology29. This non-malleability of technology, says Van de Poel,
makes it hard to govern30
.
Hughes argues that this difficulty in governing is caused by technological momentum:
once launched, and the technology is widely adopted, it’s hard for government to ban or restrict it
and hard for society to shape it31. So shaping them through the democratic process is an
important matter of timing; Hughes continues, “shaping is easiest before the system has acquired
political, economic, and value components”.
Another cause of this difficulty to govern technology, as described by Colleridge, is its
complexity and scale32. Technologies are introduced into complex systems – political,
environmental, social or economic systems – that can be too complex for either the technology
itself or society to control. Collingridge sees phases in development and that in the early stages
(while the technology is still malleable) it is possible to steer it in the right direction according to
29 Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and society, and
how they affect the current debate about artificial intelligence. Human Affairs (Bratislava, Slovakia), 30(4), 499–
511.
30 Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and society, and
how they affect the current debate about artificial intelligence. Human Affairs (Bratislava, Slovakia), 30(4), 499–
511.
31 Hughes, T. P. (1994) Technological Momentum in Smith, & Marx, L. (1994). Does technology drive history? : the
dilemma of technological determinism. MIT Press.
32 Collingridge, D. (1980) The Social Control of Technology. New York: St. Martin’s Press
20
social values and needs, although it is not clear what the consequences might be of the
technology33. Later when this knowledge has become available it is much harder – or impossible -
to change it as it has become entrenched in society34. Hughes notes that, “As they become larger
and more complex, systems tend to be more shaping of society and less shaped by it.”35
.
Conclusion
The strength of technological determinism lies in an apparent history of evidence. Street
cars and later automobiles, for example, appeared to change society fundamentally – where
people lived, the work that they did, the company they were able to keep. But the weakness of
technological determinism lies in the fact that, while it appears to hold true at a macro level, it
doesn’t hold true at a micro level. As we will see in the following chapter, government and society
enabled the automobile to hold the influence it had on society.
Social constructivism has more plausibility here as we look at the role of individuals, such
as Henry Ford, and at the shaping of automobiles from a farming devise to a commuting mode,
the active support of federal government, the development of production lines in factories that
increased the output of vehicles, and the importance of funding mechanisms that enabled even
people of mediocre means to buy one: these all point to social constructivist explanations for
transportation technologies.
Social constructivism though, as an explanation of transportation technology
development, seems a little lacking. Didn't road-building and the auto shape American cities from
33 Collingridge, D. (1980) The Social Control of Technology. New York: St. Martin’s Press
34 Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and society, and
how they affect the current debate about artificial intelligence. Human Affairs (Bratislava, Slovakia), 30(4), 499–
511.
35 Hughes, T. P. (1994) Technological Momentum in Smith, & Marx, L. (1994). Does technology drive history? : the
dilemma of technological determinism. MIT Press. p112
21
the 20th century with suburban homes and lives coming to evolve around automobility? Aren't
attempts at building walkable, environmentally sustainable neighbourhoods frustrated by car
culture? Such arguments suggest there may be a point at which a transportation technology
artifact can no longer be held in check – their usefulness and availability bringing a momentum to
a point beyond which it is impossible to control.
The question of theory is particularly important for autonomous vehicles. AV companies’
advice to governments has oscillated between presenting a technological determinist view of AVs
and a social constructivist one; one disempowers both government and society and the other
shares responsibility with them. With technological determinism there’d be little point in
governments attempting to create policy to shape how they develop or society attempting to
affect their rollout. Social constructivism provides a more positive outlook – more democratic –
society can shape AVs in a way that fits their cities best. This bodes well for AVs because if they
are developed and turn out not to be a positive addition to cities then they can be re-shaped so
that they are.
A coevolutionist perspective is more hazardous. It warns that there’s a point at which
society can’t control what happens: it’s for this reason that establishing the most relevant theory
could present a pressing case for taking an early strategic view on the technology and shaping it
to society’s benefit before it’s too late. For policy makers the aim, notes Van de Poel, is to try to
proactively steer technology during its early phases, while being alert to future possibilities and
open to stakeholder deliberation although the risks of overlooking issues and concerns that are
hard, or impossible, to anticipate at these early stages remain36
.
36 Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and society, and
how they affect the current debate about artificial intelligence. Human Affairs (Bratislava, Slovakia), 30(4), p505.
22
Chapter 4: The History of Transportation Technology Development - A
Literature Review
“there is one and only one social responsibility of business – to use its resource is and engage in
activities designed to increase its profits so long as it stays within the rules of the game, which is
to say, engages in open and free competition without deception or fraud37
”
Milton Friedman
There have been a myriad of transportation technologies over the centuries and it is
impossible to know all that were invented as some will have been abandoned on the garage floor
or never even got off the paper upon which they were designed. We can see many that did
succeed and many have failed to live up to the launch expectations or that developed in ways
that was very different to how they were originally envisioned.
Because of the sheer number of transportation technologies this chapter will review the
literature on those technologies most closely connected with the development of road
transportation that are most relevant to AVs. This has narrowed down the areas of research to the
automobile itself, the building of freeways (with grade separation interchanges, a significant
enabler of private car travel) and Transportation Network Companies (TNCs), which align closely
with robotaxis, the chief use case of AVs today.
The first section of this chapter will examine how the history of these technologies have
been explained in the literature and will consider evidence of whether the dynamics of
technological determinism, social constructivism or coevolution best explain their development. In
looking at who or what has driven this technology development, the second section of this chapter
37 Friedman, M. (1970, September 13). A Friedman doctrine – the social responsibility of business is to increase its
profits. New York Times https://www.nytimes.com/1970/09/13/archives/a-friedman-doctrine-the-socialresponsibility-of-business-is-to.html
23
will consider whether the promises made about the technology have been realized by considering
the impacts and outcomes of these technologies. To understand better the reasons for these
impacts and outcomes the analysis will then turn to consider the processes, players and
stakeholders to see if they hold clues.
The Automobile – A brief history of its development
The early history of the automobile reveals an exploratory period of trial and error
involving different actors, working to find a profitable use case. There was no immediate success,
no sense that cars could be foisted upon what was initially an unwilling public. It took time, effort
– and a coincidence of other dynamics – for the technology to prove a success in the consumer
marketplace.
When the automobile first arrived on American streets they were often not welcome. In the
first decade of the twentieth century drivers had rocks thrown at them, were physically attacked,
were pulled from their vehicles, shot at by farmers, and found roads booby-trapped with glass or
tacks38. Driving at that point in history was for fun, a daring pursuit for wealthy sportsmen, or a
mode of travel for a tiny number of wealthy doctors, business men and engineers39. Jackson
suggests the Literary Digest reflected most people’s view of automobiles at that time – they were
a curiosity and toy – a ‘pleasure vehicle’ – and “more an offspring of the bicycle than a successor
to the horse-drawn carriage”40
.
38 Weingroff, R. F. (n.d.) On the Road with Woodrow Wilson. Federal Highway Administration.
https://www.fhwa.dot.gov/highwayhistory/wilson.pdf
39 Flink, JJ. (1990) The Automobile Age. MIT Press; 1990 p28
40 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press.
24
Automobiles were also a symbol of class divide. Soon-to-be President, Woodrow Wilson,
told the New York Times in 1906, “Nothing has spread socialistic feeling in this country more than
the use of automobiles. To the countryman they are a picture of arrogance of wealth, with all its
independence and carelessness.” With the perceived threat of socialism abroad it is
understandable that car-makers wanted to change the perception of their vehicles. Key to doing
so was to market the vehicles to ordinary people so they weren’t seen as a plaything of the rich.
John Parson, a Chicago banker, who was president of the American Automobile
Association (AAA), insisted Wilson’s perception was outdated, urging him to travel through the
Midwest where, “Scores of well-to-do farmers are now automobile owners, and where three years
ago the horseless vehicle was regarded with suspicion, entirely different sentiments are
entertained now. Just as soon as any class of people realize that the automobile is useful for the
ordinary work and duties of life apart from its pleasurable attributes as a touring car, it will be
regarded with favor”41. Soon others saw how they might benefit from the technology – and the
farming use case developed. Indeed it has been estimated that more than a third of automobiles
were owned by farmers in 191942 and this market was key to changing perceptions.
As ‘use cases’, the farmer - and the country doctor - were compelling reasons to welcome
the motorcar as they were seen to feed families and to save lives. For rural dwellers, they
provided an important means of earning a living and gaining access to social or cultural places
previously only available to urban dwellers. Manufacturing cars for the large rural market that
were affordable and designed to navigate country roads helped to prevent animosity towards the
vehicles. Slowly the form and function of automobiles adapted to the most promising market.
41 Weingroff, R. F. (n.d.) On the Road with Woodrow Wilson. Federal Highway Administration.
https://www.fhwa.dot.gov/highwayhistory/wilson.pdf p13
42 Akron Weekly Pioneer Press (1919, June 26). Early Automobiles. University of Northern Colorado.
https://www.unco.edu/hewit/doing-history/colorado-farmers-ranchers/transportation/early-automobiles.aspx
(Retrieved 10 June 2024)
25
In the early years of the 20th century, the cost of a car was prohibitively expensive for most
people, which kept sales low (Figure 1); but by the second decade the average man could aspire
to purchase one as the costs declined
43
.
Year Number of privately
owned automobiles
1900 8,000
1905 77,000
1910 458,000
1915 2,332,000
1920 8,312,000
1925 17,481,000
1930 23,035,000
Table 1: Cars privately owned in the US 1900-193044
The affordability was created through a coming together of technology advancements that
combined to create a seismic change. The car was new technology but so too were the methods
of production which enabled the car production to flourish and become affordable. The price of
Ford’s Model Ts dropped from $950 in 1910 to $290 in 1924, enabled by fast production lines that
were churning out the vehicles at a rate of one every ten seconds in 192545. Competitors had to cut
their prices to remain competitive so all cars became cheaper. Importantly, methods of financing
a car purchase changed with buyers not having to pay the whole cost before driving away;
43 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press. p159
44 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
45 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press. p160
26
General Motors’ success (their sales eclipsed that of Ford after the war) was largely to do with its
‘buy now, pay later’ financing schemes46
.
The cost of running a car also declined as gasoline became more widely available and
cheaper (notably after the Texas and California oil discoveries) and road quality improved. An
automobile ecosystem was developed, fueled by suburban real estate development bringing
economic growth with cars and road-building enabling the home purchases. This coincidence of
financial, economic and social changes combined with an auto-ecosystem bolstering purchases;
the numbers of cars on streets soared.
By the 1920s, the private car had become no longer a luxury but a necessity of the
American middle class47. During this decade, Robert and Helen Lynd’s social study of ‘Middletown’
found the automobile had become “an accepted essential of normal living” and the primary focal
point of urban family life, families that would rather go without food than give up their car48
.
Within a short space of time Americans had become dependent on them and their lives seemingly
shaped by them.
The essential nature of the automobile was more officially confirmed in the President's
Research Committee on Social Trends report of 1933, “imperceptibly, car ownership has created
an ‘automobile psychology’; the automobile has become a dominant influence in the life of the
individual and he, in a real sense, has become dependent upon it”49. The appeal was compelling;
46 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
47 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press.
48 Interrante J. (1980) The road to Autopia: The automobile and the spatial transformation of American
culture. Michigan quarterly review. 1980;19(4):502
https://quod.lib.umich.edu/m/mqrarchive/act2080.0019.004/84:16?g=mqrg;rgn=full+text;view=image;xc=1;q1=in
terrante
49 Quoted in Wachs, Martin., & Crawford, M. (1992). The Car and the city : the automobile, the built environment,
and daily urban life. University of Michigan Press. p72
27
the automobile offered the fastest, most direct route to any destination, it didn’t suffer from
tiredness and opened up social opportunities that were unable to be reached before50
.
The automobile: technological determinism, social constructivism or coevolution?
The automobile’s unrelenting expansion and domination over other forms of mobility came
to be viewed as natural and inevitable51. Over a long period of time it is possible to discern
technological determinism at work in their development; cars were shaping behaviors – where
people lived, where they worked (although the process of suburbanization was well underway
before the car), if they worked and how they went about their lives. It opened up social networks
too, bringing friends and families closer together and widening the accessibility of cultural and
leisure pursuits. Kline & Pinch have noted that many historians of the automobile use a form of
technological determinism to explain change; but the authors argue that users of cars themselves
acted as agents of technological change52: it was the people that shaped a use case that
enhanced their comfort and convenience, and the industry responded with a profitable product
that was shaped around the needs and desires of consumers.
The key actors in the development of the motor car included the inventors and engineers
such as Karl Benz and Henry Ford as well as automobile manufacturers, notably General Motors.
The banks and car makers had a role here and so too did the government by creating regulations,
safety standards and policies. Local government was willing to invest in roads which made driving
easier and tax on fuel provided a good funding steam. Enabling legislation, the introduction of the
50 Wachs, Martin., & Crawford, M. (1992). The Car and the city : the automobile, the built environment, and daily
urban life. University of Michigan Press. P73
51 Urry, J. (2008). Governance, flows, and the end of the car system? Global Environmental Change, 18(3), 344.
https://doi.org/10.1016/j.gloenvcha.2008.04.007
52 Kline, R., & Pinch, T. (1996). Users as Agents of Technological Change: The Social Construction of the Automobile
in the Rural United States. Technology and Culture, 37(4), 763–795. https://doi.org/10.1353/tech.1996.0006
28
concept of ‘jaywalking’ and keeping children from playing on streets were all policy decisions that
favored the motorcar and enabled its development to become the dominant mode of
transportation In cities. Consumers were also important as early adopters and as actors who
influenced demand, shaped the development of new car features and usage patterns. The media
and advertisers also played a role promoting motor cars and shaping public perception.
The auto industry ecosystem became a behemoth over the 20th century, incorporating
vehicle manufacturers, suppliers of parts and components, dealerships (the retail arm of the
industry), the aftermarket for repairs and maintenance services, finance and insurance,
technology providers, supply chain providers, energy providers and marketing and advertising
agencies. Together they shaped the product, how it was powered, how safety and environmental
issues were defined, and ensured that their interests were listened to by policymakers53
.
The growing ecosystem profited from the expansion of sales and helped to increase the
need for and desirability of automobiles and turn them into cultural icons. Public and private
organizations, engineers, auto clubs, and manufacturers were coming together formally and
informally to encourage the growth of the industry. Encouraging the growth in car ownership and
helping to create the necessary infrastructure to support it including networks of streets, arterial
roads, and freeways54
.
Size and a wealthy membership brings clout and auto owner associations grew and
strengthened in political power. In making sure policymakers heard their messages the industry
used its own lobbyists and public relations staff in Washington and collaborated with trade and
business associations55. It became adept at forming coalitions with willing partners for specific
53 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press p183
54 Wachs, Martin., & Crawford, M. (1992). The Car and the city : the automobile, the built environment, and daily
urban life. University of Michigan Press. p9
55 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press
29
pieces of legislation or for specific public relations purposes, gaining something of a halo effect
when it joined with an organization with a moral purpose (such as the American Automobile
Association’s support in educating children about road danger). Intense auto lobbying paid off as
the federal government helped the auto industry in a number of ways including protecting it from
foreign competition. Between 1913 and 1922 import tariffs ranged from 30 to 45 percent, and
from 1922 to 1934 from 25 to 50 percent; so in practice competition could only be domestic56
.
The federal government also helped by not regulating – including on safety. Indeed, until
1966 the automobile was the only form of transportation in the US without federal safety
standards even though it accounted for 92% of transport-related deaths and 98% of injuries57. To
avoid regulation the industry created a narrative of blaming the driver and encouraged the
development of the ‘three Es’: engineering (highway design), education (of the driver), and
enforcement (of the rules of the road). This engineering mantra was a strategy that served to
distract and insulate the industry from regulation; its intent was not to save lives but to shape
how the problem of highway deaths was defined58. Luger notes, “auto industry officials who,
seizing the initiative on safety with their superior resources, deflected attention away from the
contribution of vehicle design to injuries. By being able to shape the public perception of traffic
safety issues, industry officials prevented governmental scrutiny of their design decisions and
preserved their managerial autonomy”59
.
The vehicle annual style changes squeezed out the smaller auto manufacturers,
concentrating the power of the auto industry into the hands of a small number of companies.
56 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press
57 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press p54
58 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press
59 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press p183
30
Smaller car companies couldn’t afford to abandon tools and dyes before they were worn out or
used up – or the necessary advertising to show what was new; forty three companies left the
industry between 1923 and 1926 and by 1935 only ten remained60. By the two decades after the
World War II the auto industry was “a tightly knit, joint-profit-maximizing oligopoly, earning
profits well above the average for all manufacturers leading to antitrust investigations61.” The
industry had become incredibly rich and powerful.
Coevolution
It’s clear f that the proliferation of cars was not the product of technological determinism.
A myriad of players - and economic, financial and social coincidences - came together to provide
a clear path for technological development and societal infusion. But can we simply say it's social
constructivism? There were dynamics at play that make it closer to a coevolution model;
specifically the eventual setting of a developed car form and function (that was no longer
malleable) and the complexity of the ecosystem that made it impossible to soften this form and
function once it had become ‘locked-in’.
While there were many other actors that had influence, the public did - after some
hesitation - respond positively. Just as the bicycle followed the pattern of what Kline & Pinch have
described as a process of interpretative flexibility – the malleability of technology – before a final
version of it is settled upon, after use cases of ‘a play thing for rich city dwellers’ and farm
equipment, by the end of the 1920s, ‘closure’ had occurred and a standard car had developed.
This artefact was built for comfort, speed and for ‘go-anywhere, go-anytime’ use for whoever
60 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press
61 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press p41
31
could meet the low financial barriers. More types of vehicles met the diverse needs of different
users. Farm labourers, “had stopped using their autos for grinding their grain, plowing their fields,
or carrying their produce to town; instead, they had begun to buy tractors and pickup trucks in
large numbers— new artifacts that manufacturers developed partly in response to these novel
interpretations of the car. The users, so easily overlooked in writing the story of technology, had
made their mark”62. This shaping of the domestic car aligns with the coevolutionists view of
technology development – momentum had been produced, closure had occurred in the car as an
artifact, and at a point it became too complex for government or society to control the processes
at play even if there had been the impulse to do so.
Once society and the economy became ‘locked in’ to the form and function of the
developed car – one that was broadly acceptable to the public – the malleability became lost as
massive profits were made for those producing and selling them and the associated
infrastructure, products and services63. Further, the ecosystem that had developed became so
complex and sprawling that government and society could not reverse it, “as billions of agents
co-evolve and adapt to form a system of interdependent agents and relations—a complex
assemblage or system that ‘constitutes’ the ‘steel-and-petroleum’ car64”. This momentum has
been created and been so firmly embedded that even today cities have found it very hard to carve
out active travel complements to the road system including cycling and transit infrastructure. Yes,
the car shaped where people lived and what they did with their time but consumers were
signalling to the ecosystem how they wanted cars to be shaped and what they wanted them to be
able to do to improve their convenience and comfort.
62 Kline, R., & Pinch, T. (1996). Users as Agents of Technological Change: The Social Construction of the Automobile
in the Rural United States. Technology and Culture, 37(4), 763–795. https://doi.org/10.1353/tech.1996.0006
63 Arthur, B. (1994). Increasing Returns and Path Dependence in the Economy. University of Michigan Press, Ann
Arbor.
64 Urry, J. (2008). Governance, flows, and the end of the car system? Global Environmental Change, 18(3), 344.
https://doi.org/10.1016/j.gloenvcha.2008.04.007
32
Highway-building - A brief history of its development
Highway-building, while not a technological artifact in a traditional sense, is a
transportation technology because it is the application of scientific knowledge, tools, materials,
machinery, and engineering techniques.
An important technological advance was the development of consistent, quality road
pavement that enabled vehicles to travel safely at speed. Road surfaces built for lightweight
bicycles or horse-drawn vehicles needed to provide traction but they impeded the speed of
motorcars. A period of experimentation and development had started in the nineteenth century
and at the turn of the next, city councils were looking for surfaces that would meet the needs of
both autos and horses.
The shift in materials from cobblestone and gravel to asphalt in the late 19th century had
enabled more stable passage of horses, bicycles and early motor vehicles reducing the risk of
slipping or skidding and had the added benefit of reducing the noise the affected surrounding
homes. The first source of asphalt was mined in Trinidad but soon sources were discovered within
the US which enabled its faster and cheaper transportation. But it took decades for asphalt to
develop into a stable, reliable pavement that wasn’t inclined to stickiness in the summer,
slipperiness in winter and general cracking and wear and tear. Differences in sources, composition
temperature when laid and other variables produced many years of variable – and at times poor -
quality. By the First World War however, the asphalt industry had reached maturity with more
consistency of product65
.
Meanwhile the further development of concrete as a road-building material was also
developing. The first concrete pavement was laid in Bellefontaine, Ohio in 1893, but it soon
65 Holley, I. B. (2003). Blacktop: How Asphalt Paving Came to the Urban United States. Technology and
Culture, 44(4), 703–733. https://doi.org/10.1353/tech.2003.0165
33
deteriorated. Better construction equipment soon enabled wider and longer slabs for two lanes of
traffic and a smoother ride. While concrete roads were noisier than asphalt they were found to be
more durable and provided a surface more suitable to cars, but not for horses.
The superior surface created by concrete was demonstrated in the first road built entirely
for cars. In 1908, William Kassam Vanderbilt II created a toll road for high-speed racing of
automobiles as well as ordinary travel outside of competitions, the Long Island Motor Parkway.
Built of reinforced concrete, it featured engineering as well as this material innovation, including
banked turns, guard rails and bridges that precluded the need for intersections that would slow
vehicles. This private thoroughfare was built for automobile speed and set a standard for
highway-building.
By 1910, the scale of need for road-building was becoming clearer and the focus of
demands moved from pressure on local and state governments (which could only fund smaller
incremental changes using property tax) to pressure on the federal government for significant,
large-scale construction. A social or moral imperative had developed too; good roads started to
be seen as a democratic tool, bringing rural dwellers the same rights, and privileges as urban
dwellers.
Pressure for better roads began with demands by bicyclists for routes to the country in the
late 19th century. It then developed as part of a Progressive reform movement to improve the lives
of rural populations, especially to enable access to markets in towns and cities. Mud-bound
farmers were being perceived as deserving of the same social, political, religious, and educational
opportunities available to urban residents and this provided a social justification for road-building
by the federal government: poor roads became a moral wrong and their improvement a matter of
democratic fairness66
. At this time, the US had about 2.2 million miles of rural roads, of which less
66 Seely, B. E. (1987). Building the American highway system: engineers as policy makers. Temple University Press.
p.35
34
than 10% had "improved" surfaces such as gravel, stone, sand-clay, brick, shells, or oiled earth
because travelling between states was considered a luxury, something only for wealthy
travelers67. Within five years though these principled concerns around fairness between rural and
city dwellers were overshadowed by a new rationale for road-building — the need to serve the
automobile68
- and only federal government was capable of providing the necessary funding.
Vanderbilt’s highway on Long Island provided an inspiration for a new long-distance
highway stretching from coast to coast. The Lincoln Highway, conceived in 1912 by Carl G. Fisher,
combined the same engineering know-how, notably grade separation, with the use of gravel for
the road surface which was soon replaced with a preference for concrete. The invention of selfpropelled steam-powered machines for mixing cement - a major technological breakthrough in
the first decade of the 20th century – made realistic the use of concrete for such large-scale
highway-building69. Concrete’s durability and longevity provided a smoother, faster driving
surface, improved traction, weather resistance, and allowed for standardization and predictability
– and machines for mixing became more widely available.
Concrete ‘seedling miles’ were used to provide a showcase to gain support and funding for
the new the Lincoln Highway. An ‘Ideal Section’ was built during 1922 and 1923 in Lake County,
Indiana, funded by federal and county government with contribution from the United States
Rubber Company, the president of which was one of the team creating the Highway, the Lincoln
Highway Association. The Ideal Section comprised: a 40-foot wide concrete pavement 10 inches
thick (maximum loads of 8,000 pounds per wheel); no grade crossings (or advertising signs); a
minimum radius for curves of 1,000 feet, with guardrail at all embankments; and curves
67 Weingroff, R. F. (2023, June 30) The Lincoln Highway. US DOT Federal Highway Administration.
https://highways.dot.gov/highway-history/general-highway-history/lincoln-highway
68 Seely, B. E. (1987). Building the American highway system: engineers as policy makers. Temple University Press.
p225
69 Holley, I.B. (2008). The Highway Revolution, 1895 to 1925. Carolina Academic Press, Durham: 160.
35
superelevated (i.e., banked) for a speed of 35 miles per hour. Together the materials used and the
engineering features were the underpinnings of a new type of highway that enabled speed,
throughput and long-distance travel. The key technological breakthrough of highways was grade
separation, which enabled higher speeds and a more efficient throughput of vehicles.
Sufficient funding for the highway’s large scale building however remained elusive for
decades as the project failed to attract sufficient federal or private sector funding (notably, Henry
Ford refused on principle to contribute). Eventually the federal government moved to help fund a
national road-building program. Weingroff cites five reasons why arguments for federal funding
for road-building became so persuasive after the turn of the century. First, the growing
involvement of farmers, who needed to get their goods to markets, and the needs of others in the
goods-roads movement in addition to the introduction of Rural Free Delivery, which depended on
passable roads for mail. Second, the increasing number of automobiles, especially after Ford’s
introduction of the Model T, an expectant middle class consumer market was gathering strength
alongside the growing strength of the American Automobile Association (AAA). Third, the Supreme
Court in Wilson v. Shaw (1907) decided that Congress had the power “to construct interstate
highways” under its constitutional duty to regulate interstate commerce. Four, the founding of the
American Association of State Highway Officials in December 1914 gave states an effective voice
for advocating a national road improvement program. Fifth, the appointment of the very
persuasive, Logan Page, as Director of the Office of Public Roads (OPR) in 1905 with his approach
of applying persuasive ‘apolitical’ judgment, based on “irrefutable data, free of political taint and
corrupt influence”; a belief in the rhetoric of data and metrics that continues today70 71. So a
70 Weingroff, R. F. (n.d.) On the Road with Woodrow Wilson. US DOT Federal Highway Administration.
https://www.fhwa.dot.gov/highwayhistory/wilson.pdf
71 Note: The perceived maturity of a science “has come to be judged by the extent to which it is mathematical.
First come astronomy, mechanics, and the rest of theoretical physics. Of the biological sciences, genetics is top
dog, because it has theorems and calculations. Among the so-called social sciences, economics is the most
mathematical and offers its practitioners the best job market, as well as the possibility of a Nobel Prize” (Davis and
36
variety of processes and players acted to facilitate the significant financial support of roadbuilding.
National security was an important selling point. From a global viewpoint, the world wars
were used to justify large-scale road-building in an area of policy in which federal government
had not been perceived to have a role. Political involvement was justified by portraying a threat of
foreign invasion and the need for adequate mobilization of troops across the country if the need
arose. President Roosevelt had seen from his active service in Germany what an efficient national
road transportation system should look like in an advanced economy and it was entirely different
to what existed back home. His personal commitment to change this and the pressure from
various industries - especially business, rubber, and auto manufacturing – strengthened the
argument for a significant federal investment.
At a national level, outrage over road deaths had escalated and the blame was often
pushed onto either driver behaviour (especially speeding), insufficient driver education or careless
pedestrians72. The solution to the safety problem was to find engineering solutions that’d support
high speed rather than reduce speed; speed was thereby reframed to an extent as an engineering
matter. Speed was prioritized and framed as efficiency and was enabled by devising a number of
technological and engineering solutions as noted, especially grade separation as a solution to the
safety problem by eliminating the conflict point of cross-travelling motorists, wide bends, lack of
distracting advertising and clearer lines of sight were the safety solutions all of which enabled the
motorists to drive faster.
Of the individuals who had an impact on road-building one was especially significant -
Larry Page, Director of the Office of Public Roads and President of the American Association for
Hersh 1987, 53). The art of city building that’s the work of urban planners is not held in such high esteem even
with the buzz of ‘smart city’ in the mid to late 2010s.
72 Norton, P. D. (2011). Fighting Traffic: The Dawn of the Motor Age in the American City. The MIT Press. pp75,
247.
37
Highway Improvement (a forerunner of the Federal Highway Administration). His Vice President
was the President of New York Central Lines and its board of directors included four senior people
in railway companies, a Congressman, engineering journalist, a university President - logistical
and road interests were also strongly represented. Railroad companies had an interest in roadbuilding because they wanted to continue to be a key player in the movement of goods - in
combination with road movement rather than in competition with it. To some extent it was a
broad church of differing interests with a mission to work collaboratively, because “the time had
come when all of the agencies striving for the common cause of road improvement should work
together73”. Harnessing the knowledge and experience of all partners was emphasized, “there
should be no more playing at cross purposes” and that the movement for better roads “should be
so systematized and everywhere placed on so high a plane of honest and earnest effort that the
cheap charlatanism of the professional promoter” should not be permitted nor (with a shot across
the bows of those intent on pernicious democratic involvement) “the bungling efforts of the wellmeaning but uninformed citizen”74
.
Although there was a diversity of professions, there was a singular outlook – that this was
an engineering matter, that they were dealing with facts, numbers, science and logic. This focus
on data rather than qualitative measures (such as around well-being, public health or livability)
had important implications for public policy attention. The definition of a successful and efficient
system had been reframed as a matter of mathematical throughput – and here highways became
the obvious solution. The context though was a sense of booming optimism and a national faith in
technological progress75
.
73 American Association for Highway Improvement. (1912). Papers, Addresses and Resolutions before the American
Road Congress 1911. Waverly Press, Baltimore USA
74 American Association for Highway Improvement. (1912). Papers, Addresses and Resolutions before the American
Road Congress 1911. Waverly Press, Baltimore USA
75 Jackson, K. T. (1987). Crabgrass Frontier: The Suburbanization of the United States (1st ed.). Oxford University
Press. p162
38
The concerns and strong direction of the Office for Public Roads appeared justified as it
was clear that congestion in cities had become a significant issue. To cope with increasing
congestion ‘parkways’, following the Long Island example, became increasingly popular with its
concrete pavement, bridges and tunnels to separate vehicles from local traffic, and limited access
through toll gates76. The principle of the parkway soon spread across the country.
While roads built for speed was part of the success of the auto and construction lobby it
was also a demonstration of the power of truck and freight lobbyists, industries that stood to gain
greatly from road-building which would reduce the monopoly power of the freight railroads. In
most cities, truck operators and state road engineers persuaded city officials to build parkways
with a view toward maximizing traffic flow.
The aim of the first meeting of the annual American Road Congress in 1911 was to find
solutions to the problems of road improvement in order to increase public prosperity and welfare.
The US President sent his support in a telegram stressing the value of good roads to farmers: “The
effect that [good roads] will have in increasing the value of farms, in making the lives of farmers
and their families much more full of comfort, and in the general benefit conferred by greater ease
of intercommunication the country over, can not [sic] be exaggerated”. A program of national
road-building had strong political support and this soon translated into strong financial support
with the Federal Aid Road Act of 1916, which provided up to half of the cost to states to build and
improve roads. Weingroff notes, “The Federal Aid Road Act of 1916 was, in many ways, the
embodiment of Page's ideals. Each state would have a highway agency with engineering
76 Jackson, K. T. (1987). Crabgrass Frontier: The Suburbanization of the United States (1st ed.). Oxford University
Press. p166
39
professionals to carry out the federal-aid projects. Page and his engineers had approval authority
so they could ensure the projects were designed and constructed properly”77
.
Arguments in favor of road-building appealed not just to engineers and the developing
ecosystem of businesses that benefited but to a broad section of the population78. Consultants
were hired to develop street traffic plans in the 1910s and 1920s usually commissioned by
downtown businesses who believed central business district traffic congestion was driving
customers to the suburbs79. City government had cause to support this view as declining central
business district property assessments would reduce their funding80. Focus was on trips made
within cities, rather than between cities and municipal planners knew that roads were not just a
means of facilitating movement but also of shaping urban form and of this their plans were
cognizant.
The main trust for road improvements though came from this developing ecosystem of
business interests which by the 1920s had included a coalition of powerful private pressure groups
including tire manufacturers and dealers, oil companies, service-station owners, road builders,
land developers all lobbying for better streets81. The growing ecosystem around the car and roadbuilding had expanded its reach and complexity. Businesses argued that traffic congestion
dropped real estate values by increasing the cost of doing business downtown and viewed
highway-building as a form of “social and economic therapy” and justified public spending on it
77 Weingroff, R. F. (1996, Summer) Federal Aid Road Act of 1916: Building The Foundation. Public Roads. US DOT
Federal Highway Administration .https://highways.dot.gov/public-roads/summer-1996/federal-aid-road-act-1916-
building-foundation Retrieved September 17, 2024
78 Jackson, K. T. (1987). Crabgrass Frontier: The Suburbanization of the United States (1st ed.). Oxford University
Press. p164
79 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
80 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
81 Jackson, K. T. (1987). Crabgrass Frontier: The Suburbanization of the United States (1st ed.). Oxford University
Press. p164
40
using the theory that improvements would pay for themselves by increasing property-tax
revenues along the route82
.
From 1919 states started to introduce fuel taxes, mainly to fund rural road-building, but
more significant funding soon came from federal coffers. Further national legislation, the Federal
Road Act 1921, made 200,000 primary roads eligible for federal funding on a fifty-fifty matchfunding basis and in the following decade the total miles of highway surfaced across the country
doubled83. At this stage though the percentage funding breakdown between state and federal
governments was about 57 to 43, with user fees (gasoline taxes and auto licenses) being the main
sources of funding rather than general taxation84
.
With the number of automobiles increasing significantly from 1925 to 1929 (up 30% from
20.1 million to 26.7 million85) traffic congestion started to intensify. More fundamental changes
were needed and transportation planners argued for more parkways or expressways that had the
capacity and could handle the need for speed and safety. There was much agreement that these
roads would solve the congestion problem for good by allowing only limited access and creating
grade separation to enable greater traffic volumes at higher speeds and with fewer collisions86
.
The funding streams though were not enough to cope with growing demand so the federal
government itself introduced its first gasoline tax in 1932 - a specific funding source for highways.
The tax raised funded construction and also enabled the federal government to have more control
over transportation priorities and policies and power shifted from local to state and federal
82 Jackson, Kenneth T. (1987) Crabgrass Frontier: The Suburbanization of the United States, Oxford University
Press.
83 Larry R. Ford (1994) Cities and Buildings Baltimore: Johns Hopkins University Press p 233.
84 Childs, W. R. (2017, August). How Public and Private Enterprise Have Built American Infrastructure. Origins
https://origins.osu.edu/article/how-public-and-private-enterprise-have-built-american-infrastructure
85 Brown, J. R., Morris, E. A., & Taylor, B. D. (n.d.). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2)
86 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2), 166.
41
government87. Cities had to persuade state and federal governments if they wanted support for
road-building, which revealed different perspectives and priorities: in contrast to cities, state and
federal government prioritized rural projects and resisted funding urban projects88 in addition to
favoring projects to improve travel between cities, rather than in-and-around cities. The
government support given to road-building (and in turn to the auto ecosystem) was not however
provided equally to transit.
The rise of the automobile led to a replacing of transit use which was not expected. Foster
notes that, “mass use of the automobile, at least as a tool for urban commuting, caught too many
experts by surprise. For these and other reasons, neither street railway managers nor public
officials took effective action to bolster mass transit systems during the 1920s”89. Foster notes,
“Hindsight makes clear that the explosion of automobile suburbs signaled the demise of the
electric railway”90. Jackson notes how this demise was accelerated by a perception that streetcars
were a private sector concern and not something that should be propped up by government and
by regulation which prevented many streetcar companies from increasing fares. With General
Motors buying up streetcar lines and replacing them with buses, this form of transit became
unviable91
.
Road-building, for the federal government, had a purpose beyond mobility; it was a tool
to boost the economy; it was undertaken to create jobs, not just to provide transportation
facilities92. Jobs in the 1920s were plentiful, fueled by the booming automobile industry, but after
87 Taylor, B. D. (2000). When finance leads planning: Urban planning, highway planning, and metropolitan
freeways in California. Journal of Planning, Education and Research, 20 (2), 196–214.
88 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
89 Foster M (1981) From Streetcar to Superhighway Temple University Press (Philadelphia) p46
90 Foster M (1981) From Streetcar to Superhighway Temple University Press (Philadelphia) p49
91 Jackson, K. T. (1987). Crabgrass Frontier: The Suburbanization of the United States (1st ed.). Oxford University
Press. p170
92 Altshuler, A. A., & Luberoff, D. E. (2004). Mega-Projects: The Changing Politics of Urban Public Investment (1st
ed., pp. ix–ix). Brookings Institution Press.
42
the 1929 crash and ensuing Great Depression unemployment rates soared to 25% by 1933. Even for
those who had jobs, reduced wages led to economic hardship leading to profound social changes
including migration from the Dust Bowl and social unrest. President Roosevelt’s New Deal, which
aimed to create millions of jobs, found plenty of work in building road infrastructure. So
supporting the automobile industry at this crucial time, including through large-scale roadbuilding, had significant appeal to the federal government, providing a way to help build the
country out of the Depression and stave off a perceived socialist threat. Transportation projects
and job-creation projects were convenient travelling companions.
The appeal wasn’t just economic and mobility-related – there were political
considerations too. Infrastructure projects are generally popular with politicians because they are
something concrete (literally) that they can point to and say they have achieved. Flyvbjerg
proposed the “political sublime,” to describe the rapture politicians get from building monuments
to themselves and for their causes93. Because consequences of projects are often so complex it’s
hard (except long after they’ve been built) to determine fault and with road-building the
consequences at least in the short-term, is positive media coverage with ribbon-cutting
photographs, and more freely flowing traffic that helps local business. With the lack of investment
in transit, that happened alongside the growing perception that it was old technology, there soon
became no alternative and a point was reached whereby road-building was the only way forward.
The construction of freeways into the mid-20th century was made possible by a set of
policies and financial mechanisms that facilitated large-scale road-building. At the center of this
process was the "Grand Bargain" involving fuel taxes earmarked for the Highway Trust Fund,
93 Flyvbjerg, B. (2012). Why mass media matter, and how to work with them: Phronesis and megaprojects, in
Flyvbjerg, B., Landman, T, Schram, S., eds., Real social science: Applied phronesis (pp 95–121). Cambridge, England:
Cambridge University Press.
Flyvbjerg, B., (Ed.). (2014). Megaproject planning and management: Essential readings. Cheltenham, UK: Edward
Elgar.
43
state responsibility for construction and ownership, and significant federal financial support. This
system enabled rapid freeway expansion across the country, but also came with significant social
and environmental consequences, which were often overlooked in the rush to build.
In July 1954, President Dwight D. Eisenhower set out his motives for supporting a strategic
highway network in a speech in July 1954 listing the problems to be overcome, which were very
much about the Interstate Highway being a silver bullet for the perennial problems of the nation’s
roads:
● Safety - an annual toll of nearly 40,000 killed (a similar figure to 2023 in fact) and 1.3
million injured.
● Congestion – wasting otherwise productive time.
● Economy - bad roads nullify the efficiency in the production of goods by inefficiency in
their transport.
● Defense - "the appalling inadequacies to meet the demands of catastrophe or defense,
should an atomic war come."94
These arguments – notably the familiar silver bullets of safety and congestion were
sufficient to convince those who needed to be convinced and the Interstate Highway System (ISH)
started to be built in the 1950s with the federal government paying 90% of its costs. In doing so it
reflected the standardization to which the Lincoln Highways aspired including limited access, a
design to accommodate speeds of 50-70 mph, a minimum of two travel lanes in each direction,
12-foot lane widths, 10-foot paved right shoulder and 4-foot left shoulder. All these features
94 Weingroff, R. (2023, June 30). Original intent and purpose of the interstate system: 1954-1956. U.S. Department
of Transportation. https://highways.dot.gov/highway-history/interstate-system/original-intent-purpose-interstatesystem-1954-1956
44
enabled greater speed even though the death toll can be seen to be largely about excessive
speed.
From the 1950's the federal share of funding changed from fifty-fifty to 80% for nonInterstate system road projects and 90% for Interstate system projects95. This generous funding
formula made freeway construction an attractive proposition for state governments, as they could
access massive amounts of federal money with relatively little financial commitment on their part.
It also incentivized states to expand their freeway networks quickly. However, in exchange for this
funding, states had to adhere to strict federal design standards.
The federal government’s financial support came with conditions: states were required to
build highways that met federal design standards. These standards were meant to ensure
uniformity and safety across the interstate highway system. Design specifications covered
everything from the width of lanes and the strength of pavement to the inclusion of interchanges
and the separation of traffic flows. While these standards created a more consistent and reliable
network, they often led to freeway designs that prioritized vehicle speed and efficiency over local
considerations, such as community cohesion or environmental preservation. In urban areas, these
design standards often dictated the construction of large, multi-lane freeways that cut through
neighborhoods, leading to displacement, increased pollution, and long-term changes to the city
fabric.
One of the most significant aspects of freeway construction during this period was the
absence of an environmental impact report (EIR) process which wasn’t in place until 1970 with the
passing of the National Environmental Policy Act (NEPA). Prior to this, freeway projects could
proceed without a formal evaluation of their effects on air quality, ecosystems, noise pollution, or
community displacement. As a result, freeways were often built through densely populated urban
95 Federal-Aid Highway Program (FAHP) (2021, March 1) In Brief Congressional Research Service.
https://crsreports.congress.gov/product/pdf/R/R44332
45
areas, displacing thousands of residents, disproportionately affecting communities of color, and
contributing to the degradation of natural environments. The lack of an EIR process also meant
that concerns about long-term environmental sustainability were rarely considered in the planning
stages of freeway construction, which focused instead on moving as many cars as quickly and
efficiently as possible.
America, being a huge wealthy country with a large population spread across it was
bound to need and develop a strategy road network. What wasn’t inevitable was the change into
an auto-centered culture. While the funding for roads served many purposes the fact that there
wasn’t commensurate funding of transit prevented Americans from having the free choice to have
a transit or auto-oriented lifestyle. The funding of road-building led to such an extensive overengineered system that it was impossible for the country to be anything other than car-centric. A
tipping point for this change was perhaps reached in 1954 when President Eisenhower appointed
a committee to study the country’s highway requirements and made Lucius Clay – a member of
GM’s board of directors - its chairman. At the heart of the findings of that committee was an
inevitable accommodation for an auto-centric future.
Freeway-building: technological determinism, social constructivism or coevolution?
The history of freeway-building shows how it was urged on by a wide spectrum of actors,
including successive governments and a broad coalition of industries involved in the selling of
motor vehicles and the construction of roads. There was strong political drive for highwaybuilding; governments provided support and, importantly, funding to pay for it all. They were
responding to the farmers, to the need for a reliable mail service, to help downtowns, to ensure
troops could move fast in the event of attack, to harness the power of trucks for goods delivery,
and to create jobs in times of recession.
46
This was not technological determinism, this was a vast number of interests calling for the
new road-building engineering technology to enable them and their industries to profit and
expand. It was an array of actors who had political, economic and societal objectives that drove
and shaped the development of the technology. A key actor was the consumer and the public was
presented with road-building as the only conceivable solution to the congestion problem.
While there were many stakeholders involved in the building of freeways this doesn't
necessarily point to simple social constructivism. In freeway-building there were hallmarks of
coevolution at work, notably complexity in organizations involved and momentum with the
technological and engineering innovations that made it hard for society to control. It would have
been very hard to rein in freeway-building after it had reached a tipping point which was perhaps
passed when the federal government provided hefty financial support. A tipping point for
highway-building had been reached, and as Holley has explained, ‘The revolution that got the
nation's highways out of the mud had developed the momentum which would carry it for a
generation. The elements of the revolution are evident: federal and state funds; federal and state
organisation; an impressive array of technological developments, the machines of the trade;
knowledge of materials and methods, scientific and practical; educated engineers and contractors
with the practical experience required to apply their learning; skilled labourers to operate the
increasingly complex machinery; and tax-paying citizens sufficiently persuaded that the high cost
of paved highways was fully justified by the economic advantages. These factors, taken together,
added up to the highway revolution of 1895 to 1925’96
.
The desire on the part of cities and states to take advantage of a bargain deal with the
federal government on funding, which couldn't be guaranteed to be there in perpetuity,
encouraged this sense of momentum to build. The complexity and this momentum potentially
96 Holley, I.B. (2008), The Highway Revolution, 1895 to 1925. Carolina Academic Press, Durham, p161
47
points to a coevolutionary technological development process but it was certainly a socially
constructed one and not a matter of technological determination.
In other ways coevolution can be seen to have been at work as technological innovation
and societal changes mutually influenced one another. Post-World War II suburban growth and
the rise of car culture drove the demand for infrastructure that could connect urban centers with
sprawling suburbs. Freeways, as a technological solution, reinforced these societal trends, making
suburban living and personal automobility more accessible while also shaping cultural norms
around freedom and mobility.
Advances in civil engineering and road surface materials enabled the freeway system, but
societal feedback loops further entrenched their dominance. Freeways reshaped cities, reinforcing
suburbanization and decentralization, which in turn created further demand for highways.
Economic growth facilitated by freeways spurred investment in more infrastructure, creating a
self-reinforcing loop that reflected coevolutionary dynamics.
Transportation Network Companies (TNCs) - A brief history of their development
The most developed use case for AVs currently in the US is the ‘robotaxi’ – that is AVs,
licensed to carry fare-paying passengers. Robotaxis are already competitors to TNCs in the cities
in which they operate. App-enabled TNCs are the forerunner of robotaxis and so fitting that it
should be included as a key historical technology.
When Transportation Network Companies (TNCs) first emerged, they were seen as a
revolutionary advancement in urban mobility. These companies introduced an innovative business
model that transformed how people hail rides, replacing the traditional method of flagging down
taxis or phoning for a ride with a few taps on a smartphone. The technology that enabled this
disruption was centered on sophisticated software platforms that connected drivers—who were
48
independent contractors rather than employees—with riders in real time. The platform optimized
driver routes, pricing, and matching, making rides faster, more convenient, and often cheaper
than traditional taxi services. The appeal of flexibility for drivers and the convenience for riders
quickly allowed TNCs to scale in cities across the world.
The software platform not only made the TNC business model possible but also provided it
with a distinct competitive edge. By operating as a network company, Uber and Lyft could grow
without the capital and regulatory overhead of owning vehicles or employing drivers. This created
a highly scalable, decentralized system where the primary cost was the maintenance of the digital
infrastructure, while drivers bore the costs of car ownership and maintenance. The use of
algorithms enabled dynamic pricing (or surge pricing), allowing TNCs to balance supply and
demand more efficiently than traditional taxi companies. In time this enabled TNCs to dominate
markets or threaten to capture entire sectors of the hired transport industry. By leveraging their
technology platforms, TNCs could outcompete taxis on price, convenience, and availability,
pushing many local taxi industries out of business or to the brink of collapse.
The technology was part of tech trends of the 2010s – disrupting traditional sectors and
contributing to the perceived dawn of a new sharing economy. The apparent megatrend of a
sharing culture gained traction such that in in 2015, PWC claimed that it is ‘one of the most
important global trends and success stories of recent years and, with all certainty of years to
come’ and that, ‘43% of consumers take the view that ownership is mainly a hassle: choosing
takes time, the cost of acquisition is high, the asset has to be repaired and stored, and once the
purchase has been made the opportunity to make further choices is lost.97’The rise of app-enabled
companies such as Uber, Lyft, Airbnb, and WeWork all pointed to a future in which ownership was
97 PWC (2015) Sharing or Paring – the Growth of the Sharing Economy. PricewaterhouseCoopers International
Limited. https://www.pwc.com/hu/en/kiadvanyok/assets/pdf/sharing-economy-en.pdf
49
old school and the Millennials were leading the way out of this pointless consumerism into a
future of unencumbered borrowing.
Uber and Lyft have dominated the TNC market for nearly fifteen years. Uber was launched
publicly in San Francisco in 2011 at a time when they were generally called ‘rideshare’ companies
because of the expectation that they would enable a matching of people going the same way and
so share rides. It was thought that ridesharing would reduce car ownership, with consequent
societal benefits, including reduced energy consumption, emissions, congestion, and need for
parking infrastructure98. It was thought they’d reduce time and money and have positive
environmental impacts99
.
Like automobiles at the start of the twentieth century, the target audience in the early
years of TNCs was wealthy people because they were a relatively expensive way to travel. Uber
provided only luxury cars and rides costed 50% more than a taxi100 and were, “barely more than a
private luxury car service for Silicon Valley's elite techies”101. Ease and convenience were key to
their success and popularity. But while popular with the public they were certainly not popular
with city governments.
When TNCs were first launched onto cities the public sector was caught on the hop102; they
had not been part of the development of the technology and the sector’s advice was not sought.
Once operating in cities Uber dropped prices by up to 30 percent aiming to get customers to use
the service instead of rental cars, public buses, and subways – and forcing Lyft to cut prices too.
98 Cohen, A. and Shaheen, S. (2016) Planning for Shared Mobility. American Planning Association PAS Report 583.
99 Deakin, E., Frick, K. & Shively, K. (2012) Dynamic Ridesharing. University of California Transportation Center,
Working Papers. qt1c0421x7. University of California Transportation Center.
100 McAlone, N. (2016) This is How Uber used to look when it first started out – and how it changed over time
Business Insider https://www.businessinsider.com/ubers-design-history-2010-2016-2016-2
101 Wohlsen, M. (2014, January 3) What will Uber Do With All That Money From Google. Wired.
https://www.wired.com/2014/01/uber-travis-kalanick/
102 R. R. Clewlow, (2017, October 11) New Research on How Ride-Hailing Impacts Travel Behavior. Planetizen
(https://www.planetizen.com/features/95227-new-research-how-ride-hailing-impacts-travel-behavior
50
Traditional taxi companies responded with a combination of protest, legal action, and attempts
at modernization themselves, especially by providing an app-based platform. The aim of the price
drops was to disrupt the taxi market and achieve a virtuous circle: lower prices leading to more
customers and usage, leading to a larger supply of cars and busier drivers, enabling the TNC to
further cut prices and put more pressure on competitors103. Cheaper rides were of course popular
and it was hard for politicians to reject more affordable means of transportation.
TNCs had a sense about them that they were the David among the Goliaths of the
transportation world, battling to do what’s right against established monopolistic power of
government and the medallion taxi industry. Uber had success in by-passing regulation – Wired
magazine noted in 2014, “Reaching Uber's goals has meant digging in against the established
bureaucracy in many cities, where giving rides for money is heavily regulated. Uber has won
enough of those fights to threaten the market share of the entrenched players”. The Uber
approach was a “sharp-elbowed operating model, where it essentially storms into a city and
opens for business, all but daring regulators to stop it in the face of a public that enjoys its
services”104. It was a practice of, ‘show up, break the law and announce that they were going to
keep doing it’105
But once the technology was rolled out in cities governments did respond with regulations
and policies although some conceded at least partial defeat in the face of a public that very much
liked the Uber service. Uber got used to mobilizing its users who were an important player in
shaping how the technology developed. The Washington Post reported in 2014, “A notice sent to
103 Stone, B. (2017, February 7) The Upstarts Reprinted with permission of Little, Brown and Company in Biggs, J.
The inside story of the rise and rise of Uber. TechCrunch https://techcrunch.com/2017/02/07/the-inside-story-ofthe-rise-and-rise-of-uber/
104 Robillard, K. (2014, August 26) Plouffe at Uber A Tough Start. Politico.
https://www.politico.com/story/2014/08/david-plouffe-uber-110366
105 Attributed to David Sutton, spokesman for Who’s Driving You?, the taxi industry’s campaign against the ridesharing industry, quoted in Robillard, K. (2014, August 26) Plouffe at Uber A Tough Start. Politico.
https://www.politico.com/story/2014/08/david-plouffe-uber-110366
51
Uber users in Virginia included the email address and phone number of the ordinarily low-profile
official in charge of the decision. The notice instructed the company’s supporters to demand that
the DMV “stand up for you.” Hundreds of them did and, by Sunday, Commissioner Richard
Holcomb’s inbox was flooded”106
. Virginia Gov. Terry McAuliffe responded, “I am pleased that we
were able to work together to find a swift solution that will provide Virginia’s workers, students
and families with more transportation options”.
Cities including Austin, Las Vegas, Denver and Miami resisted Uber but they were
eventually persuaded by sustained popular demand. In Miami for example 400 Uber staffers
descended on the city for a ‘workation’, partying, handing out Uber postcards, putting up posters
on light poles, and using the hashtag #MiamiNeedsUber; the Mayor eventually conceded107
.
As TNCs grew in number, they quickly began to disrupt local taxi industries that found
themselves outcompeted on price, convenience, and availability. With an app-based interface,
transparent pricing, and faster response times, TNCs rapidly attracted a customer base. Dynamic
pricing, or “surge pricing,” allowed Uber and Lyft to manage supply and demand by raising prices
during peak periods, further improving efficiency.
Regulatory challenges emerged almost as soon as Uber and Lyft began to expand. TNCs
operated in a legal gray area, often bypassing the stringent regulations that traditional taxi
services were subject to, such as licensing, medallions, and fixed fares. Many city governments
and taxi associations saw this as an unfair advantage, leading to legal disputes, protests, and the
creation of new regulatory frameworks to govern ridesharing. Some cities banned TNCs
altogether, while others created new legislation to allow them to operate under specific
106 Helderman, R. (2014, December 13) Uber pressures regulators by mobilizing riders and hiring vast lobbying
network. Washington Post https://www.washingtonpost.com/politics/uber-pressures-regulators-by-mobilizingriders-and-hiring-vast-lobbying-network/2014/12/13/3f4395c6-7f2a-11e4-9f38-95a187e4c1f7_story.html
107 Stone, B (2017) The Upstarts Reprinted with permission of Little, Brown and Company in Biggs, J. The inside
story of the rise and rise of Uber. 7 February 2017, TechCrunch https://techcrunch.com/2017/02/07/the-insidestory-of-the-rise-and-rise-of-uber/
52
conditions. In some cases, TNCs fought these regulations aggressively, using their platforms to
mobilize public support and lobbying local governments.
Despite these legal battles, TNCs continued to grow, with Uber leading the way. By 2014,
Uber had expanded to over 100 cities worldwide, including major international markets. In
addition to traditional ridesharing, Uber and Lyft began experimenting with new service models.
UberPool and Lyft Line were introduced as carpooling services, encouraging multiple passengers
to share rides, theoretically reducing congestion and emissions.
Having to deal with individual states and cities is a time-consuming business – far easier
to have national rules. So it perhaps wasn’t a surprise when Uber hired President Obama’s former
campaign manager as senior vice president for policy and strategy and soon after former
Transportation Secretary, Ray LaHood, urged a federal approach.
TNC had to spend a great deal of money on lobbyists. The Washington Post reported that
Uber had hired at least 161 lobbyists across the US over the previous two years and spent
$475,000 in Sacramento from July to November 2014 alone to lobby California lawmakers.108
.
While tech companies may gather support of politicians, the same politicians may feel obliged to
the traditional taxi industry. On July 31, 2014 the independent, non-partisan organization that
campaigned for open government, the Sunlight Foundation, reported that, “the taxi cab industry
has donated at least $3,500 to the political war-chests of state legislators for every $1 that Uber,
Lyft and Sidecar gave”109
.
The maturing of Uber can be seen under the leadership of a new CEO in 2017. It was a
more consensual Uber that appeared – one that recognized technology companies couldn’t force
108 Washington Post (2014, December 13). Uber pressures regulators by mobilizing riders and hiring vast lobbying
network. Washington Post. https://www.washingtonpost.com/politics/uber-pressures-regulators-by-mobilizingriders-and-hiring-vast-lobbying-network/2014/12/13/3f4395c6-7f2a-11e4-9f38-95a187e4c1f7_story.html
109 Oklobdzija, S. (2014) Will Big Taxi’s massive political spending advantage leave ride-sharing groups stranded?
Sunlight Foundation
53
their way into a city and expect to flourish. Responding to Uber’s ban in London, Uber’s chief
executive, Dara Khosrowshahi said, “The truth is that there is a high cost to a bad reputation,” he
wrote. “It really matters what people think of us, especially in a global business like ours110”.
Khosrowshahi pointed to two key elements of Uber’s way out of the mess – improve public
opinion and work more closely with the public sector. Far from technology having a life of its own
– with a determinist force – this was an acknowledgement that it was very much being shaped
and limited by the public and the public sector.
Power over TNCs moved away from local government though. In California, for example,
state legislators preempted local and city regulators by creating a separate legal category that
made the platform-based companies no longer subject to city and county regulatory control111
.
TNCs became subject to regulation by the California Public Utilities Commission requiring just one
permit that would allow a fleet to operate statewide. This contrasted with the rules for taxi
companies whereby each city and county could require a permit each.
These inconsistent rules in California reflected a state attitude that some have deemed to
work in favor of tech companies whatever the fairness or outcomes of the proposal. Pernicka
found in correspondence with the Greater California Livery Association in 2017, “There is a
distinction I think in California, probably in the United States in general, and that is we are in love
with technology. . . . Anything that might impede technological development or implementation is
not even considered here. If you go to a California legislature and you say to them, in the spirit of
fairness to all drivers of these services, here is what should happen in terms of an equal labor
110 Butler, S., Topham, G. (2017) Uber stripped of London licence due to lack of corporate responsibility The
Guardian https://www.theguardian.com/technology/2017/sep/22/uber-licence-transport-for-london-tfl
Retrieved September 22, 2024
111 Pernicka, S. (2019). The disruption of taxi and limousine markets by digital platform corporations in Western
Europe and the United States: Responses of business associations, labor unions, and other interest groups. Working
paper. Retrieved from Institute for Research on Labor and Employment, University of California, Los Angeles.
https://irle.ucla.edu/wp-content/uploads/2019/06/Disruption-of-Taxi-and-Limousine-Markets.pdf
54
market . . . the response is the implementation of that requirement could dampen technological
development’. This attitude of prioritization of technological advancement over other social and
economic considerations is one that was apparent too in the relaxed approach of the California
legislature to the testing of AVs as we shall see. This reflects a cultural perception of technology
as a positive tool for improvement and Progress.
TNCs: technological determinism, social constructivism or coevolution?
TNCs and transportation disruptive technologies – such as e-scooters – provide good
example case studies of the weakness of technological determinism as it applies to mobility
technology. Much was made of shared mobility as a ‘Fourth Revolution’ but TNCs have not
significantly changed the mobility landscape. Pew research published in 2019 found, few adults
use them as part of a regular routine and only 2% say they use them every day or almost every day
and over two-thirds of users use them less than once a month. “Put differently” they conclude, “a
mere 4% of the U.S. adult population today uses ride-hailing apps on a weekly basis – a share that
is largely unchanged from 2015, when 3% of Americans reported being weekly riders112”. We also
know that people haven’t given up their cars so essentially all we have is a taxi that we order
through a smart phone rather than a phone: this is not particularly revolutionary. The notable
difference is that they can eat into the transit market or the other active travel share.
Although trials were run using TNCs as a complement to transit this hasn’t developed in a
consistent way. TNC companies were looking for a use case that would be profitable and pooling
failed to be so. Profit (or actually less of a loss) depends on selling a product consumers want to
buy; but it emerged that what consumers wanted was discounted taxi rides that were easy to
112 Jiang, J. (2019) More Americans are using ride-hailing apps. Pew Research Center
https://www.pewresearch.org/short-reads/2019/01/04/more-americans-are-using-ride-hailing-apps/
55
summon. So while laudable use cases have been trialed the eventual set artifact was the one
most able to generate revenue for the companies involved – and that was a vehicle to perform the
role of taxi in the most profitable geographic areas which favor the density of cities, a younger
age profile with more disposable income.
As with the development of the auto and freeway-building, with TNCs there have been
many stakeholders involved in their development including the companies directly involved,
government and users of the services. Does this mean they have been socially constructed? The
technology made rides more convenient which is what consumers wanted and with competitive
pricing and ease of ordering a ride it became hard to curb their use. The speed with which TNC
companies were able to scale brought a momentum that was a fundamental part of the
disruptive technology playbook. This complexity and momentum points to a coevolutionary
development process.
Is technological determinism at play?
The history of the three transportation technologies reveal that their development hasn’t
been a matter of technological determinism but it doesn’t seem to have been simply a process of
social construction. While there have been many actors involved in their development there is a
strong sense of coevolutionary forces at play: transportation technology development doesn't
determine society; and society doesn't wholly determine transportation technology development.
In all cases a point was reached whereby the form and function of the technology became set,
momentum was created and while governments have been able to impose controls on TNCs they
were not able to eliminate them.
While we may instinctively feel technological determinism would be an unwelcome
dynamic in democratic cities, how does coevolution fare? Is it producing the cities we want? The
56
next section will briefly consider the impacts of the three technologies before a deeper
examination of the processes that contributed to these outcomes, and at what point the
technologies moved from a collaborative, controlled dynamic of social constructivism to a force of
coevolution that neither government, tech companies nor the public could control. The findings
will be used later in the dissertation to see what lessons might be learnt in the treatment of AVs
before they gain momentum and become too complex to unpick from the fabric of our
transportation system.
Impacts
Cars have of course had a revolutionary impact on mobility. They satisfied a real need for
transportation, something as basic as food, clothing, shelter113. People bought them because they
met transportation needs better than existing alternatives, offered new possibilities114 and
provided a means of escape from the increasingly dense city.
Those who had looked to the future a hundred years ago were quite right when they
predicted the great increase in access, enabling people to develop friendships and relationships
outside the home, and enabling more choices in where to live and work. People who lived in the
country in particular gained as their mobility increased dramatically. The automobile brought a
significant pleasure to people’s lives by enabling them to more easily visit friends and relatives
without a complicated network of transportation or being tied to timetables and routes of transit
providers.
113 Interrante J. The Road To Autopia: The Automobile And The Spatial Transformation Of American
Culture. Michigan quarterly review. 1980;19(4):503
114 Interrante J. The Road To Autopia: The Automobile And The Spatial Transformation Of American
Culture. Michigan quarterly review. 1980;19(4):503
57
Ikuta points to women as being significant winners too as the automobile expanded their
sphere of activity115. The car also fitted extremely well with the fundamental American values of
freedom and individualism and the extension of the American Dream.
Clear winners from the advent of cars were those invested in the massive expanse of
ecosystem they created: auto manufacturers, the workers in the industry and other industries
including, petroleum, infrastructure and construction, the leisure, retail and service industry (such
as drive-in theaters and fast food restaurants), real estate speculators and those serving the car
industry – oil companies, tire and rubber companies and others.
There were many losers though – those killed or injured in road crashes and their families
and the slower, more insidious health effects of poor air quality and the stress and noise created
by busy roads . Pedestrians, cyclists, and other vulnerable road users were disproportionately
affected, as urban streets were redesigned with the car in mind. Efforts to improve safety focused
on clearing people out of the way of vehicles – impacting others’ mobility - and encouraging the
development of faster, more hazardous but more ‘efficient’ car travel over safer, slower, more
pleasant environments.
Urban design itself was profoundly transformed to accommodate the car, often at the
expense of other forms of mobility. Cities reoriented themselves around the automobile, creating
wide, multi-lane roads, expansive parking lots, and drive-through facilities. Walkability and public
spaces were sacrificed to facilitate car movement, with sidewalks narrowed, crosswalks
deprioritized, and pedestrian places shrunken. Suburbs sprawled outward, with residential and
commercial zones disconnected and only accessible by car. This car-centric urban design made it
difficult to navigate cities without a vehicle, further entrenching the car’s dominance in shaping
how people lived, worked, and moved around.
115 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
58
There were other significant consequences, particularly in congestion and livability. One of
the myths surrounding the car was that it would solve congestion problems by providing
individual freedom of movement. But, as cars became more affordable and ubiquitous, congestion
in cities worsened. Roads designed to handle fewer vehicles became overwhelmed, and the
solution of building more roads often backfired, as the additional capacity created induced
demand. Rather than dispersing urban traffic, the car centralized congestion in many areas,
creating gridlock in city centers and contributing to sprawling, car-dependent suburbs.
For those without cars, the rise of the automobile eroded mobility, particularly in cities
where investment in public transportation declined. As cars took over urban spaces, transit
systems often suffered from neglect and underfunding, leading to fewer options for those reliant
on buses, trains, or walking. This particularly affected lower-income residents, the elderly, and
people with disabilities, as well as younger populations who could not yet drive. In the middle of
the last century, the city became restructured at great public expense to meet the needs of a
minority of upper and middle class motorists116. As public funds were diverted toward roadbuilding and car infrastructure, transit systems were left to deteriorate, further cementing the
dominance of the automobile. Streetcar systems that once provided widespread urban mobility
were dismantled to make way for cars, leaving residents with few viable alternatives to driving.
Public policy became captured by the car’s needs, as governments sought to make driving
cheaper and easier. Fuel taxes were kept low to keep driving affordable, despite the
environmental and infrastructure costs. Driver license requirements were made relatively easy to
meet, ensuring mass car ownership. Traffic management practices, such as prioritizing cars at
intersections, imposing fines or delays on pedestrians for jaywalking, and creating parking
mandates for new developments, all worked to cement the car’s primacy in urban life. These
116 Flink, J. J. (1990) The Automobile Age. MIT Press
59
policies often externalized the costs of driving—creating pollution and road maintenance and
traffic management problems — a cost borne by society at large, rather than making drivers pay
for the true cost of their car use.
Freeway-building
Freeways have unquestionably brought immense increases in mobility, productivity,
and prosperity. Freeways provide a very effective way of moving vast numbers of vehicles at
speed. Many studies have evidenced substantial economic benefits from freeways because of
their combination of speed, convenience and relative safety117. The movement of goods makes a
particularly positive contribution to the economy and it's almost impossible to think of an
advanced nation without a strong network of high capacity roads.
A colossal amount of money was spent on roads which soon dwarfed what was spent on
transit. From 1956 to 1970, the federal government spent $70 billion for highways and only $795
million for rail transit”118. Inevitably the quality and reliability of transit services declined and the
losers here were those who were not able or could not afford to drive.
The winners in road-building were notably the industries associated directly with building
them but also the auto-highway ecosystem more broadly. As with the development of the auto,
people who lived in rural areas particularly benefited from road-building. Early in the 20th century
finding a market for goods was hard in rural areas where roads became impassable at some
times of the year when businesses just had to stop production because they couldn’t get their
produce to their intended market during wet or snowy weather. In the early part of the 20th
117 Such as Cox & Love, 1996; Friedlaender, 1965; Nadiri & Manumeas, 1996; FHWA, 1996
118 Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed.). Cambridge
University Press p12-13.
60
century, rural life was greatly enhanced by the combination of the automobile and road-building
with improvements to leisure, social and economic opportunities. In terms of productivity of mode,
compared to well-used trains, freeways are an inefficient way to move people and goods; in
terms of space, speed and safety trains are better. But for the individual traveler trains are not as
comfortable and convenient as cars because of the latter’s ties to schedules, placement of
stations, and lack of privacy.
People who could afford to travel also benefited - although in time, as cost came down,
more and more people could. This divide between low income people was more dramatically
demonstrated in decisions on the location of the interstate highways. Many low-income
neighborhoods were divided by the new highways, while their wealthier counterparts often got
better treatment or were able to stop the proposed projects altogether119
.
It took until the 1960s for serious macro level assessment of the impacts of freeways on
the environment to emerge, notably in the lead up to the 1970 National Environmental Policy Act
(NEPA). Issues that had to some extent been foreseen by planners over half a century before had
emerged as serious problems.
The impact of air quality on those living near high-capacity roads became more apparent
as time went on. In the late 1940s concerns started to grow about the pollutants emitted by cars
as the problems of smog and consequent health impacts became more obvious. By the late 1950s
Los Angeles residents were regularly receiving smog alerts - in 1959 the city government reported
that smog caused widespread eye irritation among residents on 187 days of the year and by 1962
this had grown to 212120. The increasing number of automobiles and the growth in automobile
emissions through the twentieth century forced governments across America to seek ways to deal
119 Brown, J. (2005). A tale of two visions: Harland Bartholomew, Robert Moses, and the development of the
American freeway. Journal of Planning History, 4 (1), 3–32.
120 Fallon, M. (2018). Self-driving cars : The new way forward. Lerner Publishing Group.
61
with worsening air quality. Again, it was the poor who were worst affected and less likely to have
the option to move.
The effects of the freeways on inner-city African American communities were particularly
severe121 122. By the 1950s, many African American families were living in some of the poorest
areas of cities, close to the centers, as White families had left for the suburbs. With the demolition
and social disruption involved, interstates provided city governments with a means of shifting
Black families away from potentially valuable real estate to be replaced by new developments
that would attract more wealthy, spendthrift (White) customers or walled-off Black communities
from White123. Road-building showed that the process can be captured by interests with nothing
to do with mobility: for economic, political and social (and specifically, racial) reasons.
As described earlier, highway-building brought jobs and economic growth not just directly
but indirectly through the access to new housing in new suburbs. There were many benefits
around increases in mobility – but these benefits came with costs to cities which were not
generally recognized in the earlier part of the previous century. Cities and states supported
freeway- building for the range of benefits they were perceived to bring, “Inner city slums could
be cleared, blacks removed to more distant second-ghetto areas, central business districts
redeveloped, and transportation woes solved all at the same time - and mostly at federal
expense”124
.
121 Mohl, R. A. (2004). Stop the Road: Freeway Revolts in American Cities. Journal of Urban History, 30(5), 674–
706.
122 Silver, C. (1984). Twentieth-century Richmond: Planning, politics, and race. Knoxville: University of Tennessee
Press.
123 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
Press.
124 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
Press. p103
62
TNCs
Winners from the introduction of TNC include people who are tech-enabled, have a bank
account and can use a smartphone, which tended to exclude more of the very old, people with
some disabilities and the very poor. Clewitt and Gouri found that college-educated, affluent
Americans had adopted ride-hailing services at double the rate of less educated, lower income
populations. They also found they were used more regularly by people living in more urban
neighborhoods of cities, while only 7% of suburban Americans do so125; in part this reflected that
37% respondents used the service mainly as a response to problems of parking.
The findings are echoed in Pew research which in 2019 found: “roughly half of Americans
ages 18 to 29 (51%) say they have used a ride-hailing service, compared with 24% of those ages 50
and older. Those whose annual household income is $75,000 or more are roughly twice as likely as
those earning less than $30,000 to have used these services (53% vs. 24%). And over half of adults
with a bachelor’s or advanced degree (55%) say they have used these services, compared with 20%
of those who have a high school diploma or less.
Ridership, Pew found, also varies substantially in different types of communities. While
45% of urban residents and 40% of suburban residents have used a ride-hailing app, only 19% of
Americans living in rural areas have done so”126
. So TNCs have served a younger, more wealthy,
more urban and better educated population. Evidence from elsewhere found the main
beneficiaries of TNCs have been more specifically the white middle class who have been provided
with a more convenient way to order a taxi service – adding to their mobility options127
.
125 Clewlow, Regina R. and Gouri S.Mishra (2017) Disruptive Transportation: The Adoption, Utilization, and Impacts
of Ride-Hailing in the United States. Institute of Transportation Studies, University of California, Davis,
Research Report UCD-ITS-RR-17-07
126 Jiang, J. (2019) More Americans are using ride-hail apps. Pew Research. https://www.pewresearch.org/shortreads/2019/01/04/more-americans-are-using-ride-hailing-apps/
127 Gehrke, S. R., & Huff, M. P. (2024). Spatial equity implications and neighborhood indicators of ridehailing trip
frequency and vehicle miles traveled in the phoenix metro region. Transportation (Dordrecht), 51(1), 271–295.
63
An early disappointment was on the issue of sharing – it was hoped they could be a
demonstration of a new sharing economy. Called ‘rideshare’ companies in the early years, it soon
became apparent that the main purpose was not to share rides128. They were not shared as
originally portrayed and were instead a predominantly single-passenger mobility service129
.
Another disappointment for cities was on the claim that TNCs could increase reach and
flexibility of transit’s fixed-route, fixed-schedule and be a complement to small transit agencies
and in large cities130. Findings on the overall impact of TNCs were thin to start and quite slow to
emerge; this has at least in part been because the TNC companies have refused to share data131
.
In fact TNCs were notably seen to have decreased bus use132 and transit more broadly133 and
replaced transit use especially in higher income areas with greater transit availability134
.
Differences in transit impacts though have been found. Manville et al, looking at Southern
California, found that it appears TNCs do not replace transit use135 while an earlier study by Rayle
et al, looking at San Francisco, found TNCs compete with public transit for some individual trips
and that a third of TNC trips would otherwise have been made by transit136. Researchers at
Carnegie Mellon found that TNCs have displaced transit most in cities with high incomes and high
128 Zenner, A. (2015, January 14) The AP bans the term ‘ride-sharing’ for Uber & Lyft. Greater, Greater Washington.
https://ggwash.org/view/36979/the-ap-bans-the-term-ride-sharing-for-uber-lyft
129 Gehrke, S.R., Huff, M.P., Reardon, T.G. (2021) Social and trip-level predictors of pooled ride-hailing service
adoption in the Greater Boston region. Case Stud. Transport Policy.
130 Hall, J. D., Palsson, C., & Price, J. (2018). Is Uber a substitute or complement for public transit? Journal of Urban
Economics, 108, 36–50.
131 Manville, M., Taylor, B. D, & Blumenberg, E. (2018). Falling Transit Ridership: California and Southern
California. UCLA: Institute of Transportation Studies. Retrieved from https://escholarship.org/uc/item/0455c754
132 Babar, Y., & Burtch, G. (2020). Examining the Heterogeneous Impact of Ride-Hailing Services on Public Transit
Use. Information Systems Research, 31(3), 820–834
133 R. R. Clewlow, (2017, October 11) New Research on How Ride-Hailing Impacts Travel Behavior. Planetizen.
https://www.planetizen.com/features/95227-new-research-how-ride-hailing-impacts-travel-behavior
134 Meredith-Karam, P., Kong, H., Wang, S., Zhao, J. (2021) The relationship between ridehailing and public transit
in Chicago: A comparison before and after COVID-19. J. Transp. Geogr. 97, 103219.
135 Manville, M., Taylor, B. D, & Blumenberg, E. (2018). Falling Transit Ridership: California and Southern
California. UCLA: Institute of Transportation Studies. Retrieved from https://escholarship.org/uc/item/0455c754
136 Rayle, L., Dai, D., Chan, N., Cervero, R., & Shaheen, S. (2016). Just a better taxi? A survey-based comparison of
taxis, transit, and ridesourcing services in San Francisco. Transport Policy, 45, 168–178.
64
childless household rates137. So it seems that the same technology can have very different impacts
depending on where it is launched and the population served. In sum though, in a 2024 study it
has been found overall that, “Early optimism for ride hailing services to complement existing
public transit services and offer individuals another shared mobility service with reduced travel
costs and improved travel times have largely proven to be unsubstantiated.” 138
Competing with transit is a serious issue especially with so many transit agencies
struggling with falling ridership since Covid. Peeling people away from transit weakens the
service overall and riders suffer from reduced service level, poorer maintenance, weakened safety
from reduced natural surveillance and reduced money for service investment from the reduced
farebox.
Early research found positive signs that congestion and car ownership would be reduced139
but here too there were disappointments. Research from Carnegie Mellon found that Uber and
Lyft led to increased road congestion in terms of both intensity (by 0.9%) and duration (by 4.5%),
and an insignificant change in vehicle ownership140. San Francisco’s transit agency found that
congestion in the city grew with average morning peak arterial travel speeds decreasing between
2010 and 2016 by -26%, while afternoon peak arterial speeds decreased by -27%. Vehicle hours of
delays on the major roadways increased by 40,000 hours on a typical weekday, while vehicle
137 Ward, J. W., Michalek, J. J., Samaras, C., Azevedo, I. L., Henao, A., Rames, C., & Wenzel, T. (2021). The impact of
Uber and Lyft on vehicle ownership, fuel economy, and transit across U.S. cities. iScience, 24(1), 101933–101933
138 Gehrke, S. R., & Huff, M. P. (2024). Spatial equity implications and neighborhood indicators of ridehailing trip
frequency and vehicle miles traveled in the Phoenix metro region. Transportation (Dordrecht), 51(1), 271.
139 Gehrke, S. R., & Huff, M. P. (2024). Spatial equity implications and neighborhood indicators of ridehailing trip
frequency and vehicle miles traveled in the Phoenix metro region. Transportation (Dordrecht), 51(1), 271–295.
140 Diao, M., Kong, H. & Zhao, J. (2021) Impacts of transportation network companies on urban mobility. Nat
Sustain 4, 494–500. https://doi.org/10.1038/s41893-020-00678-z
65
miles travelled on major roadways increased by over 630,000 miles on a typical weekday 141
.
Other research also found increases in VMT 142 and congestion143 144
.
Other early claims about positive environmental impacts of TNCs were also found to be
wrong. Research by Michalek et al found, “Traffic externalities (i.e., congestion, crashes, and
noise) associated with the shift from personal vehicles to TNC vehicles are larger than emission
externalities by an order of magnitude, [it] increases the total net external costs in all cases we
tested, including when TNCs are all electric and powered by a zero-carbon electricity grid”145. On
energy use they found a 25−40% increase despite a more efficient TNC fleet (compared to the
overall fleet of private vehicles) and that shifting travel to a TNC vehicle increases net energy use
by 41−90%. In summary, lead author of the research, Jeremy Michalek, explained, “You create
lower external costs to society when you drive your personal vehicle, on average".
On economic impacts, research from Carnegie Mellon found that broadly, TNCs have
increased economic growth, employment and wages of unstable jobs but their effects have been
different in different cities146
.
There is an underlying theme here that has echoes in the introduction of the car and of
road-building – people want convenience and comfort – so they will buy the most comfortable
option that will get them to a destination in a way most convenient to them. The more
141 SFCTA. (n.d.) TNCs & congestion. SFCTA. https://www.sfcta.org/sites/default/files/2019-
05/TNCs_Congestion_Report_181015_Finals.pdf
142 Kaplan, S., Gordon, B., El Zarwi, F., Walker, J. L., & Zilberman, D. (2019). The Future of Autonomous Vehicles:
Lessons from the Literature on Technology Adoption. Applied Economic Perspectives and Policy, 41(4), 583–597.
https://doi.org/10.1093/aepp/ppz005
143 Barrios, John Manuel and Hochberg, Yael V. and Yi, Hanyi, (2019, April 3). The Cost of Convenience: Ridehailing
and Traffic Fatalities Available. https://ssrn.com/abstract=3361227
or http://dx.doi.org/10.2139/ssrn.3361227
144 R. R. Clewlow, (2017, October 11) New Research on How Ride-Hailing Impacts Travel Behavior. Planetizen
https://www.planetizen.com/features/95227-new-research-how-ride-hailing-impacts-travel-behavior
145 Michalek, J., Armanios, D., Nock, D (2022) Ridehailing Service Equity in Normal and Rare Conditions. Mobility
21, USDOT
146 Michalek, J., Armanios, D., Nock, D (2022) Ridehailing Service Equity in Normal and Rare Conditions. Mobility
21, USDOT
66
comfortable and convenient it is, the more trips they will make, the further they will go and the
more impact this will have on the environment.
TNCs, being easier to summon, could be a safety tool for example for women and indeed
Weber finds that the introduction of Uber in cities to be associated with a reduction of personal
crimes by 5% in treated cities (about 43 personal crimes a month, roughly 41 assaults) – largely
assaults due to alcohol147; correlation of course is not causation. But this doesn’t mean they bring
safety to a city: research by Barrios et al found the arrival of ride hailing is associated with an
increase of approximately 3% in the number of fatalities and fatal accidents, for both vehicle
occupants and pedestrians.
The coevolutionary dynamic in transportation technologies has brought some benefits but
many disbenefits – with less livable streets, a degraded environment and other negative impacts
impacting, especially, poorer, transit-dependent people. Might the same be true of autonomous
vehicles? There are many positive claims made about AVs but we don’t know whether the benefits
claimed will transpire. To understand this the next section will examine why there were gaps
between the claims made about the previous transportation technologies and their outcomes. It
will cover who was involved, what their motives were, and why claims were accepted. This will
help to understand whether the negative outcomes could have been foreseen and acted upon –
and prevented – and at what stage intervention needed to have happened. An important question
is why this happened and whether this could have been predicted before the point of complexity
and momentum arrived. If this could be understood we may be able to steer the development of
AVs before a point of complexity and momentum when they are no longer controllable.
147 Weber, B. S. (2019) Uber and urban crime. Transportation Research Part A: Policy and Practice, Volume 130,
pp 496-506,
67
The gap between hopes, predictions, expectations and reality
In order to assess the claims about autonomous vehicles this part of the literature review
now turns to the processes and actors involved in the development of historic transportation
technologies. These will be taken in turn and will include an examination of the hopes, predictions,
and expectations of the technologies and a consideration of the motives of the actors involved.
The automobile
The first advertisement for a motorcar in the US had appeared in Life magazine on 3 July
1902, promoting its reliability, power, speed, endurance and ease of driving148 for the wealthy
potential buyers. The following year publications couldn’t keep up with car-maker demand for
advertising which emphasized durability, reliability, practicality and economy, the main
considerations in buying cars149. Advertising grew and by 1910 domestic automakers were
spending over $1m a year on promotion.
After autos had become within the means of the middle class the Ford Motor Company
extolled their low price, modern engineering and the pleasure they could bring150. The opportunity
to develop new markets led auto manufacturers to look beyond the rural dweller to urbanites too.
But they were still marketed in the realms of a leisure pursuit rather than as a means of
commuting in the early part of the 20th century and the advertising reflected this. Foster notes,
“the public was familiar, for example, with advertising photographs of automobiles being used to
power stationery farm machinery. While the automobile industry consciously and aggressively
created urban markets, it made no particular effort at least during the 1920s, to promote the car
148 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
149 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
150 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press.
p160
68
as a commuting tool for urban workers”151, although into this they developed as driving flourished
and transit use waned. The more uses one can find for a vehicle, the more one can sell.
The desirability of autos, rather than just the fulfilment of mobility need, was encouraged
through advertising that linked cars to demonstrations of wealth and success. After the First
World War, when the car-makers realized that customers wanted and could afford a touch of
elegance in their automobiles, the new cars began to show some style, and advertising copy
began to stress beauty over the more basic mechanical features152
. This also led to more car
purchases as people wanted the latest trends which were communicated through mass
advertising and a developing consumer culture.
Advertising also brought the car into the public consciousness as part of the formula for
the American Dream, closely linked to freedom. The motor car speeded up the process of
suburbanization with homes outside cities with an automobile being the formula for what it
meant to have achieved success in life. Fundamentally automobiles were an attractive
proposition; living with privacy and space but with the ability to travel when and where you
choose was compelling. From the 1950s and 1960s automobiles were both a symbol of freedom
and a form of self-expression. They were also a very evident practical enablement of freedom: car
drivers had already been freed from the inconvenience of public transportation, able to go where
they want, when they wanted. They also enabled access to more jobs, for women to run errands,
to shuttle children, and young people to date away from the porch.
Solutionism was key to the automobile’s success in permeating the country as the chief
means of mobility. As with other transportation technologies the promoters presented their
product as a solution to whatever were the problems of the day, providing a ‘silver bullet’ to solve
151 Foster M.S. (1981) From Streetcar to Superhighway : American City Planners and Urban Transportation, 1900-
1940. Temple University Press. p60
152 Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle Books, San
Francisco, CA
69
them. These silver bullets have tended to be around safety, environmental externalities and
congestion.
The fast growth of the auto industry and the opportunities it brought for other industries
created a powerful ecosystem. To enable expansion, and resist government constraint, car
companies looked to the problems governments faced and crafted their persuasive messages
around them, key to this was deconcentrating the congested cities. These messages were then
echoed by the myriad of supporting businesses and organizations that benefited from the growth
of the auto industry through lobbying for low regulation, protected international trade and low
cost impact on drivers.
It’s perhaps easy to see the negative environmental impacts of cars when we compare a
vision of ten lane freeways slicing through cities with the pastoral vision of horse clip-clopping
down Main Street, but in the very early part of the twentieth century cars appeared to some as
environmentally better than the horse. In New York in 1900, for example, horses produced 2.5m
pounds of manure and 60,000 gallons of urine on the streets each day. Traffic was often clogged
because the bodies of dray horses died where they fell through overwork, summer heat or
destroyed after stumbling on slippery surfaces and breaking their legs153. Horses posed dangers to
humans too; in 1909, for example, there were 3,850 human deaths in horse-related accidents in
the US154. As Flink observed, the motor car provided an attractive alternative. These pretty
hideous realities of city life were naturally seized upon by auto boosters.
Initially the motorcar ‘was considered cleaner safer, more reliable, and more economical
than the horse’155 and early advocates argued they were safer too because horses were seen as
153 Flink, J. J. (1990) The Automobile Age. MIT Press p34
154 McShane, C. (1997). The Automobile: A Chronology of Its Antecedents, Development, and Impact (1st edition.,
Vol. 6). Routledge. https://doi.org/10.4324/9781315161907 p41
155 Flink, J.J. (1990) The Automobile Age. MIT Press p35
70
"willful and wild," while automobiles did not "shy or run away”156. They also had a rapidity of
movement that could be a matter of life and death; the car-driving country doctor could literally
save lives by getting to an ailing patient sooner. Evidence was produced that claimed they were
safer than trolleybuses too157
.
There was no escaping the mounting casualties and fatalities on the roads though – and
those of children who’d been used to having streets as their playground were especially
distressing. Foster notes how the rising road deaths were justified in these years: “defenders
pointed out in 1929 that the number of automobiles had multiplied 100 fold in the past 25 years,
yet traffic fatalities had increased but tenfold” and that, “while deploring needless deaths,
defenders insisted that the automobile was safer than ever before and deftly turned fatality
statistics into justification for much higher expenditure for safer, more modern highways”158. The
use of data to justify a sickening death rate appears now as cold and irresponsible.
The early car industry didn’t focus on safety; there was something of an iron law of
marketing, that safety did not sell159. Lee Iacocca, a top Ford executive in the mid-1960s,
expressed the industry's long-held view: "Styling cars sells cars and safety does not."160. Indeed
auto companies chose not to see crashes as a vehicle problem but as a driver behavior problem.
Car companies resisted calls to make their vehicles safer and instead blamed the roads and the
drivers – or pedestrians who got in the way. It took until the 1960s for car manufacturers to take
156 Eastman, J. W. (1984). Styling vs. Safety. The American Automobile Industry and the Development of
Automotive Safety, 1900-1966:115
157 American Association for Highway Improvement (1912). Papers, addresses and resolutions before the American
Road Congress. Richmond, Virginia, November 20-23, 1911. Williams & Wilkins Co., Baltimore p122
https://archive.org/details/papersaddresses00unkngoog/page/n141/mode/2up
158 Foster, M. S. From Streetcar to Superhighway : American City Planners and Urban Transportation, 1900-1940.
Temple University Press; 1981. p60
159 Congressional Record, Senate July 26, 1965 18207 Vol. 111, Part 13 89th Congress - 1st Session
https://www.congress.gov/89/crecb/1965/07/26/GPO-CRECB-1965-pt13-6-2.pdf
160 Cited in O'Connell, J. and Myers, A. (1966) Safety Last: An Indictment of the Auto Industry (New York: Random
House), p. 5
71
some responsibility. This was reflected in a congressional debate in which one speaker noted,
“While still emphasizing that it is driver error-not vehicle error-which is to blame for the vast
majority of accidents, the car manufacturers obviously have decided that it is in their best
interests to start building more safety into their cars, regardless of where the responsibility for
accidents may rest161”.
As crash victims ran into the tens of thousands the response from the auto industry was to
colonize road space and preclude the use of it by others. Behavior change was portrayed by the
car industry as the solution and so safety education – especially for children – was initiated. To
deal with recalcitrant adults, jaywalkers were publicly ridiculed or arrested162. The privileged were
those able to access a vehicle and those not privileged were those excluded from streets without
consideration to their rights. Norton notes that in 1918 city planners pointed to the lack of
dedicated play areas as “one of the chief causes of loss of life on public streets”163; the solution
was not to consider rights to the road space but to clear children off the streets.
An important appeal of the car was its use as a tool to relieve the unsanitary,
overcrowded and congested cities164. As Flink noted, “the automobile seemed, to proponents of
the innovation, to afford a simple solution to some of the more formidable social problems of
American life associated with the emergence of an urban industrial society”
165. Congestion had
built up as a consequence of earlier technologies that had significantly increased mobility in the
nineteenth century. Up until then, suburbs had been socially and economically inferior to the core
of cities but between 1815 and 1879, as steam ferry, omnibus, railroad and cable cars emerged,
161 Congressional Record, Senate July 26, 1965 18207 Vol. 111, Part 13 89th Congress - 1st Session
https://www.congress.gov/89/crecb/1965/07/26/GPO-CRECB-1965-pt13-6-2.pdf
162 Norton, P. D. (2008). Fighting traffic : the dawn of the motor age in the American city (1st ed.). MIT Press. p38
163 Norton, P. D. (2008). Fighting traffic : the dawn of the motor age in the American city (1st ed.). MIT Press. p85
164 Foster M.S. (1981) From Streetcar to Superhighway : American City Planners and Urban Transportation, 1900-
1940. Temple University Press.
165 Flink, J. J. (1970). America adopts the automobile, 1895-1910. MIT Press. p112
72
American cities ‘turned inside out’ and a reverse pattern emerged - of suburban affluence and city
despair166. But the cost of travel to the center was high and so this suburbanization was largely a
movement of wealthy people from the city core to the suburbs. As accessibility to the city core
increased through the increase in streetcar lines, the number of jobs in city centers increased too
as did the demand for cheap accommodation for workers and more crowding. As the wealthy
moved out, their homes started to be replaced by commercial buildings or high density housing for
a growing urban poor.
As the number of streetcars increased they provided a relief valve for the crowding as it
enabled a reasonably priced means of commuting compared to rail. The central business districts
benefited during the period, thriving as they welcomed people from the city and those living
further out who gained access to the center using the streetcars. The introduction of automobiles
– especially from the later 1910s – was seen as a continuation of the streetcar as a tool to
deconcentrate the cities.
City planners broadly considered the automobile to be a solution to the growing problem
of overcrowding – a way of ‘deconcentrating’ city centers. Peter Norton endorses Mark Foster’s
finding that, ‘the majority of planners enthusiastically endorsed both automobility and the
suburban movement out of conviction’167. Planners saw cars as the basis of a new and better city,
says Norton, and both Mark Foster and Clay McShane note that urban planners were enthusiastic
promoters of cars. Importantly, the car was seen by planners as a complement to the streetcar
rather than a competitor168; there wasn’t a sense of replacing them but rather becoming a positive
addition to the transportation system.
166 Jackson, K. T. (1987). Crabgrass frontier: the suburbanization of the United States. Oxford University Press.
167 Norton, P. D. (2011). Fighting Traffic: The Dawn of the Motor Age in the American City (1st ed., pp. ix–ix). The
MIT Press. p10
168 Foster M.S. (1981) From Streetcar to Superhighway : American City Planners and Urban Transportation, 1900-
1940. Temple University Press.
73
It was anticipated that urban traffic congestion and parking problems would disappear
because automobiles were more flexible than streetcars running on fixed rails, and they took up
only half the space of horse-drawn vehicles169’. But the number of automobiles rose rapidly –
almost tripling in the 1920s – a rate that was not anticipated.
Would it even have been possible for urban planners to prevent some of these negative
outcomes? Planners’ approach to the introduction of cars at the time, says Guerra, was either to
ignore them entirely or describe them as a silver bullet to current mobility woes170. At the
international city and regional planning conference held in New York in 1909 planners
“underestimated and misunderstood the future impacts of the automobile”171. These planners’
focus was on shifting people out of center of large cities that were severely crowded,
“Suburbanization was seen at the time not as a problem, but as a strategy for allowing people in
congested cities to escape to areas where they could enjoy higher quality housing, healthier
lifestyles, and parks and open space”172. Given these conditions, planners were probably right but
the scale and complexity of change meant they couldn’t foresee the extent of suburbanization.
The negative externalities revealed themselves over a period of decades by which time it
was too late to reign-in the automobile – even if there was the societal or political will to do so.
Cars had become something most people both wanted and needed – and political, economic,
financial and social structures and processes had become aligned to support them.
The impacts on air quality, the physical division of neighborhoods by busy streets, the
reduction in physical activity, the loss of children’s freedom of play in the city, the decline of
169 Flink, J.J. (1990) The Automobile Age. MIT Press p35
170 Guerra, E. (2016). Planning for Cars That Drive Themselves: Metropolitan Planning Organizations, Regional
Transportation Plans, and Autonomous Vehicles. Journal of Planning Education and Research, 36(2), 210–224.
171 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
172 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
74
transit with consequent loss of accessibility for those without cars or who couldn’t drive – all
these were not what consumers had been promised in advertising. Its naive however to think it
might be otherwise; the massive auto ecosystem existed to expand profits not to fix societal ills.
Naturally they sought means of selling more cars – but it wasn’t in their remit to manage
strategic vehicle movement, a responsibility that rests with the public sector. Messaging around
cars was based on an understanding of what consumers wanted – safety, comfort and
convenience; and around what regulators wanted – a driver of the economy, the provision of jobs
and a way of making cities better.
Freeway-building
Those with an interest in the proliferation of automobiles also had an interest in the
proliferation of roads. The auto ecosystem promoted both but they weren’t the only supporters
who pushed for it to be the central means of transportation.
City governments in the early 20th century had few tools at their disposal to deconcentrate
cities and, as noted, the auto brought the deconcentration they wanted. Their objective of
deconcentrating cities however was not achieved without a struggle because initially chambers of
commerce wanted to keep density at the center as it meant more customers for the core city
businesses they represented. But as traffic congestion became an increasingly pressing problem
the traffic appeared to be deterring customers from coming to the city centers. But city planners
and engineers had a solution.
City engineers were experiencing a golden age in their profession – respected, increasingly
well-paid and seen as scientific problem-solvers of the modern city’s challenges. To traffic
engineers, “deconcentration was not a way to fight traffic. It was unconditional surrender”,
75
explains Norton173. There was a solution as far as the engineers were concerned – and they could
find it through their scientific calculations, methods, devices - or street widening projects; and
largely the public and city leaders believed them.
Faster roads were seen as a way of making downtowns more accessible and stemming
the flow of jobs and businesses to suburbs174 175. It was believed that expressways could help to
shepherd people into downtown and lure businesses back from the suburbs. "Superhighways,"
claimed the head of the Chicago Regional Planning Association in January 1940, "offer the
opportunity to protect the regional value of the central business area" and "to enhance the values
of ... decadent areas and help restore them to a tax paying condition176". What this could look like
had been demonstrated in the months leading up to the Second World War, when the most
powerful motor manufacturer of the time, General Motors (GM), captured the public imagination
with a vision of future transportation technology.
Industrial and theatrical designer, Bel Geddes, designed GM’s hugely successful Futurama
exhibition at the 1939-1940 New York World’s Fair. Here, Geddes showcased a vision of 20 years
into the future when there would be high-speed automated roads. The following year, in his book
Magic Motorways, Geddes loosely described the technology required – a national system of
radio-controlled electric cars, propelled by electro-magnetic fields. A cog track which would be
set into the road and cars would be fitted with cog wheels that connected with the track and
173 Norton, P. D. (2008). Fighting traffic : the dawn of the motor age in the American city (1st ed.). MIT Press. p132
174 Brown, J. (2005). A tale of two visions: Harland Bartholomew, Robert Moses, and the development of the
American freeway. Journal of Planning History, 4 (1), 3–32.
175 Taylor, B. D. (2000). When finance leads planning: Urban planning, highway planning, and metropolitan
freeways in California. Journal of Planning, Education and Research, 20 (2), 196–214.
176 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
Press.
76
propelled the car forward177. But the star of the show at the World Fair was really the
demonstration of high- speed roadways within Geddes’ model which provided strong inspiration .
Futurama was a clear demonstration of silver bullets at work to solve the car problem – it
appeared to solve the congestion problem, make roads safe, make the driving task easy and
enjoyable, and improve the environment. But it came at a cost that wasn’t much considered - as
noted by Wired magazine: ‘The progenitor of the optimism-soaked hybrid of amusement park and
educational diorama later perfected by Walt Disney, Futurama was a 17-minute pitch by General
Motors that showed its audience a world that had solved transportation by signing over the
ground floor of city and country to the car.’178 Such impacts on livability – the bargain that had to
be struck - were of course not part of the messaging.
The experience of Futurama was more about vision than practical technological
development and with the Second World War intervening, plans for the realizing the more
futuristic elements of the vision – the automated roads - would have to wait; but the highway
system, it was felt, could not.
What worked in terms of promotion of the auto also worked for road-building – their
interests were intertwined. Promoters used familiar messages to promote freeways as a silver
bullet to the problems facing drivers, cities and governments – safety, congestion and
environmental externalities.
After the war, a significant selling point of a national freeway system was safety: the need
to move troops from one side of the country to the other was perceived to be crucial to national
security. Lobbyists also argued that highways had contributed immensely to the country’s
economic growth and politicians and state and local officials insisted they saved everyone
177 Fallon, Michael. Self-Driving Cars : The New Way Forward, Lerner Publishing Group, 2018. ProQuest Ebook
Central, https://ebookcentral.proquest.com/lib/socal/detail.action?docID=5444003. Chapter 1
178 Davies, A. (31 December 2017) The World’s Fair Future and the quest for our next utopia. Wired.
https://www.wired.com/story/worlds-fair-1939-futurama-utopia/
77
money179. Business saw the importance of reliable and fast routes to transport goods, people were
able to access more jobs and industries developed around a burgeoning ecosystem including
concrete, asphalt, and civil engineering.
The processes at play with freeway-building after the Second World War were very
different to the earlier part of the century. A key feature of the early part of the century was the
influence of city planners and more local, tailored road-building that worked with the fabric of the
city. Early freeway plans were created within cities and were often multi-modal, reflecting a
broad understanding of the inter-dependencies of transportation with growth, urban renewal, and
revitalization of downtowns.
As noted, in the period immediately after the Second World War II though, those planners
who perceived transportation planning as a task of drawing together mobility with land use in
harmony became marginalized. Because cities lacked the ability to pay for urban freeway
development, their planning staff lost control over the process180. Local planners outside the state
and federal highway bureaucracies, who tended to have more intimate knowledge of the
neighborhoods and districts through which the urban freeways would run, could only recommend
minor changes, subject to state and federal approval181. As Lovelace observed, “this resulted in
the preemption of park land … the division of neighborhoods, and the destruction of the fabric of
historic districts”182. The new breed of professionals with a narrower focus on traffic service and
179 Leavitt, H. (1970) Superhighway, Superhoax. Doubleday & Co (New York) p232.
180 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
181 Brown, J. (2005). A tale of two visions: Harland Bartholomew, Robert Moses, and the development of the
American freeway. Journal of Planning History, 4 (1), 3–32.
182 Lovelace, E. (1993). Harland Bartholomew: His contributions to American urban planning. Urbana, IL:
Department of Urban and Regional Planning, University of Illinois, Urbana-Champaign. p135
78
perceived traveler safety – a very different set of values from those expressed at the 1909
conference and by like-minded early planners183
.
Ambitious freeway proposals were popular with city officials, business groups, and
motorists, but they were very expensive and city councils didn’t have the funds to pay. As noted,
once state and federal government took the lead in road-building – by providing most of its cost -
a new corp of engineers with a very different perspective dominated. State and federal engineers
focused on maximizing traffic throughput and to a lesser extent on increasing safety. An
important pitch for freeways was the grade-separation which removed a significant point of
conflict between vehicles that caused crashes. As noted, grade separation was considered safer
as well as more efficient. The Automobile Manufacturers Association - the representative body of
the auto companies – explained to Congress, “We have a direct interest in safer, more freeflowing highways. The future of our industry depends on them.184” Safety and congestion were
important but so too were the needs of the industry – which Congress would have recognized
meant jobs – and these of course were important politically and economically too.
These were different objectives than those of the early planners and significantly affected
the freeways ultimately built in cities. The State highway engineers’ more narrow, technical view
of transportation planning was imposed indiscriminately on urban and rural areas. Maximizing
safe traffic flows and minimizing costs while adhering to standardized designs were their
priorities. Engineers’ hard metrics and scientific techniques were presented as incontrovertible;
they were seen as objective, rational and apolitical. The metrics of this science - speed and
throughput – was seen as superior, while the planner’s art of working with the grain of the
community and nature, with aesthetics, economic symbiosis and community, were not nearly as
183 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2), 161–177.
184 Leavitt, H. (1970) Superhighway, Superhoax. Doubleday & Co (New York) p151
79
important. State highway engineers and Federal Highway administrators at the Bureau of public
roads adopted a uniform, hard-nosed, technocratic attitude: they saw their role as just building
highways and other issues such as housing and the relocation of displaced families with the
responsibility of other agencies185. Importantly, there was strong political support for this
engineering vision and Logan Page (President of the American Association for Highway
Improvement) “had successfully manipulated the political system to press the engineering vision
of freeways”186 in his role as the first Director of the Office for Public Roads.
The ‘expert’ engineers operating at state level were appealing politically by the mid-1950s
because they seemed to provide a way of ending the political squabbling over roads. The Bureau
for Public Roads’ (BPR) point system for assessing how well a highway served traffic was
according to Business Week at the time, “a scientific, statistical measure of highway conditions
and needs” that enabled ranking without fuss and was adopted by nearly half the states because
it was “a way to take the road-building program out of the pork barrel”187. But the BPR engineers
were always active in politics, even though they claimed they operated above it because of their
technical training188
.
Research by Brown et al points to planners’ inability to foresee the impacts of private cars
and this contributed to the proliferation of an engineering-dominated vision of urban highways –
a vision that ignored issues of public health, social cohesion and livability189. But responding to
calls for strategic road-building, “City planners dropped dutifully into line and added to the
185 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
Press. p116
186 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2), p170.
187 Seely, B. E. (1987). Building the American highway system: engineers as policy makers. Temple University Press.
p230
188 Seely, B. E. (1987). Building the American highway system: engineers as policy makers. Temple University Press.
p231
189 Brown, J. R., Morris, E. A., & Taylor, B. D. (n.d.). Planning for Cars in Cities: Planners, Engineers, and Freeways in
the 20th Century. Journal of the American Planning Association, 75(2)
80
choruses of commercial and civic organizations”, noted Jackson190 and, “Elected officials bowed to
private pressure and the public purse was opened.”
A minority of planners were concerned however that expressways and increasing
decentralization, would make the situation worse. One outspoken critic was Theodore J. Kent,
chairman of the City and Regional Planning Department of the University of California at Berkeley
and planning director for San Francisco. He warned as early as 1949 of a costly freeway for the
Bay Area that would be hopelessly overcrowded and choked by the time it is completed191. Kent
recognized the need to focus on a region-wide public transit system, to limit the demand for more
freeways and private automobiles to a small express-highway network near the center of San
Francisco192
.
Sectoral interests pursued their own reasons to support plans for freeways and with them
the growing domination of the auto industry. They used persuasive techniques that were
sophisticated and insistent. The public relations advice on the new national freeway system,
presented by General Louis Prentiss to the American Roadbuilders Association (ARA), included a
ten-point PR plan for the road-building in American Road Builder magazine in 1957 which was
accompanied by a slogan in large font with the message, ‘a few must suffer so that many may
benefit’193. His advice to state and local highway officials included calling the editor of the local
newspaper giving engineering, traffic and economic reasons for the new highways; to call bankers
as respected local people; and ‘organize militant support’ for the highway plans by recruiting
190 Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford University Press.
191 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
press.
192 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
press.
193 Leavitt, H. (1970) Superhighway, Superhoax. Doubleday & Co (New York) p116
81
well-respected local people to the cause194. This was a good demonstration of the use of these
silver bullets as well as the highly organized and resourced ecosystem.
The ARA also produced a half-hour video, widely distributed, that framed opposition as
selfish and against the interests of the community and the nation, “small-minded if we tried to
fight it” says a lead character in the embedded storyline. The importance of Progress was
emphasized and that the highways would be designed and created with scientific detachment
and that they were an economic imperative, ‘indispensable’ to national defense and a way of
reducing crashes195
.
A belief that America could build its way out of the problems was supported by business,
politicians and numerous interests within a vast, complex ecosystem that was impossible to
dismantle. The generous federal funding, availability of manpower, and towns and cities not
wanting to be left behind created a momentum of change. The automakers used trade
associations and lobbying to promote their positions and former auto industry executives came to
hold influential positions in government and on official commissions196 197
.
What wasn't predicted at the time was the destruction of the urban core that followed the
building of the interstates; the ensuing sprawl was not foreseen either. But engineers did foresee
the consequences of fixes such as the street widening; Norton notes, ‘Engineers warned that the
construction of new traffic facilities would not in itself end congestion, since new capacity would
invite new demand. But automakers saw no threat in this’. Referring to a New York City engineers
warning that new roadways “would be filled immediately by traffic which is now repressed
194 Leavitt, H. (1970) Superhighway, Superhoax. Doubleday & Co (New York) p119-20
195 Video available at https://www.hemmings.com/stories/when-the-highway-came-to-hilldale-a-look-at-how-theinterstate-system-was-sold-to-america/ Retrieved 09 September 2024
196 Mohl, R. A. (1993). Race and space in the modern city: Interstate-95 and the black community in Miami. In A. R.
Hirsch & R. A. Mohl (Eds.), Urban policy in twentieth-century America (pp. 100–158). New Brunswick, NJ: Rutgers
University Press.
197 St. Clair, D. J. (1986). The motorization of American cities. New York: Praeger.
82
because of congestion,” Automotive Industries calmly observed that this was “an interesting
thought from a sales standpoint”198.’
TNCs
To capture a profitable business, TNCs developed from an expensive, niche service to a
cheaper option for relatively young urbanites. While assumed to have benefits for the
environment, their appeal was really convenience, ease of use and price.
Early Uber persuasion messages focused on ease and convenience of ordering and paying
for a ride and for traveling in style. An Uber news webpage from 2010 points to the appeal:
“hassle free & in style”199. Around this time Lyft was launched, originally as Zimcar, and its selling
point was more around fun and affordability. Speaking to Techcrunch in 2012, Co-founder, John
Zimmer said, ““Uber is this amazing, luxurious transportation experience,” Zimmer said. “Our
vision for Zimride is to have everyone participate”200
. Like cars a hundred years before, ridehail
companies moved to provide a more egalitarian offer.
Lyft initially differentiated itself with a more community-oriented brand, encouraging
users to sit in the front passenger seat and interact with drivers. Despite this, both Lyft and Uber
operated on similar principles: neither owned vehicles nor employed drivers, instead relying on a
network of independent contractors who used their own cars. This model allowed for rapid
expansion with relatively low overheads, helping TNCs grow beyond the US and into international
markets by the mid-2010s.
198 Norton, P (2008) Fighting Traffic. MIT Press, p156
199 Graves, R. (2010, September 13) Uber means business. Uber. https://www.uber.com/newsroom/uber-meansbusiness/ Retrieved 14 June, 2024
200 Cutler, K., (2012) Zimride’s Lyft Is Going To Give Uber Some Lower-Priced Competition Techcrunch 22 May 2012
https://techcrunch.com/2012/05/22/zimrides-lyft-is-going-to-give-uber-some-lower-priced-competition/
Retrieved June 14, 2024
83
As with most technology – including the car in the 20th century - the cost of the innovation
drops over time and often significantly201. Not long after Lyft’s launch, Uber introduced a cheaper
alternative to the luxury rides with Uber X and introduced other types of vehicles enabling cheaper
rides. A new advertising campaign used a number of slogans that emphasized choice, company
scale, and a sense of rider control: ‘Choice is a Beautiful Thing’, ‘All Around the World’, ‘Be The
Boss’, ‘Go Global’, ‘Get it Done’, ‘Make it a Night Out’. Pictures showed young urbanites, young
stylish urbanites, and young stylish urbanites carrying many bags of small, expensive-looking
purchases.
In the early days of TNCs, they were understood as a carpooling service – matching people
going the same way. There was expectation that they would enable shared rides, pooled rides
leading to less congestion and carbon emissions202. Travis CEO of Uber on launching the Uber Pool
service explained the benefits to VMT and congestion, “Two people taking a similar route are now
taking one car instead of two. And when you chain enough of these rides together, you can
imagine a perpetual trip — the driver picks up one customer, then picks up another, then drops
one of them off, then picks up another”.
For the American Planning Association they saw equity benefits based on existing shared
ride practices. Helping low income people and people of color was a notable attraction;
extrapolating data from the National Household Travel Survey, the APA said, “ridesharing users
still tend to have lower incomes, and Hispanics and African Americans carpool more than other
racial and ethnic groups” and, citing research by Liu and Painter, 2012, “Studies indicate that
ridesharing may serve an important role in enhancing mobility in low- income, immigrant, and
201 Kaplan, S., Gordon, B., El Zarwi, F., Walker, J. L., & Zilberman, D. (2019). The Future of Autonomous Vehicles:
Lessons from the Literature on Technology Adoption. Applied Economic Perspectives and Policy, 41(4), 583–597.
https://doi.org/10.1093/aepp/ppz005
202 Behroozi M. Understanding the Impact of Ridesharing Services on Traffic Congestion (2023). In: Heydari B,
Ergun O, Dyal-Chand R, Bart Y, eds. Reengineering the Sharing Economy: Design, Policy, and Regulation. Cambridge
University Press; 2023:119-145
84
nonwhite communities where travelers are more likely to be unable to afford personal
automobiles and obtain drivers’ licenses (Liu and Painter 2012203)”.
Another important part of the TNC appeal was that they were presented as a more
environmental choice - rides ‘on demand’ so you didn’t have to own a car, part of the sharing
economy. The future with Uber was imagined in the early years as one in which far fewer people
own cars but everybody would have access to them204. Uber told MSNBC ‘Morning Joe’ program,
that innovation was how you lift up whole industries, about the good Uber was doing for serving
under-served areas and the environmental benefits of pulling cars off the road205
.
An American Planning Association (APA) report in 2016 looked at evidence from pre-smart
phone carpooling and considered that a new surge in pooling could result with TNCs that would
capture the same benefits including: foregoing vehicle purchases; increased use of alternative
active travel; reduced vehicle miles traveled; increased access and mobility for carless households
and reduced fuel consumption and greenhouse gas emissions206. Planners saw the potential
benefits of Uber, while emphasizing the limited amount of research so far. But the TNC companies
were very guarded with their data, citing commercial confidentiality, so it was hard to understand
impacts.
The introduction of Lyft Line and Uber Pool – providing genuinely shared rides - gave hope
that this was how things would develop; but they didn’t. Whether by user preference, by the
availability of sharing or the lack of profitability, sharing rides did not emerge as a significant part
of the TNC offer.
203 Liu, C., and Gary Painter, G. (2012). Travel Behavior among Latino Immigrants: The Role of Ethnic
Neighborhoods and Ethnic Employment. Journal of Planning Education, 32 (1): 62–80.
204 Wohlsen, M. (2014, January 3) What will Uber Do With All That Money From Google. Wired.
https://www.wired.com/2014/01/uber-travis-kalanick/
205 Scola, N. (2015, February 12) The Obamization of Uber. Politico.
https://www.politico.com/magazine/story/2015/02/obama-uber-115166/
206 Cohen, A. and Shaheen, S. (2016) Planning for Shared Mobility. American Planning Association. PAS Report 583
85
By the late 2010s, studies began to show that TNCs were contributing to increased
congestion in some cities as many drivers were spending significant amounts of time driving
between rides, contributing to “dead miles.” The promise of reduced emissions through shared
rides had not materialized as expected, as momentum built and the convenience and affordability
of ridesharing encouraged more people to use TNCs in place of public transportation or walking.
In response to these criticisms, Uber and Lyft shifted their environmental messaging away
from sharing and onto sustainability. Both companies invested in electric vehicles and
autonomous vehicle research. Uber made bold claims about its commitment to becoming a zeroemissions platform by 2040, while Lyft announced plans to transition to 100% electric vehicles by
2030.
Naturally, as a private sector company, seeking external investment, the TNC companies’
primary focus was less on societal benefits and more on optimizing user experience and scaling
their business model. Public messaging around societal benefits however helped to fend off
regulation and to be accepted in cities. While they recognized the potential for reducing car
ownership and supporting a "mobility as a service" vision, the immediate goals centered around
growing their user base, increasing ride frequency, and expanding market share.
2014 and 2015 saw a TNC focus on safety following court action that claimed they were
too casual about vetting its drivers207. Rider safety was a weak point for Uber, following reports
from India and elsewhere of sexual assaults in Ubers208. As with environmentalism, on safety Uber
retained the messaging but moved the focus. Uber joined forces with Mothers Against Drunk
Driving (MADD), which helped to shift the focus from those relatively rare occurrences of serious
207 Scola, N. (2015, February 12) The Obamization of Uber. Politico.
https://www.politico.com/magazine/story/2015/02/obama-uber-115166/
208 Scola, N. (2015, February 12) The Obamization of Uber. Politico.
https://www.politico.com/magazine/story/2015/02/obama-uber-115166/
86
sexual and other assault to the far more common death and injury resulting from DUI (to
encourage using Uber instead).
The safety theme played out in Uber advertisements such as with the image of a child
accompanying the slogan, ‘Safest Rides on the Road’ and by humanizing its drivers - including
female drivers - who are shown as ‘regular guys’ rather than the suit and sunglass-wearing
drivers in earlier advertisements that appeared a little sketchy.
Uber also tried to persuade city governments of the economic benefits of the company. In
2016 Uber summarized the advantages they brought to cities: ‘A city with Uber has more
economic opportunities for residents, fewer drunk drivers on the streets, and better access to
transportation for those without it”. The introduction of Uber was about taking a trip more easily,
about safety, and about bringing jobs to cities – familiar messages from the introduction of other
transportation technologies.
Lyft, the biggest of the TNC companies after Uber, created promotional messaging that
conveyed a more friendly ride – more fun and hip than ordinary taxi rides and with the implication
that it contrasted with the darker problems affecting Uber’s image. Lyft’s brand was defined in
the early use of the huge pink, fuzzy mustache that adorned the front of Lyft vehicles in the early
years which, as co-founder Zimmer said is, “part of the fun; it’s part of the whimsy. It’s trying to
get people to smile. Even if a driver won’t put it on his car, he’s probably smiling about it and
talking about it”. More specifically, “The mustache is actually a symbol of trying to connect and to
have fun.209
”
Lyft’s advertising and social media posts reflect this idea of friendliness and fun which
helped to differentiate itself from Uber. Lyft made use of pop art, celebrity driven ad campaigns
(such as ‘Undercover Lyft’ whereby celebrities acted as drivers only revealing their identity at the
209 Logomyway (n.d.) The Lyft logo and its history. Logo my Way. https://blog.logomyway.com/lyft-logo/
87
end of the ride to much surprise and delight to the culturally well-informed riders). Not only were
TNCs promoted as ‘fun’ but ‘magical’ too. The CEO of Uber in 2014 told Wired magazine, “Uber
should feel magical to the customer," said Travis Kalanick in an interview, "They just push the
button and the car comes”210. (This idea of fun has echoes of the early automobile ‘pleasure car’
and the sense of fun and ‘magic’ is precisely the word an AV CEO has used in his tweets to
describe the AV experience. Like other things in life, what starts as fun and magical can eventually
become work-a-day and even tiresome).
As part of the ‘smart city’ and ‘big data’ hype in the 2010s, TNCs continued the tradition
of engineering, metrics-led approach of the road-building pioneers. Uber’s staffing has been high
on data people, “Kalanick employs a data-science team of PhDs from fields like nuclear physics,
astrophysics, and computational biology to grapple with the number of variables involved in
keeping Uber reliable. They stay busy perfecting algorithms that are dependable and flexible
enough to be ported to hundreds of cities worldwide”211
. But the data was strongly guarded and
the impacts and activities of companies hard to discern: so the full picture was not available to
consumers or governments until long after the companies had rolled-out in cities, momentum had
gathered, a great many people had invested in the technology and it was too late to make any
significant changes.
Synthesis and implications
The history of the development of transportation technologies shows that there has been a
coevolutionary force at work. There have been lots of players involved in their development and
210 Wohlsen, M. (2014, January 3) What will Uber Do With All That Money From Google. Wired.
https://www.wired.com/2014/01/uber-travis-kalanick/
211 Wohlsen, M. (2014, January 3) What will Uber Do With All That Money From Google, Wired.
https://www.wired.com/2014/01/uber-travis-kalanick/
88
an ecosystem formed and strengthened around them. Their ecosystems gained deeper complexity
as the number of people and organizations that benefited from their success expanded. Their
development acquired momentum as funding was invested and their uses coalesced around a
form and function that responded to the greatest consumer demand. The momentum and
complexity grew to a point where the technologies were no longer controllable by society or
government.
But coevolution has been bad for cities. Yes, there’s been significant increased mobility but
this has come with significant costs in terms of congestion, equity, environmental externalities,
deaths and injuries. There have been winners and losers – and the losers have generally been the
poor. There has been a big gap between what was expected or promised and the outcomes for
cities and their people.
Key actors have had incentives that don't actually include overall strategic systems
mobility - it's been more about selling cars, building or expanding road infrastructure or selling
rides. The actors involved are not philanthropic, third sector, charitable causes - for the most part
they’re private sector businesses that want to expand rapidly and where success is measured by
attracting investment and selling stuff.
It's been helpful to the promotion of transportation technologies that planners, politicians,
academics, consultants and others that surround the ecosystem alight upon reasons to lend their
support. This has included reasons such as that the technology has the potential for job creation,
equity benefits, economic growth, or reduced parking needs.
The transportation ecosystem has used a standard set of silver bullets (congestionbusting, safety and environmental improvement) and supplemented with promises that they can
also tackle other needs of the day (deconcentrating cities, defending from foreign attack or
reducing car ownership). The promotion of transportation technologies works like the promotion
of any other product - you establish who your audience is, what they want, and therefore you
89
craft messages to appeal to them, segmenting as necessary; so of course they'll play up their
contribution to safety, environmental improvement, congestion relief or whatever the government,
regulators, investors, potential customers need to hear to enable them to grow their business. This
is less about duplicity and dishonesty and more about standard marketing practices and
capitalism.
How a transportation technology appears to be in the early years is not an indicator of
how it will develop. The auto was not all about helping farmers and providing fun for reckless
young men, freeways were not all about bringing shoppers back to city centers and TNCs were not
all about getting people to transit stops. Sure, they were use cases but they weren’t the last word
in the technology evolution. These were technologies that in the early years were malleable and
they developed into the artefact that would be most appealing to the greatest market bringing
the most significant profit. They also came with significant costs to cities as the automobile
became a vehicle to take you to and from work every weekday and to anywhere else at the
weekend; highway building extended its tentacles deeper across the country with wider and wider
roads; and TNC companies focused on dense, wealthier areas where more rides could be sold.
90
Chapter 5: The Path to AVs
"Damn the driver and spare the car.212"
Ralph Nader
This chapter introduces the current AV cycle and draws out the commonalities we see
between it and the history of transportation technologies, drawing out patterns. The first section
sets out the history of AV development; the second examines the actors and promoters involved
and the similarities with the history of previous transportation technologies; the third section looks
at how AVs have been promoted, again to see the parallels from the past. The final section draws
together the findings and points to tentative conclusions.
AVs – A brief history of their development
While the motorcar was working its way into becoming a household necessity, the limits of
automotive technology were being pushed by ambitious, ingenious engineers. The early history of
autonomous vehicle technology was very much about plucky entrepreneurs tinkering and
experimenting with their inventions. The first driverless vehicle in the US was probably Houdina’s,
‘American Wonder’, which he showcased in New York City in 1925. American Wonder was radiocontrolled from the car behind which was able to steer it, use its brakes, gears, clutch and horn213
.
212 Nader, R. (1965) Unsafe at Any Speed: The Designed-In Dangers of the American Automobile. Grossman
Publishers p89
213 Time Magazine reported the experiment on Monday, Aug. 10, 1925:
‘Science: Radio Auto
‘In Manhattan, an empty touring car lounged against a Broadway curb. A man stepped on the running-board but
did not approach the controls. Pedestrians gaped to hear the chauffeurless machine start its motor, shift into gear,
lurch away from the curb into thick traffic. Down Broadway it went, looping uncertainly back and forth across the
91
The reportedly hair-raising drive in New York – and subsequent crash – was a somewhat hapless
venture which did little to inspire others. It took over two decades for any significant development
in AV technology to happen again in the US.
In 1953, Arthur Barrett, Jr – rigged up a wire in the ceiling of a warehouse and other
gadgetry to enable a vehicle to be propelled forward; and so the ‘Guide-O-Matic’ was born.
Barrett’s attempt at some form of automation captured the spirit of the 1950s: it was a period of
expansion and invention.
Four years after Barrett’s contraption, on a 400 foot strip of highway outside Lincoln,
Nebraska, a test site was created for a more ambitious experiment. Engineers built a car that
received electrical signals from circuits buried within wires in the pavement which controlled its
steering and braking and warned of obstructions. Signals along the sides of the highway were
picked up by the car’s sensors alerting it to an intersection or where it might encounter other
vehicles”214. Witnessed by dozens of highway officials from state and local government, journalists
and car manufacturers, the experiment demonstrated a growth in interest in vehicle automation -
and the attention of government.
With a similar technology, in 1958, Joseph Bidwell, head of General Motors (GM) Research
Engineering Mechanics Department, modified a Chevrolet for automation, equipping it with a pair
of coils at the front so it could detect the alternating current of a wire set in the roadway signaling
street. It missed a cowering milkwagon, blew its horn, dodged a speeding fire-engine. Motorcycle police escorted
the vagrant down Fifth Avenue, where a particularly wild lurch brought the man on the running-board to the
steering wheel, not in time, however, to avoid a crash with a car full of cinematographers.
‘The automobile was Inventor Francis P. Houdina's American Wonder, controlled by radio waves sent from a
following car. Two sets of waves were used, caught by antennae on the Wonder's tonneau, introduced to circuitbreakers operating small electric motors, which in turn operated steering wheel, clutch, brake, gears, horn.’
214 Fallon, M. (2018) Self-Driving Cars : The New Way Forward. Lerner Publishing Group. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/socal/detail.action?docID=5444003
92
to the car’s steering which would adjust accordingly215. But this rash of experiments didn’t gain
momentum for the development of the technology because other more urgent matters forced auto
makers to turn their attention to more immediate changes needed for road safety.
The high accident and death rate on US roads led to efforts to improve safety in road
travel into the 1960s and dominated thinking on vehicle development. The interest in automation
largely collapsed following the passage of new federal auto safety laws and strict safety
regulations from the mid-1960s, which took the focus of private sector research away from
futuristic engineering and onto satisfying the new safety standards.
The subsequent fuel economy regulations of the 1970s kept private sector researchers
busy into the following decade. But some research continued: a GM/RCA automated car
experiment in 1975 in Princeton, NJ received the attention it deserved – the modified version of
the vehicle, guided by buried electronic cables in the road, was a forlorn media event that
narrowly avoided a crash216
.
The increasing development of computing power - and artificial intelligence in particular -
through the late 1970s and 1980s sparked a renewed interest in the technology. By this time
federal governments had taken a leading role not least because the technology was strongly
linked to the growing defense industry. In 1985 the Defense Advanced Research Projects Agency
(Darpa) produced the first Autonomous Land Vehicle (ALV), which used on-board cameras to
send images of the battlefield to the six computers that controlled it. Because of the complexity of
processing and responding to the images the ALV was only able to travel 3 mph but the
technology soon developed.
215 Scribner, M. (2014). Self-driving Regulation. Competitive Enterprise Institute.
https://cei.org/sites/default/files/Marc%20Scribner%20-%20Self-Driving%20Regulation.pdf Retrieved 20
September 2024
216 Fallon, M. (2018) Self-Driving Cars : The New Way Forward. Lerner Publishing Group. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/socal/detail.action?docID=5444003.
93
The federal government had brought on other players - notably universities that were also
making headway. In 1986, engineers at Carnegie Mellon built Navlab 1, an autonomous vehicle
that could travel at 20mph. The role of governments during this period of technology development
was that of incentivizer, funder and convener – bringing together the private sector, auto
companies, universities and other interested organizations to advance the technology.
With the 1989 fall of the Berlin Wall, Glasnost and the thawing of east-west relations,
defense spending started to be reassessed. Some of the work continued but was reconceptualized for civilian purposes217. One such re-purposed technology was the ALV technology.
The actors involved (the government, the military, technology companies) were, to an extent,
leading the use of the technology but there was a sense of ‘technology solution looking for a
problem’. The developers looked to other use cases outside the military since eastern bloc
hostilities had ended and military funding began to be reduced. There appeared to be potential as
auto technology – and to monetize its use.
The Gulf War from 1990 - and the precision (‘smart’) bombing of Operation Desert Storm -
demonstrated though that there was value in continued investment in technology for the
battlefield. The demonstration of the success of the technology used implied that technological
strength equaled military strength - and this helped to rally public opinion in favor of the war
which could be fought without so many military casualties, even if in fact the technology used was
not as advanced as it appeared on home televisions218
.
The context was a perception of rapidly advancing technology and a broader interest and
investment in that technology. Of particular note was the growth in ownership of home
217 Flam, F. (1994, February 4). Ex-defense scientists come in from the cold. Science, vol. 263, no. 5147, pp. 623+.
Gale In Context: Biography, link.gale.com/apps/doc/A14808738/BIC?u=usocal_main&sid=bookmarkBIC&xid=592e1300. Retrieved 11 Apr. 2023.
218 CSIS (n.d.) The Gulf War. Center for Strategic and International Studies.
https://www.csis.org/programs/emeritus-chair-strategy/us-strategic-and-defense-efforts/lessons-war/gulf-war
Chapter 7 p483
94
computers; between 1990 and 1997 the proportion of households owning a computer rose from
15% to 35%; by 1990 a third of the highest earning households owned a computer219
.
ALV technology wasn’t the only military equipment that could be repurposed for
commercial development – GPS (Global Positioning System) did too. The technology, combined
with other new communication technology, was thought to have a range of potential uses
including tracking and fleet management as well as having the potential to reduce congestion by
enabling more direct routes, better wayfinding, reduced uncertainties, and - as it developed –
enabling the avoidance of high traffic areas.
During this decade, the US federal government was deeply involved with transportation
technology and heavily funding them: Intelligent Vehicle Highway Systems (IVHS), later called
Intelligent Transportation Systems (ITS), was one such project. It aimed to integrate vehicles and
highway infrastructure with emerging technologies including sensors, communications, and
automation with an overarching objective to increase the effectiveness of surface transportation
systems, to improve safety, accessibility, energy efficiency, environmental quality, congestion and
enhance economic productivity220
. ITS focused on real-time traffic management, vehicle-toinfrastructure communication, and accident prevention technologies. Such objectives reflect the
silver bullet promotion of more historic technologies discussed in the previous chapter.
With Congressional funding of $650m, another project, the Automated Highway System
(AHS) was tasked to "develop an automated highway and vehicle prototype from which future
fully automated intelligent vehicle-highway systems can be developed"221. A National Automated
219 Bureau of Labor Statistics (1999, March, 4). Computer Ownership up sharply in the 1990s. US Department of
Labor Statistics. https://www.bls.gov/opub/btn/archive/computer-ownership-up-sharply-in-the-1990s.pdf
220 TRID (1994) What is IVHS? IVHS Architecture Bulletin, Issue Number 1. IVHS America.
https://trid.trb.org/view.aspx?id=389133
221 Ferlis, R. A. (2007, July/ August) The Dream of an Automated Highway Public Roads. USDOT Federal Highway
Administration.
https://highways.dot.gov/public-roads/julaug-2007/dream-automated-highway
95
Highway System Consortium (NAHSC) was tasked with delivering on the AHS vision and was
made up of a mix of public and private sector interests including auto manufacturers,
infrastructure builders, engineering, communications, state and local transportation agencies, and
respected academic transportation centers. Such actors reflect the historic pattern of players in
the development of transportation technologies.
The work of the NAHSC culminated – and ended - in August 1997 with an AHS proof-ofconcept demonstration on I-15 in San Diego, CA, where more than twenty fully automated
vehicles operated flawlessly, "hands-off, feet-off", for four days on two protected lanes - the
culmination of the government’s programme that had begun in the 1980s. Add to 1997 demo. The
cars were equipped with sensors and automated controls driving safely at high speeds without
human input using guidance from technology embedded in the roadway. The aim was to
“eliminate traffic flow variances caused by human distractions; safely increase traffic density in
the lane because of the tighter operating tolerances possible with fully automated control; and
manage entries and exits so that AHS lanes maintain optimum speed and spacing in heavy
traffic”222. Although the work had attracted the support of a range of partners including GM and
Lockheed Martin, the technology still required significant development. The additional cost, effort
and slow progress however meant it failed to achieve momentum.
By the early 2000s, advances in computing, sensors, and artificial intelligence began to
change the approach to automated vehicles. A Congressional mandate (section 220 of the Floyd
D. Spence National Defense Authorization Act, 2001) stated that “It shall be a goal of the Armed
Forces to achieve the fielding of unmanned, remotely controlled technology such that . . . by 2015,
one-third of the operational ground combat vehicles are unmanned.” To this end, the U.S.
222 Public Roads (1997, July/August) Demo ’97: Proving AHS works. US Department of Transportation: Federal
Highway Administration. Issue No: Vol. 61 No. 1 https://highways.dot.gov/public-roads/julyaugust-1997/demo-97-
proving-ahs-works Accessed 28 November 2024
96
Department of Defense’s Defense Advanced Research Projects Agency (Darpa) initiated a
Challenge in 2004 – a driverless car competition following a 150 mile route in the Mojave Desert
with a $1 million prize for the fastest autonomous vehicle to complete the difficult course in less
than 10 hours. Fifteen cars competed on March 13, 2004 with the most successful completing
around 7 miles of the 142-mile route.
Although none of the competitors actually finished the race, it was an important step
forward as clearly engineers were making real progress – notably those from Carnegie Mellon,
one of the more successful teams that managed to build and deploy a driverless vehicle. A rush of
enthusiasm greeted the Challenge and renewed interest emerged in autonomy but this time
without the use of connectivity between road and vehicle; the focus instead moved to the ability
of the vehicle to drive independently using a combination of sensors, radar and cameras –
building from the AHS battlefield technology.
Much of the history of AV development was developed on the basis of vehicles connected
to something else to take instruction on actions to take, rather than vehicles that operated
autonomously. Houdina’s ‘American Wonder’ was connected to the vehicle behind (Vehicle to
Vehicle, V2V), Barrett’s Guide-O-Matic took instructions from wires in the ceiling (Vehicle to
Infrastructure, V2I) and Bidwell’s modified Chevrolet took instruction from wires set in the
roadway (Vehicle to Infrastructure, V2I). The culmination of the Automated Highway System – the
1997 highway demo, used a radar reflective stripes and magnets imbedded in the roadway223. The
first Darpa Challenge marked a move towards autonomous, rather than connected, systems
because the purpose was primarily for the battlefield in remote or hostile environments where
infrastructure could be expected to be unreliable, vulnerable to enemy interference or non223 FHWA. (1997 July/August). Demo ’97: Proving Ahs works. Public Roads. Issue No: Vol. 61 No. 1
https://highways.dot.gov/public-roads/julyaugust-1997/demo-97-proving-ahs-works Retrieved 28 November 2024
97
existent. Because of this framing of the technology, developers turned their attention towards
sensors and computer vision rather than connected infrastructure.
The rapid development of the various technological components raised the possibility of a
civilian consumer market for the technology. The attraction to AVs over CAVs for tech developers
was that they wouldn’t be beholden to government to fund costly investment in infrastructure
upgrades to enable connectivity. While connectivity hasn’t been the guide to AV development
since Darpa, it is however still expected to play a supportive role to AVs in the future and is being
developed as part of trials in some states.
A second Darpa Challenge in 2005 saw significant improvements in the technology. This
time, five autonomous vehicles successfully completed the 132-mile desert course. The challenge
demonstrated the viability of autonomous vehicles in off-road environments and urged on the
development of technologies key to AV development including LIDAR, GPS, and advanced sensors.
The 2007 Darpa Urban Challenge marked the first time autonomous vehicles were tested
in an environment that more closely resembled real-world driving, sparking major interest from
industry players, particularly tech companies. In 2009, Google launched its Self-Driving Car
Project, creaming-off engineers who had participated in the Darpa challenges. By 2010, Google
AVs had completed over 100,000 miles of autonomous driving on real-world roads. Uber also
started developing its autonomous vehicle division in 2014, seeing AVs as a natural evolution for
its ride-hailing business. By 2015, other companies like Ford, GM, and Audi had begun their own
AV projects. Auto companies, seeing a potential market, started work in earnest.
The federal government had taken the lead in AV development during this time, funding
the ITS program in the 1990s and the Darpa Challenges. Government had invested colossal
amounts of public funding, led and provided the financial incentives for others to climb aboard
with their programs. By 2015 significant interest had accumulated and state-level enabling
98
legislation started to develop a new industry with proving grounds springing up across the United
States and investment started pouring in.
The phase of AV development since 2010 sits within the context of a broader period of
technological change. ‘Smart city’ technology, of which AVs are an important feature, was
trumpeted as a key part of the ‘Fourth Industrial Revolution’, including by Dan Doctoroff,
Chairman and CEO of Sidewalk Labs (a subsidiary of Google) back in 2016224
- with changes
predicted to produce a massive change that would have far-reaching consequences like previous
industrial revolutions.
The killing of Elaine Hertzberg in Tempe, AZ, in 2018 by a distracted safety driver
operating a modified Uber AV in Phoenix, AZ, was an inflection point. But fairly quickly attention
moved and further AV development was back underway. A further inflection point was the crash
in October 2023 involving a Cruise robotaxi in San Francisco. The empty vehicle dragged the
victim twenty feet, leaving her under a wheel, and then the AV company involved attempted to
hide the video of the full event leading to the withdrawal of its license to operate in California.
Today, many states are hosting AV testing and an ecosystem of tech industry, engineers,
and auto companies have emerged alongside public affairs organizations such as the Autonomous
Vehicle Industry Association whose members include Waymo, Cruise, Ford, Volvo, Volkswagen,
Uber and UPS, bringing together auto companies, logistics, TNCs and tech companies – an array
of industries that benefit from road use, reflecting the ecosystems of previous transportation
technologies.
224 USDOT Volpe Center (2016, October 26) Reimagining Transportation from the Internet Up with Dan Doctoroff.
US Department of Transportation, Volpe Center.
https://www.volpe.dot.gov/news/reimagining-transportation-from-the-internet-up-with-dan-doctoroff Retrieved
September 22, 2024
99
Actor Involvement
What could be seen towards the close of the 20th century and the early part of the 21st was
a continued role for the key players that enabled the development of the car, road-building, and
TNCs in the earlier part of this century. The auto industry has been joined by the engineering
sector but also the tech sector, which today includes some of the most valuable businesses in the
world, including of course Google. It has also found its champions in academia, politics and
elsewhere.
The auto companies involved have included GM, Ford, Volvo, Tesla and Audi. Tech
companies have also been heavily involved, notably Google, but also a myriad of other companies
involved in the software. Engineering companies have been involved, including those developing
the sensors and other equipment. These companies provide the foundation of a sprawling
ecosystem.
A vital actor in the development of AVs has been the investors in the companies involved.
A massive amount of investment has been put into the development of AV technology –
investment that of course expects a return. In 2023, strategy and management consulting firm,
McKinsey & Company, estimated AVs would create $170-230bn in revenue by 2035225
.
The importance of venture capital and other investment sources is reflected in the social
media posts of Cruise, a tech company that along with Waymo has dominated the robotaxi
business in the past few years . Cruise’s tweets project a company that’s enjoying soaring success
225 Deichmann, J., Ebel, E., Heineke ,K., Heuss, R., Kellner, M. and Steiner, F. (2023, January 6) Autonomous
driving’s future: Convenient and connected. McKinsey & Company.
https://www.mckinsey.com/industries/automotive-and-assembly/our-insights/autonomous-drivings-futureconvenient-and-connected#/
100
- fast clocking up millions of driverless miles and generating massive investment. On 15 April
2021, Cruise tweeted, “Today, @Walmart joins @Microsoft @GM @Honda & institutional
investors in a previously announced funding round bringing the combined equity investment to
$2.75B. Redefining transportation requires strategic partnerships” and “The Cruise Origin is taking
its next steps. We’ve secured a multi-year $5B line of credit with @GMFinancial to purchase
thousands of Cruise Origin vehicles, and the first test vehicles have just begun production. ….
Combined with existing committed resources, this gives us access to more than $10B in capital to
scale Cruise over the next several years.”
These massive amounts of funding and big-name funders project an image of power and
success and importantly, belief in the product’s potential. Waymo, in contrast does not openly
celebrate funding matters in such detail. AV development is an expensive business and there is
intense pressure for the AV companies to appeal to investors so enable them to continue with
their work. The risks are that, in order to impress investors, AV companies ‘run before they can
walk’, putting at risk the trust in the technology.
Giving evidence to a congressional committee on AVs, Professor Philip Koopman, a highly
respected expert on AV safety at Carnegie Mellon University, noted: “It is important to remember
that there is tremendous financial and competitive pressure to deploy the technology as fast as
possible. Given a near vacuum of regulatory requirements on the safety of the computer driver
itself (other than recalls, often after significant harm has been done), companies have a huge
incentive to cut corners on safety and bet on getting lucky to meet their milestones and for that
next round of funding….. the reality is that in the scramble to keep their companies viable,
executives might prioritize deployment milestones and getting the next feature working rather
than safety. Regulations provide the guard rails and a level playing field so that American
companies as well as foreign competitors can’t cut corners on safety if they think that the
risk/reward outcome in doing so is appealing. There is nothing inherent to this technology that
101
guarantees that safety will be improved. If we do not explicitly require safety, we are likely to get
the least amount of safety the industry thinks it can deliver while prioritizing showing progress to
investors. ”226. Professor Koopman reflects the grave dangers of reliance on venture capital on safe
testing.
AV companies have given the impression that the technology is more advanced than it is
to impress investors but also to give the appearance of momentum. Public-facing material – and
surveys227 – reflect what Forlano calls an “already-here-ness” of AVs’228. It is not always made
clear whether what’s described is the practical product available or a theoretical concept: usually
it’s the latter. The ‘already-here-ness’ for example was expressed at the South by Southwest
Conference in March 2023 during which the CEO of GM, speaking with the CEO, Kyle Vogt,
encouraged people to just ‘take a ride’ because “the technology is here right now”229. In reality full
autonomy does not yet exist and some heavy-weight AV companies have folded in its pursuit,
including in 2022 ArgoAI (which was backed by Ford and Volkswagen) and Ghost (which had
partnered with OpenAI) which closed in April 2024 and Cruise itself in December 2024, which has
folded back into GM and stopped its work on robotaxis.
Killed by a software engineer
This blurring of the state of advancement of the technology is reflected in survey research
from J.D. Power in 2023 which revealed that consumers think that the technology is more
developed than it is. Fifteen percent of respondents thought fully automated vehicles were
226 US Congress (2023, July 26). Philip Koopman Testimony on Self Driving Cars
https://d1dth6e84htgma.cloudfront.net/IDC_Philip_Koopman_Ph_D_Testimony_Self_Driving_Cars_AV_Hearing_2
023_07_26_c2ebaa103f.pdf
227 Such as the J.D. Power survey of 2023
228 Forlano, L. (2019). Stabilizing/Destabilizing the Driverless City: Speculative Futures and Autonomous Vehicles.
International Journal of Communication 13(2019), 2811–2838
229 Interview available at the SXSW website https://schedule.sxsw.com/2023/events/PP1143207 Retrieved 26 July
2024
102
already available for purchase or lease and a further nine percent thought they would be within a
year. The cause could be seen as a deliberate confusion on the part of tech companies – this is
particularly true of Tesla, the CEO of which, Elon Musk, has been saying his company would be
producing fully autonomous vehicles imminently since 2015. Indeed Tesla has been claiming such
capabilities for its ‘Autopilot’ systems for years and as a result has been the subject of court
action over exaggerated claims of its capabilities of its ‘Autopilot’230. In a court case in Florida
following the death of Stephen Banner in 2019, the judge found evidence that Tesla “engaged in a
marketing strategy that painted the products as autonomous” and that Musk's public statements
about the technology “had a significant effect on the belief about the capabilities of the
products”231. Bryant Walker Smith, a University of South Carolina law professor, told Reuters news
agency that the judge's summary of the evidence suggests "alarming inconsistencies" between
what Tesla knew internally, and what it was saying in its marketing232
.
J.D. Power’s accompanying white paper, ‘Mobility Confidence Index Survey’, measured
understanding of descriptions of varying levels of automation and noted, “The use of “Autopilot”
remains troubling. It is a top selection to describe L2 automation and appears in the middle for all
three automation descriptions illustrating consumer confusion. It is therefore not surprising that
consumers incorrectly (and frequently) cite “Autopilot” and “Tesla” as examples of AVs that are
available today for purchase/lease”. The press notice reflects this: “consumers inaccurately cite
examples of personal vehicles available to purchase or lease today, as 22% indicate that “Tesla”
or “Autopilot” are fully automated.
230 For example, coverage by the (2024, April 28), Lawsuits test Tesla claim that drivers are solely responsible for
crashes. Washington Post. https://www.washingtonpost.com/technology/2024/04/28/tesla-trial-autopilot-lawsuit/
231 Jin, H. and Levine, D. (2023, November 21) Judge finds evidence that Tesla knew about Autopilot defect.
Reuters. https://www.reuters.com/legal/judge-finds-evidence-that-tesla-musk-knew-about-autopilot-defect-2023-
11-22/ . Retrieved December 18, 2024.
232 Jin, H. and Levine, D. (2023, November 21) Judge finds evidence that Tesla knew about Autopilot defect.
Reuters. https://www.reuters.com/legal/judge-finds-evidence-that-tesla-musk-knew-about-autopilot-defect-2023-
11-22/ . Retrieved December 18, 2024.
103
There are instances of the use of ‘conditional’ tense from tech companies (“the technology
could save thousands of lives”), which appears to be a deliberate blurring of the current readiness
of the technology. At times the use of the conditional tense appears to be dependent on the
acceptance by a recalcitrant public or Luddite naysayers and at other times on a loosening of
regulation or speed of scaling (or both).
On scaling, for example, this is implied in Cruise’s quote when announcing its partnership
with Mothers Against Drunk Driving (MADD), one of the nation’s leading advocacy organizations,
could be seen to imply achieving the safety required is a matter of scaling: “This tragic reality
drives the mission at Mothers Against Drunk Driving (MADD) and our team at Cruise to safely
bring autonomous vehicles to more communities with urgency. We know that our AVs – which
cannot drive under the influence, don’t get drowsy, and don’t text while driving – can and will
save lives and reduce collisions on the road”. (This actually has echoes of safety promotion of cars
over horses in the last century with the former “would not shy or run away”). In turn, MADD
National President Tess Rowland responded to Cruise, “We are grateful to Cruise for their
partnership and inspired by technology’s potential to ensure nobody needs to worry about the
potentially reckless behavior of drivers around them. Together, we can create a nation where
there are No More Victims,”. Cruise goes on to say, “Drunk driving is a factor in 31% of all traffic
crash fatalities in the United States. MADD’s vision is a world with “No more victims of
drunk/drugged driving.” Cruise knows that world is possible and it’s our joint mission to help
achieve it”. The bottom line though is that Level 5 autonomy does not exist and the technology
that is powering robotaxis (the current highest level of automation) is not yet proven.
There can be seen to be a feedback loop by which the car and tech companies talk up the
technology; media (including social media posts) echo the optimism because it doesn’t have the
resources to investigate and because it makes for interesting news or Likes; and academics tend
to talk about AVs as if the technology is ready or nearly ready (such as asking about attitudes
104
even though it’s about a product that does not yet exist) because it’s topical and more likely to be
published or it’s new and it’s easier to find an angle not already covered in the literature.
The role of the federal government in AV development has been important, especially in
the earlier years – just as it was for the development of the automobile and freeways. It has
assumed a number of roles. Congress provides federal funding and sets national transportation
objectives which are implemented by the U.S. Department of Transportation and the National
Highway and Traffic Safety Administration (NHTSA) upholds vehicle safety design standards.
Using the Federal Motor Vehicle Safety Standards it ensures that all vehicles conform to key
safety components including brakes, a steering wheel, seatbelts and mirrors – and assumes a
human is at the wheel. The fundamental role performed by NHTSA is to approve the safety of
vehicles – setting the minimum safety standards for vehicle features to prevent unreasonable risk
of injuries and fatalities. It also approves exemptions from safety standards and investigates
defects and mandates recalls of vehicles.
Since 2016, the federal government has published guidance to encourage the development
and rollout of AVs but has not set regulation. The first of such guidance was a joint publication
from NHTSA and the USDOT, the Federal Automated Vehicle Policy released in 2016 in the final
months of the Obama presidency233. This ‘Policy as guidance’234 aimed to accelerate the
development, describing the power of the technology which had, ‘the potential to uproot personal
mobility as we know it, to make it safer and even more ubiquitous than conventional automobiles
and perhaps even more efficient, self-driving cars have become the archetype of our future
transportation’.
233 USDOT and NHTSA. (2016, September) Federal automated vehicles policy: Accelerating the next revolution in
roadway safety. US Department of Transportation.
https://www.transportation.gov/sites/dot.gov/files/docs/AV%20policy%20guidance%20PDF.pdf
234 USDOT and NHTSA. (2016, September) Federal automated vehicles policy: Accelerating the next revolution in
roadway safety. US Department of Transportation.
https://www.transportation.gov/sites/dot.gov/files/docs/AV%20policy%20guidance%20PDF.pdf p6
105
The objectives of the federal guidance focused on safety followed by improved access for
people with disabilities, aging populations, for communities where car ownership is prohibitively
expensive, or those who prefer not to drive or own a car. In a nod to technological determinism,
Secretary for Transportation, Anthony R. Foxx, stated in its Introductory Message, ‘As the
Department charged with protecting the traveling public, we recognize three realities that
necessitate this guidance. First, the rise of new technology is inevitable. Second, we will achieve
more significant safety improvements by establishing an approach that translates our knowledge
and aspirations into early guidance. Third, as this area evolves, the “unknowns” of today will
become “knowns” of tomorrow.’ (An interesting fourth reality might be whether the “knowns” of
yesterday become the “knowns” of today).
A further interesting aspect in this report is the importance given to the role of the general
public who are seen as an important stakeholder, a role that has not been taken forward in any
formal and consistent way. The original ‘policy as guidance’ describes a more extensive two-way
dialogue with the public: “DOT expects and intends this Policy and its guidance to be iterative,
changing based on public comment; the experience of the agency, manufacturers, suppliers,
consumers, and others; and further technological innovation.’ However, Tom Cohen has pointed
out that public engagement is a key weakness in AV development more generally235
.
The extent of public comment on the federal government position appears to have come
from a public meeting held in November 2016. The executive summary of the meeting notes,
‘Public commenters in attendance included representatives from a broad range of organizations,
such as automotive manufacturers and suppliers, vehicle fleet operators, technology companies,
industry associations, and advocacy groups’. The general public appear not to be represented in
235 T. Cohen, J. Stilgoe, and C. Cavoli (2018) Reframing the Governance of Automotive Automation: Insights from
UK Stakeholder Workshops. Journal of Responsible Innovation 5: 3 p257–279.
https://doi.org/10.1080/23299460.2018.1495030
106
fact but instead representatives of the growing ecosystem familiar from the history of
transportation technologies.
Within the document, themes from the history of transportation technology are
immediately evident: safety, reduced congestion, environmentally improvement. The promise of
improved mobility access is a newer feature, reflecting 21st century concerns for equity but which
were evident in TNC promotion. It’s a theme that repeats in later iterations of government
guidance. In September 2017, further NHTSA voluntary guidance, Automated Driving Systems 2.0:
A Vision for Safety (ADS 2.0) is described in an accompanying government press release: ‘From
reducing crash-related deaths and injuries, to improving access to transportation, to reducing
traffic congestion and vehicle emissions, automated vehicles hold significant potential to increase
productivity and improve the quality of life for millions of people’236. Of course, one might say the
same of transit investment.
The document also sets out the recommended division between federal and state
responsibilities with NHTSA setting standards for vehicles, enforcing compliance, investigating
and acting on vehicle defects, communicating and educating the public on vehicle safety. State
responsibilities are around licensing and registering vehicles, enforcing traffic laws, conducting
safety inspections if they choose and regulating insurance and liability.
Such federal activity that appeared to have much zeal, waned from 2016. This left
something of a vacuum in AV regulation that state governments increasingly filled. The reason for
this change can be seen to have been political237. Nevertheless, in 2017, Sen. John Thune (R-S.D.),
236 USDOT and NHTSA (2017, September). Automated driving systems 2.0: a vision for safety. NHTSA.
https://www.nhtsa.gov/sites/nhtsa.gov/files/documents/13069a-ads2.0_090617_v9a_tag.pdf
237 In 2022 the San Francisco Examiner quoted Steve Shladover, a research engineer at Partners for Advanced
Transportation Technology at UC Berkeley, and an AV consultant for the DMV (8 Apr 22): ‘During the Trump
Administration, when “regulations were considered evil,” the already slow process of regulating AVs was put on
pause…. In an ideal world, the federal government would have stepped out earlier and would have taken the lead
on (regulations), but that didn’t happen,” Shladover added that the Biden Administration restarted the process of
crafting AV regulations.
107
chairman of the Senate Committee on Commerce, Science, and Transportation, and Sen. Gary
Peters (D-Mich.) introduced S. 1885, The American Vision for Safer Transportation Through
Advancement of Revolutionary Technologies (AV START) Act which laid out a federal framework
for autonomous vehicle regulation238; it was a cross-party effort.
This bill had four key consequences that would have impacted on the power distribution
between federal and state governments. It ‘(1) establishes a framework for a federal role in
ensuring the safety of AVs; (2) preempts states from adopting, maintaining, or enforcing any law,
rule, or standard regulating an AV or automated driving system regarding certain safety
evaluation report subject areas; (3) sets conditions for AV introduction into interstate commerce
for testing, evaluation, or demonstration; and (4) applies certain safety exemptions239. The bill,
though, stalled in the Senate; even though it was revised by its authors it wasn't enough to satisfy
critics who had concerns about crash liability, passenger safety, data security, privacy and
regulatory accountability240. In spite of the technology development that preceded and followed
the bill – and increasing amount of state legislation – the federal government did not progress an
AV regulatory framework.
Further federal guidance, Preparing for the Future of Transportation: Automated Vehicles
3.0 was published in 2018. It focused on three key areas: advancing multi-modal safety; reducing
policy uncertainty; and outlining a process for working with U.S. DOT241. It again stressed the
primacy of safety in the reasons for AVs and a need to inform (rather than ‘establish dialogue
238 US Senate Committee on Commerce, Science & Transportation (2017, September 28). Thune and Peters
introduce S. 1885, The AV State Act. Press release. US Senate. https://www.commerce.senate.gov/2017/9/thuneand-peters-introduce-s-1885-the-av-start-act
239 US Congress (2017) S.1885 – AV Start Act https://www.congress.gov/bill/115th-congress/senate-bill/1885
Retrieved 4 January 2024
240 Davis, J. (2017, December 5). New AV START Discussion Draft Addresses Some Stakeholder Concerns.
Eno Center for Transportation. https://enotrans.org/article/new-av-start-discussion-draft-addresses-somestakeholder-concerns/
241 USDOT (2018, October 4). Preparing for the Future of Transportation: Automated Vehicles 3.0. US Department
of Transportation.https://www.transportation.gov/av/3 Retrieved 27 November, 2023
108
with’) the public. In 2020, the federal government published, Ensuring American Leadership in
Automated Vehicle Technologies: Automated Vehicles 4.0, the title of which indicated the key
issue was being in pole position in the global technology race although it’s substance did little to
build on previous guidance.
In July 2023, the House Subcommittee on Innovation, Data, and Commerce held a
legislative hearing, ‘Self-Driving Vehicle Legislative Framework: Enhancing Safety, Improving
Lives and Mobility, and Beating China’. With the aim of promoting the need for a national
framework, the hearing covered a number of key issues: AV safe operation and deployment; the
economic importance of America’s leadership in global automotive design, manufacturing and AV
technology; the impact of AVs on the workforce; the role AVs in improving the mobility of people
with disabilities; and data privacy and cybersecurity issues. Themes of the economy, jobs, safety
are all familiar themes in transportation tech promotion.
Practical ways forward emerged during the hearing notably that more data was needed to
understand how safe AVs were that were currently being tested (as opposed to how safe AVs
theoretically could be). It was proposed that this could be done by strengthening NHTSA’s data
requirements while allowing an expansion of their exemption authority from 2,500 to 100,000 AVs
per year in order to gather that necessary data. The hearing also discussed the roles of Federal,
state, and local government in AV regulation including on liability and the need to protect workers
displaced by automation.
Evidence to the committee by Professor Philip Koopman noted the very lax national
picture of regulation; “Companies can shop for a state that they think has favorable rules, and
many states have been eager to create very favorable rules. Any company that can get a vehicle
working well enough to do road testing should find getting permission a simple matter of filling
out some paperwork and providing evidence of insurance coverage in an increasing number of
states. If they can manage to test for a while, say a year, without a major crash, they can likely
109
get permission to deploy without a safety driver. Under the preferred industry rules, things are
even more lax. There are no substantive impediments to deploying and maturing the technology.
Even California does not require conformance to any industry safety standards. Not even road
testing safety standards. Foreign companies come to the US specifically because it is so much
easier to test and operate here compared to other countries”. Professor Koopman provides here a
very succinct summary of the overall regulation of AVs which, although California is seen as more
strict, it is still lax in its regulation of AVs.
A hearing of the House Subcommittee on Highways and Transit on September 13, 2023
discussed, ‘The Future of Automated Commercial Motor Vehicles: Impacts on Society, the Supply
Chain, and U.S. Economic Leadership.’ As expected from the title, the discussion centered on US
AV leadership and the economic implications of it alongside the benefits of improved road safety.
The focus on economics is an important one and one that helped tech developers in the last
century and has been used very strongly in today’s promotion of AVs in some states.
In summary, federal action to regulate has been slow and has still not produced national
legislation. AV discussions have focused on economics and even foreign policy rather than being
solely about mobility. In this vacuum states have developed their own with differing levels of
regulation and with different experiences. John Thune, who originated the START Bill with Gary
Peters has been appointed Majority Leader in the Senate. With Trump's second presidency – and
his appointment of Elon Musk to a senior government position in efficiency – we may well see
renewed national interest in autonomous vehicle technology.
Other actors, who should be involved in the development of AVs, are urban planners. Their
involvement though has been slow, reflecting their pace in previous transportation technologies.
Although local government has been encouraged (such as through the American Planning
Association) to “prepare for AVs”, a review by Faisal et al finds this hadn’t happened by 2019.
They conclude, “urban planning as a profession is largely unprepared for AVs. Urban and
110
transport planners need to be aware, smart and proactive about the potential impacts,
particularly in terms of the potential for renewed urban sprawl.242” Sprawl is foreseen as a
consequence, along with more VMT (vehicle miles travelled), more vehicles on the road and
increased sprawl243
.
The lack of action on the part of planners is not due to lack of remit. Erick Guerra has
noted, “Large MPOs [metropolitan planning organizations] have both the staff and the mandate
to consider the implications of driverless cars”; he points to federal law244 that provides their
authority – indeed their responsibility - to plan strategically 25 years ahead through the
requirement to publish a strategic comprehensive plan. Surveying the biggest 25 metropolitan
planning authorities (MPOs) in 2014, Guerra found that none had included planning for AVs in their
long-range plans – and only one mentions them. Guerra found the reasons included that the
potential impact seemed too far removed from current decision-making, that there were a
number of other potential game changers and to a lesser extent there was some skepticism and
some lack of awareness at the time of writing the long-range plans (which could be up to four
years before). The strongest influence on not including AV’s in their long range plans was
uncertainty on when the technology might be ready, on consumer response, and on regulations
that might emerge and impacts of the technology. “[P]artly as a result of these uncertainties,”
says Guerra, “planners stated that driverless cars remain far removed from the types of day-today investments and policy decisions that long-range planning supports and justifies …
242 Faisal, A., Kamruzzaman, M., Yigitcanlar, T., & Currie, G. (2019). Understanding autonomous vehicles: A
systematic literature review on capability, impact, planning and policy. Journal of Transport and Land Use, 12(1),
45–72.
243 Kaplan, S., Gordon, B., El Zarwi, F., Walker, J. L., & Zilberman, D. (2019). The Future of Autonomous Vehicles:
Lessons from the Literature on Technology Adoption. Applied Economic Perspectives and Policy, 41(4), 583–597.
https://doi.org/10.1093/aepp/ppz005
244 Metropolitan Transportation Planning (2014). US Code Title 23 Chapter 1 s134 - Metropolitan Transportation
Planning.
111
interviewees consistently stated that there is insufficient evidence to justify projections of
driverless cars influence on congestion, collisions or fatalities”.
Researching the preparation of larger US city councils (over 100,000 population) five years
later, Freemark et al, reflected the Faisal et al findings of the same year. Although just over half of
their respondents said their city prioritized technological innovation, the same share agreed that
they were waiting for federal or state level legislation before moving forward. Summarizing the
views the authors conclude, “A cohort of the leaders we surveyed believes that their efforts will be
preempted by higher levels of government, so why pursue policy now?”. This waiting for upper
tiers of governance to respond has echoes of the past. As Erick Guerra notes, ‘the planning
profession has a somewhat poor track record of preparing for new transportation technologies’.
He cites Brown, Morris and Taylor’s 2009 work that argues that planners’ inability to foresee the
impacts of private cars in the 20th century contributed to the proliferation of an engineeringdominated vision of highways245. Brown et al references the agenda of the First National
Conference on City Planning in which planners’ approach to the introduction of cars was either to
ignore them entirely or describe them as a silver bullet – a remarkable optimism around the
technology that would resolve current mobility woes246
.
The weakness of involvement of local government planners in AV development is a
concern for Guerra, who warns, “Planners may yet again fail to influence the relationship between
cities and a new transportation technology by either misunderstanding driverless cars or seeing
them as a solution for contemporary planning problems, such as road congestion or climate
245 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2)
246 Brown, J. R., Morris, E. A., & Taylor, B. D. (2009). Planning for Cars in Cities: Planners, Engineers, and Freeways
in the 20th Century. Journal of the American Planning Association, 75(2)
112
change”. McAslan et al suggested the lack of progress has been due to confusion and lack of
agreement, which makes it hard to plan247
.
Academics have largely lent support to AV development. While there has been a
significant amount of research on autonomous vehicles much of it is from an engineering or
computer science perspective. Cavoli shows there’s far more technical writing than writing about
the long-term impact on the real world; his 2017 study found the majority of the literature
comprised engineering, computer science, and mathematics with social science made up less than
6% of the total; he pointed to an urgency to better understand the social and behavioral
implications of AVs248. This engineering and computer science focus has continued249 and has
echoes of the rational, engineering-dominated approach of the historic transportation
technologies described in the previous chapter.
A lack of examination of the longer-term consequences of the technology has echoes from
the past when few saw the consequences of the auto or the interstate highway system.
Assumptions were made and repeated without independent thought on the part of the agencies
and individuals involved. Cohen and Cavoli point out that some key assumptions have been
readily repeated across disciples and by policy-makers – specifically on accident reduction and
on use of road space250. Early research found optimistic impacts of AV technology including that it
247 Devon McAslan, Max Gabriele & Thaddeus R. Miller (2021) Planning and Policy Directions for Autonomous
Vehicles in Metropolitan Planning Organizations (MPOs) in the United States, Journal of Urban Technology, 28:3-4,
175-201
248 Cavoli, C. et al., 2017. Social and behavioural questions associated with Automated Vehicles. A Literature
Review. London: Department for Transport.
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/585732/soci
al-and-behavioural-questions-associated-with-automated-vehicles-literature-review.pdf
249 McCarroll C and Cugurullo F (2022) No city on the horizon: Autonomous cars, artificial intelligence, and the
absence of urbanism. Front. Sustain. Cities 4:937933.
250 Cohen, T., & Cavoli, C. (2019). Automated vehicles: exploring possible consequences of government
(non)intervention for congestion and accessibility. Transport Reviews, 39(1), 129–151
113
would increase safety and improve people’s welfare 251. Assumptions made have been engineering
ones – based on engineering thinking rather than social or behavioral thinking. And there is one
common assumption: AVs are good for cities252
.
Academics who have looked at more behavioural impacts have published papers on
‘acceptance’ of AVs – at times as if they are more than a concept and already exist or that they
will eventually exist and operate as safely as claimed. There are similarities with the arrival of
TNCs where academics optimistically considered the potential for them and perhaps there’s a
sense that if you don’t appear optimistic then you appear somewhat ‘Luddite’.
Another cheerleader for AVs have been management consultancies, such as Deloitte and
McKinsey, that have been involved with AV research because there is a market for knowledge of
future trends. Their clients expect horizon scanning, a knowledge of opportunities and challenges
afoot and it is in their business interest to have some level of knowledge of these. To encourage
clients to buy into their services there can be a tendency to promote things as significant change –
otherwise why might you need a consultant to advise you? Writing about transportation
technology, McKinsey for example in 2019 claimed, “Radically new dynamics in the early 20th
century transformed cars and, in turn, the world. Here’s why the next great inflection point is upon
us, auguring changes no less profound.253
”
The hype is very apparent in research by Arbib & Saba for consultancy, Rethink X, which
set out a future vision in which shared AVs would dominate and only a tiny, ‘immaterial’, number
of people would buy an AV. They forecast, ‘By 2030, within 10 years of regulatory approval of fully
251 Fagnant, D.J., and K. Kockelman. 2015. Preparing a Nation for Autonomous Vehicles: Opportunities, Barriers
and Policy Recommendations. Transportation Research Part A: Policy and Practice 77: 167–81
252 McCarroll C and Cugurullo F (2022) No city on the horizon: Autonomous cars, artificial intelligence, and the
absence of urbanism. Front. Sustain. Cities 4:937933.
253 Dhawan R, Hensley R, Padhi A, and Tschiesner A (2019, February). McKinsey Quarterly. Mobility’s Second Great
Inflection Point. McKinsey. https://www.mckinsey.com/industries/automotive-and-assembly/ourinsights/mobilitys-second-great-inflection-point
114
autonomous vehicles, 95% of all U.S. passenger miles will be served by transport-as-a-service
(TaaS) providers who will own and operate fleets of autonomous electric vehicles providing
passengers with higher levels of service, faster rides and vastly increased safety at a cost up to 10
times cheaper than today’s individually owned (IO) vehicles… Vehicle users will stop owning
vehicles altogether, and will instead access them when needed. The TaaS disruption will end the
model of car ownership itself’
254. These bold assertions (made with a somewhat vague
explanation of method) are the result, they say, of, ‘the huge cost savings that all individual car
owners will have when they choose to stop owning a car and use TaaS instead’. Such predictions
have not come to pass but indicate the early hype of the technology.
Reflecting on their research into the public perception of AVs, Hewitt et al notes, ‘the
public is yet to be fully convinced about autonomy in cars. Not only were participants more
anxious about higher levels of autonomy, but they also reported lower expected performance and
lower perceived ease-of-use’
255. Behavior theory suggests familiarity with the technology would
win over consumers but consumer acceptance does not appear to be increasing in spite of the
increase in information about AVs.
In an AAA survey of 949 adults across the US in 2023, they found “attitudes toward fully
self-driving vehicles have become increasingly apprehensive. This year there was a major increase
in drivers who are afraid, rising to 68% as compared to 55% in 2022. This is a 13% jump from last
254 Arbib, J. and Seba, T. (2017, May) Rethinking Transportation 2020-2030: A Rethink X Sector Disruption Report.
Rethink X. https://www.rncan.gc.ca/sites/www.nrcan.gc.ca/files/energy/energyresources/Rethinking_Transportation_2020-2030.pdf
255 Hewitt, C., Politis, I., Amanatidis, T., & Sarkar, A. (2019). Assessing public perception of self-driving cars: the
autonomous vehicle acceptance model. Proceedings of the 24th International Conference on Intelligent User
Interfaces, p526.
115
year’s survey and the biggest increase since 2020”256. In part this may be the attention heuristic of
the media sharing stories of robotaxis doing bad things.
Investment in early adopters is one strategy for advancing public acceptance. Other survey
data from J.D. Power, fielded in July 2023, shows that of 408 consumers living in Phoenix and San
Francisco that J.D. Power surveyed, only 2% lost trust in the technology during a ride experience
while 47% gained trust. “These positive firsthand experiences”, reported J.D. Power, “help educate
other consumers, providing balance to news stories that often focus solely on the negative aspects
of AVs”257. This concern that the coverage of negative incidents involving AVs reflect an attention
heuristic – considered unfair compared to the numerous incidents involving human-driven vehicles
every day. There’s something in this of course but shared TikTok videos and photos on X (Twitter)
are part of the patchwork of evidence of how the testing is going in the absence of full disclosure
of incidents and data by the AV companies themselves.
J.D. Power’s Mobility Confidence Index Study has highlighted this issue of consumer
confidence as shown in Figure 3. They have measured how comfortable people felt about riding in
a robotaxi, sharing the road with them and other measures of comfort; they found declining trust
even before the nationally reported Cruise incident in October 2023 and withdrawal of its permit
to test 258. Their research found just 20% of all consumers nationally are comfortable with AV
technology being tested on streets and highways near them. It also found that, looking at riders,
256 AAA conducted in January 2023 and reported in March 2023. https://newsroom.aaa.com/2023/03/aaa-fear-ofself-driving-cars-on-the-rise/ 1,000 adults across the US, ‘The margin of error for the study overall is 4.3% at the
95% confidence level.’
257 J.D. Power (2023, October 19). ‘New Consumers Not Embracing Robotaxis in their Community’. J.D. Power.
https://www.jdpower.com/sites/default/files/file/2023-
10/2023141%20U.S.%20Robotaxi%20Experience%20Study_0.pdf
258 The study is based on responses from 3,000 vehicle owners, age 18 and older, who completed a 15-minute
online survey. Recruitment occurred through a leading non-probability-based panel provider. The study results
were balanced to basic census demographics (age, region, gender) to be nationally representative. The sample is
comprised of 49% male and 51% female. 80% of respondents have annual incomes less than $125,000 and 44%
stated they have a university degree or higher.
116
feeling comfortable riding in an AV has declined 5 Index points in the two years to 2023. Almost
half (47%) of Riders, who tend to be open to the technology in general, are less comfortable with
them being tested in their market despite their experience and non-riders say they don’t think a
robotaxi drives any better than a human. These issues of ‘comfort’ were interpreted as ‘trust’ by
the accompanying press release.
Left to consumer demand currently ‘the public’ would reject AVs, but they are continuing
to be promoted and the technology developed. If indeed lives would be saved this has to be a
good thing - but we don’t know that. To achieve momentum there does need to be a swell of
public support for the technology. This may come as people get more used to them and more are
available in more cities – but momentum is reliant on the cost and the performance of the
vehicles – especially on their safety.
Figure 1: Mobility Confidence Index Attributes – Level of Comfort with AVs. Source: J.D. Power
2023 U.S. Mobility Confidence Index Study
117
As far as the actors are concerned there are familiar companies involved in the
development of AVs – auto companies but tech companies as well. Like the automobile and
highway-building, an ecosystem has developed around AVs and joined with the wider auto
ecosystem. Many tech companies stand to gain from AV development and governments have had
an important role in allowing AV technology to develop with little interference. As with the historic
examples, planners have not been very vocal about the possible problems with AVs; they could
provide a long-term strategic viewpoint and work to ensure the public knows not just the safety
implications but also the likely impacts on cities so that the public can make an informed choice
about how they want their cities to develop and how far AVs should be deployed.
Promotion of AVs
Having looked at the history of AV development and the actors involved, this third section
now turns to the way in which AVs have been promoted – again to establish if there are any
patterns repeated from the transportation technologies examined in the previous chapter.
The promotion of AVs into the 2010s found robotaxis to be an appropriate use as it tapped
into the hype around a sharing economy and would be an acceptable use-case for regulators. AV
companies have a number of audiences, notably: legislative and regulatory bodies; shareholders
and investors; riders and potential riders. They will of course work to ensure consistency of
message across platforms so that central messages come through whether it’s on their website, in
media interviews, at conference keynotes or in social media but there are differences in
emphases.
118
Sharing was seen to be one of the three ‘revolutions’ of transportation (the others being
electrification and automation)259 and confidence was projected that the future of AVs would be
shared. Founder of shared mobility company, Zipcar, Robin Chase, warned of the danger of
otherwise, an AV future share or not shared was the ‘heaven or hell’ future260. In spite of the
continued increase in car ownership, claims about the future of sharing have continued. A Waymo
blog written by the joint CEOs on 26 July 2023 states “Given the tremendous momentum and
substantial commercial opportunity we’re seeing on the ride-hailing front, we’ve made the
decision to focus our efforts and investment on ride-hailing.261”. Meanwhile other tech companies
working towards autonomy – notably Tesla – are focusing on the private, unshared market. The
sharing economy though has failed to live up to the hope of the 2010s as car ownership has
continued to grow262
.
Technological Determinist Rhetoric
As we’ve seen there isn’t a good case for saying that there is a technological determinist
force in technology development; nonetheless, technological determinism has been used as a tool
to persuade people of the merits of AVs. In 2016, WSP Parsons Brinckerhoff reflected technological
determinism in the headline for the introduction to their paper, ‘Driverless and autonomous
vehicles (AVs) are coming, and they will be transformational’. Such phraseology is clear in
mainstream media and also in consultant and academic writing including the Harvard Kennedy
School of Government (‘Autonomous Vehicles Are Coming: Five Policy Actions Cities Can Take
259 Sperling D. Three Revolutions. Island Press; 2018.
260 Chase, R. (2014, April 3) Will a World of Driverless Cars Be Heaven or Hell? Bloomberg.
https://www.bloomberg.com/news/articles/2014-04-03/will-a-world-of-driverless-cars-be-heaven-or-hell
261 Mawakana, T. and Dolgov, D. (2023, July 26). Doubling down on Waymo One. Waymo.
https://waymo.com/blog/2023/07/doubling-down-on-waymo-one/ Retrieved 20 February 2024
262 Federal Reserve data https://fred.stlouisfed.org/series/CXUVEHPURCHLB0402M Accessed 23 July, 2024
119
Now to Be Ready’)263. ‘The data-driven car is coming – whether you like it or not’, heralded
Information Week264
.
Elsewhere, the National League of Cities wrote: ‘The unstoppable forces of automation
and artificial intelligence are rapidly changing the way we move through, work in and design
cities. Technological advancements are transforming the mobility environment …. [B]efore long,
autonomous vehicles will be ubiquitous on roadways’265
.
It's not just the media and consultants that have implied AV development has something
of a technological determinism force – it has been reflected in professional planning literature and
from planning scholars. There’s a sense that the technology has its own trajectory and momentum
and is developing apace. The 2018, American Planning Association publication, ‘Planning for
Autonomous Mobility’, warned urban planners, ‘Regardless of the exact timeline, AVs are coming,
and they will irrevocably change transportation systems, the built environment, and our
communities….. Today, AVs are poised to disrupt the built environment and planning practices in
ways that may be hard to imagine and will be difficult to immediately determine’266
.
In the same year, the joint author William Rigg noted elsewhere, ‘With the rise of shared
and networked vehicles, autonomous vehicles, and other transportation technologies,
technological change is outpacing urban planning and policy. Whether urban planners and
policymakers like it or not, these transformations will in turn result in profound changes to streets,
land use, and cities’. The message appears to be that AVs will be rolled out and that urban
263 Fagan, M., Comeaux, D., and Gillies, B. (2021, March). Autonomous Vehicles Are Coming: Five Policy Actions
Cities Can Take Now to Be Ready. Taubman Center for State and Local Government, Harvard Kennedy School of
Government. https://www.hks.harvard.edu/centers/taubman/publications/policy-briefs/avs-are-coming
264 Norton, P. (2021). Autonorama: The Illusory Promise of High-Tech Driving (1st ed.). Island Press. p177
265 Perkins L., Dupuis N., and Rainwater B. (2018) Autonomous Vehicle Pilots Across America – Municipal Action
Guide. National League of Cities. https://www.nlc.org/resource/autonomous-vehicle-pilots-across-america/ ,
266 Crute, J., Riggs, W., Chapin, T. and Stevens, L. (2018) Planning for autonomous mobility. American Planning
Association PAS Report 592. p.3.
120
planners and policy-makers will not have a role to play in how this happens even if they don’t like
it.
This appeal of technological determinism saps a sense of human agency and it appears to
have had an effect. In a 2019 global survey of almost twenty thousand people, Ipsos Mori
reported that, “for many people, opinions about technology are shaped by a sense of fatalism.
They feel as though technological change is inevitable and that they have little say. We see this
fatalism very clearly in public dialogue. The public tends not to feel that they have a voice or a
choice in the way new technological products and services are developed. For self-driving cars,
people are likely to be concerned about whether the technology will become the norm, making
people dependent upon it and crowding out other options. Some will feel they have less
independence and less control in a self-driving world’267
.
AV Companies and the Silver Bullet Promotion of AVs
AVs have been promoted in a number of ways – through news media, social media, and
publicity events, such as city pop-ups where an AV company might hand out early access tickets
to ride in their vehicles. Publicity also comes from more specialized sources including tech
conferences, and consultancy and academic publications.
When Google first launched their self-driving technology in 2010, they promoted it as the
quintessential silver bullet for crashes, the environment, commute time and car sharing. They
proclaimed: “more than 1.2 million lives are lost every year in road traffic accidents. We believe
our technology has the potential to cut that number, perhaps by as much as half. We’re also
confident that self-driving cars will transform car sharing, significantly reducing car usage, as well
as help create the new ‘highway trains of tomorrow’. These highway trains should cut energy
267 Castell, S. and Stilgoe, J. (2019) Driverless Futures. Ipsos Mori, London.
121
consumption while also increasing the number of people that can be transported on our major
roads. In terms of time efficiency, the U.S. Department of Transportation estimates that people
spend on average 52 minutes each working day commuting. Imagine being able to spend that
time more productively.268”
More recently, in an effort to promote the technology, Mary Barra, CEO of GM (the parent
body of Cruise), in an interview in 2021 with i3 magazine, summarized what are commonly
claimed benefits of AVs: “We believe the societal benefits and business opportunities of
autonomous vehicles will be significant, and we intend for GM to be a leader in their development
and deployment. Reshaping cities and the lives of those who live in them has tremendous societal
implications. Since we believe that all AVs will be EVs, these efforts will advance our vision of zero
crashes, zero emissions and zero congestion, and help us build a more sustainable and accessible
world. Self-driving cars are real, and they are going to help rewrite the rules of mobility269”. Here
we see the traditional silver bullets firing off: safety, congestion and the environment all under an
umbrella of the smart city, seismic change.
Cruise, acquired by General Motors in 2016, expressed a commitment to safety over many
years. Following the silver bullets promotion of the past, GM’s Self-Driving Safety Report of 2018
described how it envisions the future: ‘Zero crashes to save lives; Zero emissions to leave our
children a healthier planet; and ‘Zero congestion to give our customers back their precious
time270’. This message has been refined over the years since with less emphasis on a healthier
planet and an increasing emphasis on safety.
268 Thrun, S. (2010 October 9) What We’re Driving At. Google. https://googleblog.blogspot.com/2010/10/whatwere-driving-at.html
269 Stevens, C. (2021, January 4) GM: Leading the future of mobility. Consumer Technology Association.
https://www.cta.tech/Resources/i3-Magazine/i3-Issues/2021/January-February/GM-Leading-the-Future-ofMobility
270 GM (2018) Self-driving safety report. GM.
https://www.gm.com/content/dam/company/docs/us/en/gmcom/gmsafetyreport.pdf
122
A review of Cruise’s tweets since 2021, when their robotaxis were launched on city streets
(San Francisco was the first), reveal a number of themes that reflect their brand. The themes echo
the tweets of Kyle Vogt, since he took the role of Cruise CEO in February 2021 until his departure
from the company in November 2023. Both X accounts show a brand development that regularly
refers to sustainability (chiefly that their vehicles are electric and use renewable solar power),
occasionally to affordability and especially to safety. On January 8, 2021 Vogt referred to the, ‘full
benefits of autonomy at scale - safer, cleaner, cheaper, and more accessible transportation’ and
later that month Cruise tweeted the core sell for their AVs as, "Safe, comfortable, and affordable".
The overall promotion of AVs changed in the last few years. The earlier promotion of
benefits including fun (with its echoes of early car promotion as ‘pleasure car’ and previous TNC
promotion), futuristic, and environmental benefits has given way to a stronger focus on issues of
safety. This reflects public apprehension expressed in national surveys. Tekedra Mawakana, CoCEO of Waymo, tweeted on 15 December 2022 that, “Safety is at the heart of everything we do
@waymo and our first responders are an important part of how we safely operate.”.
The earlier references to car sharing have been dropped and in fact Waymo (and Cruise)
robotaxis don’t allow for sharing rides beyond one party and usage and congestion have also
lessened as promotional messages as it’s become more clear that it's likely AVs would mean more
of both.
In Cruise’s Safety Impact Report of 2022, they set out their vision, “we are committed to a
transportation future that is safer, cleaner, more accessible, and more equitable”. Kyle Vogt, as
CEO of Cruise, tweeted on 30 June 2023, “There were 13,384 drunk-driving crash fatalities in 2021
— an average of 37 per day. This is truly unacceptable which is why I’m excited to share that
@cruise has teamed up with @maddnational to combat impaired driving and make our roads
safer. Our AVs never get drunk or impaired, and this partnership will help make our communities
safer for all road users.”
123
Like Uber, AV companies have joined with MADD including recently collaborated with
Waymo, match-funding donations271 and with Cruise272. Uber, Waymo, Cruise and GM are all
sponsors of MADD; the association provides an alignment that provides a halo effect, a positive
air of moral mission that echoes promotional methods of previous technologies.
More recently, leading AV company Waymo echoed these attributes on their website
landing page which bears the slogan, ‘The World’s most experienced driver. Making it safe and
easy for people and things to get around – without the need for anyone in the driver’s seat’. The
benefits are described as being, ‘convenient, consistent and safe’ and ‘a sustainable way to
move’. One quote is given from ‘Eva, San Francisco’, saying, ‘The fact that Waymo provides
convenience, prioritizes safety, and is committed to sustainability makes it by far my favorite
mode of transportation’. Further down the homepage, a subheading asks, ‘Why we’re here’ and
lists, ‘1.35m deaths worldwide due to crashes every year; 36,096 road deaths in the US in 2019;
94% crashes involve human error in the US’273. Here we see the same promoted attributes and no
mention of congestion.
As with the introduction of automobiles in the last century the selling point of safety is
highlighted and the failures of human drivers decried. As with freeways – the new technology can
fix this safety problem. In an interview with Fortune magazine in March 2023 Kyle Vogt, for Cruise,
said, “As a society, we’re killing an entire football stadium of Americans yearly because of
distracted driving. We’ve turned a blind eye to that for over a century because there wasn’t a
better alternative. Now there is one.”
271 MADD web notice https://secure.madd.org/page/59232/donate/1?locale=en-US Retrieved 14 June 2024
272 MADD (2023, June, 30) MADD Partners with Cruise to Advance Shared Goal of Ending Impaired Driving For
Good. MADD. https://madd.org/press-release/madd-partners-with-cruise-to-advance-shared-goal-of-endingimpaired-driving-for-good/
273 Waymo (2023) The world’s most experienced driver. Waymo. www.waymo.com Retrieved 26 November, 2023
124
The veracity of the AV main selling point – their safety – is still not clear. Being safer than
a human driver depends on what we compare them against. Given we already have the
technology to not start an ignition without checking breath for alcohol, to alert for distraction, and
to limit the speed of vehicles should we not compare the safety to drivers who do not DUI, aren’t
distracted and don’t speed – this is a much higher bar. Indeed, if AV companies are serious about
safety – if indeed safety is their “mission”, then they should use their influence to insist on these
technologies in all cars. It’s worth bearing in mind that if GM (of which Cruise is a subsidiary) were
serious about safety then all new GM vehicles could be fitted with technology to limit speeding,
distraction and DUI.
Carnegie Mellon safety expert, Professor Philip Koopman has said, it is far too early to
make assessments on their safety because – given humans average one fatal crash per 100,000
miles the AV companies would need to do at least 300m miles to make an assessment274. But
Waymo and Cruise have used their own research and presented it as irrefutable evidence of
safety.
Silver Bullet for congestion
Optimistic claims about congestion were made by academics nearly ten years ago,
notably by Fagnant and Kockelman, who found one shared AV could replace nine conventional
cars275 leading urban planning scholars to see opportunities to put right the problems that came
with the motorcar. Urban designers, say Duarte & Ratti, “should embrace AVs as a catalyst of
274 Koopman, P (2021, October 22) Debunking AV Industry Positions on Standards and Regulations. Safe
Autonomy. https://safeautonomy.blogspot.com/2021/10/debunking-av-industry-positions-on.html
275 Fagnant, D. J., Kockelman, K. M., & Bansal, P. (2015). Operations of Shared Autonomous Vehicle Fleet for
Austin, Texas, Market. Transportation Research Record, 2563(1), 98–106. https://doi.org/10.3141/2536-12
125
urban transformation” and their introduction to cities, “represents a unique opportunity to
reimagine the way we think and design road space” and put right some of the wrongs from the
last century276
.
AV companies have talked about the technology also being a resolution for congestion. In
an interview with Fortune in March 2023, Cruise CEO Kyle Vogt said: “A lot of the congestion
problems are because of how humans drive. Think about how cars accelerate in an accordion
fashion at a green light or how people slow down during an accident on a highway and do the
rubbernecking thing. Humans are so bad at driving277.” The implication is that AVs – without the
perilous human driving so badly – will alleviate the congestion problems.
Faisal et al take aim at the “frequently claimed” benefit of AVs - that they will reduce
congestion through optimum use of road spaces using platooning technology. They note, “These
studies rarely consider the scenario that an effective platooning will only work if all AVs are
travelling from a defined origin to a defined destination in a dedicated lane ….Moreover if a nonAV enters into a platoon the efficiency of platooning will reduce …More importantly, the saved
road spaces are likely to be occupied by the induced trips expected to be generated by less
mobile people to day”. History has shown time and again that if technology or road-building
creates more road space, more people will take to their cars to fill it and we can expect more
vehicles and more congestion as the ‘cost’ of travel decreases.
276 Duarte, F., & Ratti, C. (2018). The Impact of Autonomous Vehicles on Cities: A Review. The Journal of Urban
Technology, 25(4), p14.
277 Wahba, P. (2023, March 16) The CEO of GM’s Cruise thinks Driverless Cars will Rule the Road in Five Years.
‘Humans are so bad at driving’. Fortune. https://fortune.com/2023/03/16/gm-cruise-kyle-vogt-driverless-carsautonomous-vehicles/
126
Silver bullet for the economy
Government responses to AVs to some extent has been colored by politics – notably about
AVs providing jobs, which has gained the support of politicians keen to demonstrate they are progrowth and pro-jobs.
Google, Uber and Lyft, through the Self-Driving Coalition for Safer Streets, a trade
association of which they are members, hailed the early federal guidelines for AVs saying, “State
and local governments also have complementary responsibilities and should work with the federal
government to achieve and maintain our status as world leaders in innovation”. What is a
mobility tool – and potentially a tool for greater mobility equity – has been framed as a hurdle in
a global tech race.
The message at the congressional hearing in July 2023 seems to contribute to fear around
rising Chinese power. The hearings emphasized the theme of American technology leadership and
(as reflected in the name of the legislative hearing) the fear of China, which has been used as a
way to urge investment and relaxation of regulation.
A strong feature of the promotion of AVs has been the effect on the economy both
nationally and at city-level. AV companies have used the themes from the congressional
committees to justify a loosening of regulation – indeed may have crafted the messaging through
their lobbying efforts. The then-CEO of Cruise, Kyle Vogt, emphasized this in an interview with
Fortune, “I’m deeply concerned that the U.S. is moving too slowly and that Chinese competitors
could get a leg up from how their government is not just opening the doors from a regulatory and
infrastructure standpoint but also investing heavily.” This concern may be genuine or could have
been a way of excusing slower progress on scaling on the part of Cruise.
127
Fixing cities’ political problems - including equity
In the early years of the introduction of the motor car a strong ethical appeal was the
accessibility it gave to the village doctor who would much easily be able to reach their ailing
patients and so help save lives. Boosters for road-building argued there was an equity need to
afford rural citizens the same rights as urban dwellers. Similarly, an equity argument has been
used around the use of AVs and people with disabilities or mobility needs - it’s a persuasive one.
On March 3, 2023 Cruise tweeted: “We’re advancing our commitment to making the future of
transportation more inclusive by convening the Cruise Accessibility Council, a group of leaders in
the accessibility community to inform our approach to developing products and features that
make Cruise accessible for all.” Two weeks later, Cruise combines the inclusive message with one
of economic productivity: a tweet on 17th March 2023 states, “The latest study from
@natdisability looked at the macroeconomic impact of autonomous vehicles, and the numbers
are staggering. AVs would: Bring 9M people into the workforce, including 4.4M jobs for people
with disabilities - Generate over $120B in federal revenue savings”.
On 23 March 2023, Cruise combined the inclusive message with the familiar silver bullets:
“Any one of these statistics on its own is unacceptable. All of them together create a status quo
we must forcibly reject. We believe all-electric, AV ridehailing can improve our communities by
delivering on cleaner, safer, more accessible, and more inclusive mobility.”
It wasn’t just Cruise that have highlighted the potential accessibility benefits of AVs;
Waymo had already launched the ‘Waymo Accessibility Network’, a formal collaboration with
disability advocates “whose unique perspectives”, it noted in October 2022, “will help ensure we
shape the future of transportation for everyone as we continue to scale”.
Cruise had been developing the Origin – and it had been featured on their website – a
vehicle pod that can accommodate six passengers sharing – showing a disabled rider entering the
vehicle using a ramp. Cruise was struggling with design approval by the time it ceased offering
128
rides towards the end of 2023. I gained insight from an interview with Barry Einsig, a consultant
who works globally on transportation technology, has worked for departments of transportation
and assisted in the development of CALSta strategy for AVs. He noted some of the problems with
the Origin: “There are a couple of problems: first, they haven’t been certified by NHTSA because
of their crash ratings; and second, because they are designed to carry more than one person they
are classed as mass transit so must accommodate disabled riders”. So being wheelchair
accessible wasn’t in fact an ethical principle on the part of Cruise but mandated through
legislation, through the Americans with Disabilities Act. Development of the Origin was longgrassed in July 2024 - so currently no robotaxi can accommodate a wheelchair – and in spite of
the claims around helping people with disabilities – never actually has, although some AV
companies are close to engineering wheelchair compatibility.
It’s worth adding that a large proportion of disabled people do need a driver who can help
them if they need it. This is perhaps the same for other categories of people who have said that
they feel discriminated against by TNC drivers because of their appearance or their needs. These
were vocalized at the 10 August California Utilities Commission meeting with fathers expressing
approval that their daughters were safer with an AV (safe from a potentially predatory TNC
driver) and a member of the LGBTQ+ community who spoke of drivers refusing to allow them to
ride. AVs may not be the right policy response here – the better response is surely to identify the
individuals who are acting in a discriminatory way and impose sanctions to prevent it happening
again. Otherwise the losers are the many good drivers who will help a disabled person or will wait
for a daughter to close the front door before pulling away or the LBGTQ+ rider who would benefit
from a human driver’s ‘eyes on the street’. A robotaxi can’t do this.
There are other scenarios where a robotaxi won’t benefit riders. If it’s programmed to drop
a rider at a place that has turned into a dangerous spot (or simply someone you don’t want to
talk to) you can’t ask it to keep driving. A robotaxi incident in March 2024 in Los Angeles saw a
129
Waymo pull up into the middle of an active crime scene – a rider can’t ask the ‘driver’ to keep
driving, and blind people would be particularly vulnerable. And it won’t see you at the curb and
recognize you need it to pull up closer to you because of a disability – in fact one of the
complaints about robotaxis is they double park and so don’t get close enough to the curb,
presenting dangers even for riders without disabilities.
Summary – echoes of past
There are many echoes of the past in the development of autonomous vehicles, which
should alert us to the possibilities that the outcomes might be the same.
From the start governments have been helpful as the enabler, incentivizer, funder and
convener of the technology. Once a product was envisaged, that was commercially marketable,
private sectors joined the development with enthusiasm. This happened with the cars and with
freeways too and an ecosystem developed of interests dependent on roads. TNCs were allowed by
state government and they gained much investment (indeed Uber has hardly operated at a profit)
and enough public support to enable momentum to be created for them to become embedded in
the transportation system.
Important players in the development of AVs have been the car companies, which of
course are involved in order to sell more cars. If there was a belief that autonomous vehicles
would lead to people giving up their cars and fewer cars on the road, would they be actively
furthering the development? It seems unlikely. Tech companies are involved to increase profit
either by selling their technology or by attracting investment funding or sell their company as an
asset to an auto company. Investment companies, auto companies and tech companies are some
of the most wealthy sectors of the economy with significant power and influence on governments,
130
on the economy and on jobs. For this reason they gain the support of government which responds
by creating an encouraging and more relaxed regulatory environment.
The promotion messages are framed as silver bullets that will solve problems of safety,
congestion and the environment any other problems of the day. For cars and freeways these
problems were decongesting cities, for TNCs they were around equity – a theme that has been
picked up by AV companies too.
Early thinking (before momentum was created and a complex web of beneficiaries
emerged) was optimistic: cars wouldn’t compete with streetcars, freeways would cut congestion
that’d help transit, and TNCs would help connect people to transit. For AVs we’ve had similar
optimism and use cases (like those of TNC) have included connections to transit stops but trials
haven’t become widespread because they are simply not as profitable as provision in dense urban
settings which tend to have quick access to transit anyway. This is what happened with TNCs and
there’s little reason to believe that AV trials won’t follow the same path.
The general acquiescence of urban planners and academics is common to historic
transportation technologies and to AVs. There have been warning bells (as there were is the past)
but generally the response is either muted or is optimistic that they’ll bring a better future.
Planners didn’t foresee the massive impact of the motorcar nor highway-building and generally
misjudged the impacts of TNCs, thinking they’d be shared – as many planners and academics
have assumed with AVs.
Case studies will follow to explore whether what appears to be happening at a national
level is happening on the ground.
131
Chapter 6: Case Studies
‘before one can determine what level of service, what technology, and what routes should
constitute a transportation system, one must first ask, “What kind of city do you want?”278’
Warren Quenstedt, deputy administrator of the National Capital Transportation Agency
and deputy general manager of WMATA
Case study objectives
The objective of the case studies is to consider to what extent each area follows the
pattern of previous transportation technologies and whether the national picture is reflected at
city level practice.
To ensure the study had broader lessons, two test areas were chosen that had very
different contexts both politically, socially, economically, demographically and environmentally.
They needed to be well-established test areas to provide enough detail and documentary
evidence to prove useful. Phoenix, AZ and San Francisco CA were chosen as they had a
comparatively long history of AV use, opposite state politics (one Republic and one Democrat),
different weather patterns and street layouts. One case study is on the west coast while the other
is inland in the southwest region of the US. They also differ in terms of mode use with Phoenix
being very car-centric and San Francisco having a strong transit tradition.
Some US test areas have been happy to allow the AV technology to develop with little
interference from government; Phoenix is an important example here. At the other end of the
278 Schrag, Z. M. (2006). The Great Society subway : a history of the Washington Metro (1st ed.). Johns Hopkins
University Press. p.282.
132
spectrum some states have managed AV trials more closely. California has been seen to be more
restrictive but even then has been accommodating to AVs at the state level.
Case study Methodology
The case studies draw on relevant literature, archival research, official letters from public
agencies, senior politicians, public sector planning strategies and other documents, observation of
video recorded workshops and open meetings of public agencies, media analyses, promotional
material and direct interviews.
A wide range of documents were reviewed as part of these case studies including state
visioning and framework documents for autonomous vehicles, city General Plans, planning notices,
media reports, governor statements, recordings of CPUC workshops (for San Francisco), and state
and city media releases.
The case studies include interviews with key players in government, local media, and other
organizations involved to provide a deeper understanding of the AV testing in the city. Key
organizations were identified and then individuals chosen as most closely associated with the
programs. The interview schedule was approved by the Institutional Review Board of the
University of Southern California in March 2024. The interviews took place as Zoom video calls
between October 2023 and March 2024. Although each semi-structured interview was tailored
towards the role of the interviewee, a set of common questions formed the foundation of the
interview questions and these are set out in Appendix 1.
133
Case Study of San Francisco, CA
‘For a successful technology, reality must take precedence over public relations, for nature cannot
be fooled.’
- Richard Feynman, Rogers Commission Report on the Challenger Crash279
Background
Why San Francisco was chosen
In detail, San Francisco was chosen as a case study because of its scale and
comparatively long history of AVs, its strong state regulation, its challenging street layout and
weather, its population density and profile, transportation culture, and because it is in the
epicenter of the US technology sector.
The city has a scale of AV operations that provides a greater amount of data and evidence
than smaller testing centers; NHTSA’s test tracking tool shows the area to have one of the largest
street testing site in the US280. California was also one of the earliest states to trial the technology,
first passing regulations in 2012. As of November 5, 2024, 31 testing permits with a driver were in
force in the state. As of 24 July 2024, seven permits were in place to AV companies testing without
a driver. As of 11 January 2024 three AV companies had a permit to charge fares281. A Californian
279 Fenn, J. and Raskino, M. (2008) Mastering the Hype Cycle. Harvard Business Press. p47
280
NHTSA (n.d.) Test tracking tool [Infographic]. NHTSA. Retrieved October 24, 2024 from
https://www.nhtsa.gov/automated-vehicle-test-tracking-tool
281 NHTSA. (n.d.) Autonomous Vehicle Testing Permits. NHTSA. https://www.dmv.ca.gov/portal/vehicle-industryservices/autonomous-vehicles/autonomous-vehicle-testing-permit-holders/
134
city makes for a very interesting city also because the state is considered to be subject to a higher
level of regulation.
San Francisco presents as an ambitious city to trial AVs because of its geography,
topography and street layout which combine to test the limits of the technology in a way that
flatter, more modern cities do not. The comparatively narrow street width combined with many
hills (the city has forty four) provides complex roadways and diverse terrain that makes it an
attractive challenge to AV companies282. The hills preclude a complete grid road layout, impair
visibility, necessitate lower speed and create a high number of dead-end streets. The weather
also provides a challenge – notably the morning fog but also the chill and wind. Kyle Vogt (former
CEO of AV company, Cruise) tweeted on X in July 2023, “Our theory was simple: if we can make
AVs work in a city like SF - with its fog, hills, & traffic - they’ll work just about anywhere.283
”
The city also makes for an interesting case study because of its culture of transit use: 33%
of San Franciscans take transit to work – and in 2010, 17% of all daily trips were taken by
transit284. But ridership has taken a knock since Covid; survey data indicates the proportion of San
Franciscans who are frequent users (use public transit at least once a week) decreased from 59%
in 2019 to 53% in 2023 while those who had driven dropped by 65% in 2019 to 60% in 2023285; and
282 US Congress (2020, February 11) Jeffery Tumlin statement to the House of Representatives Committee on
Energy and Commerce Subcommittee. US House of Representatives.
https://docs.house.gov/meetings/IF/IF17/20200211/110513/HHRG-116-IF17-Wstate-TumlinJ-20200211.pdf
283 Vogt, K. (@kvogt). (2023, July 27). Our theory was simple: if we can make AVs work in a city like SF - with its fog,
hills, & traffic… https://x.com/kvogt/status/1684603723535007746
284 Henderson, J. (2013). Street Fight: The Politics of Mobility in San Francisco (1st ed.). University of
Massachusetts Press. p10
285 City of San Francisco (2023) City Survey Results: Resident use of different modes of transportation. City of San
Francisco. Retrieved August 5, 2024. https://www.sf.gov/data/city-survey-muni-andtransportation?_gl=1*hfyhwi*_ga*MTUyMjkzODQ2MS4xNzIyNzM1MDc4*_ga_BT9NDE0NFC*MTcyMjg3Mj
QwOS4zLjEuMTcyMjg3NDExNi4wLjAuMA..*_ga_63SCS846YP*MTcyMjg3MjQwOS4zLjEuMTcyMjg3NDEx
Ni4wLjAuMA..#find-out-more
135
the number of people who said they use public transit daily dropped from 20% to 15% - some of
these changes can be accounted for by the increase in working from home286
.
Although the city has seen decreases in public transportation use it is still high compared
to the national figure of less than 2%287; only New York City has a higher transit use. The transit use
reflects San Francisco’s density, which has also made it attractive to AV companies because of
the challenge it presents and the future market impact. One AV company has said that AVs in San
Francisco encounter 32 times as many roadway users or objects to perceive, classify, and respond
to as they do on Arizona’s suburban roads288
.
The attractiveness of the city to AV companies was summarized in a blog marking the first
million miles by Cruise, in 2021, which explained that the density, high pedestrian, scooter and
cyclist numbers, double-parked vehicles, construction zones, road closures, and other road
hazards and “instances of reckless driving, such as aggressive cut-ins and stop sign blowthroughs” makes for an extremely challenging dense urban environment. In a blog post in 2023,
Cruise explains, “Despite these challenges, and in fact because of them, we intentionally chose to
deploy our AVs in San Francisco. Driving in one of the most difficult environments in the US has
allowed us to deliver a safer experience where the need is most pressing, and has accelerated our
progress towards deploying AVs in other places where this need exists.” 289
.
San Francisco is something of a ‘spiritual home’ of technology because of its proximity to
Silicon Valley from where many tech companies started, including Uber, Airbnb and Twitter. Many
286 City of San Francisco. (2023, April 13) City survey results https://www.sf.gov/reports/april-2023/2023-citysurvey-results Retrieved 5 August 2024
287 Henderson, J. (2013). Street Fight: The Politics of Mobility in San Francisco (1st ed.). University of Massachusetts
Press p10
288 US Congress (2020, February 11) Jeffery Tumlin statement to the House of Representatives Committee on
Energy and Commerce Subcommittee. US House of Representatives.
https://docs.house.gov/meetings/IF/IF17/20200211/110513/HHRG-116-IF17-Wstate-TumlinJ-20200211.pdf p3
289 Zhang, L. (2023, April 4). Cruise’s safety record over 1 million driverless miles. Cruise.
https://www.getcruise.com/news/blog/2023/cruises-safety-record-over-one-million-driverless-miles/
136
more flocked there to benefit from the agglomeration effect of a vibrant, entrepreneurial, tech
community. Nearly 20% of the city is employed in the tech sector that comprises nearly a third of
the private sector payroll290. A significant number of residents therefore might be expected to be
more accepting of new transportation technologies in a city that was the first to experience Uber
and among the first to experience electric scooter and bike sharing.
San Francisco’s city government has not had to ‘court’ AV companies to start up there;
indeed they never invited them to use their streets as a testing ground. This may mean San
Franciscans feel less in thrall to such businesses and may be more critical of their impact because
they lacked agency in their introduction (reactance theory291). This may be exacerbated by the
history of transportation protests in the city that has demonstrated the population is motivated to
make their voice heard on matters that concern them. Together this should produce a more open
public discussion about the technology and therefore more revealing as a case study.
While there appears to be little detailed academic research into the experience of AVs in
San Francisco, there has been a great deal of media coverage – and more recently it has been
very negative. Indeed San Francisco has become something of a poster child for the problems with
AVs; no other city in the US has experienced the problems and negative public feedback received
by San Francisco. For this reason too it makes for a useful case study and critical city for analysis.
David Zipper, Senior Fellow at the MIT Mobility Initiative, told the San Francisco Chronicle
that the city “is the canary in the coal mine for autonomous vehicles because so many companies
are in Silicon Valley and see San Francisco as a high-profile showcase292’. For this and the many
290 In September 2022 Ted Egan, San Francisco chief economist, told the San Francisco Chronicle that the share of
San Francisco jobs in tech grew from 3.6% in 2006 to 18.7% in 2021, Egan said. Looking at tech’s share of the
private-sector payroll, the trajectory is even steeper: It went from 5.4% of payroll in 2006 to 32.8% in 2021
https://www.sfchronicle.com/sf/article/sfnext-poll-tech-17445059.php
291 Brehm, S. S., & Brehm, J. W. (2013). Psychological Reactance: A Theory of Freedom and Control. Academic
Press.
292 Cano, R. (2023, January 31) Self-driving cars are causing mayhem on S.F. streets, officials say. Will their
expansion be restricted? San Francisco Chronicle. https://www.sfchronicle.com/projects/2023/self-driving-cars/
137
other reasons San Francisco is a significant host city for AVs and a very good subject for extensive
research.
City profile
The city and county of San Francisco are coterminous. Its population of around 808,000 is
contained within 47 square miles with a density of 18,629 people per square mile293, making it one
of the most dense in the United States. Its streets were designed and laid in the 19th and early
20th century and so are more narrow than those of more modern cities such as Los Angeles and
Phoenix, AZ.
Being at the head of a peninsula, San Francisco is surrounded by water on three sides, so
access and optional routes are restricted. ‘In one sense’, notes Jason Henderson, ‘San Francisco is
a seven-by-seven-mile walled city with limited access and egress points’294. These geographical
features constrain city growth and – alongside its density - increase the need for greater mobility.
The density, street layout (that is not totally tied to a grid pattern), and congestion
combine to make the city an outlier in the Bay area in road collisions. It has one of the lowest
crash fatality rates per capita but one of the highest fatality rates per miles driven; this reflects
the shorter travel distances and slower speeds in the city compared with much of the region295
.
San Francisco is a city where active travel is common. In 2019, private auto trips
represented half (48%) of all trips while walking and transit each accounted for nearly one-quarter
293 US Census Bureau (n.d.) Quick Facts. US Census Bureau.
https://www.census.gov/quickfacts/fact/table/sanfranciscocitycalifornia/PST045222
294 Henderson, J. (2013). Street Fight: The Politics of Mobility in San Francisco (1st ed.). University of Massachusetts
Press p13
295 MTA (2022, October) Fatalities from crashes. Metropolitan Transportation Committee
https://vitalsigns.mtc.ca.gov/indicators/fatalities-from-crashes
138
each (22% each)296. Census data shows that the city has seen a decline in the number of
commuters using public transportation – slipping from having the third highest in the country in
2019 to the sixth in 2022297. The BART system shows that it carried fewer than half of the
passengers it did in 2019 having recovered poorly from the pandemic298
.
The pandemic surge in working from home had a significant effect. In metro areas with
over 1 million residents, San Francisco and San Jose had the highest percentage of home-based
workers in 2021 at around 35% enabled by their strength in the information and technology
sectors, which were most easily adapted to working from home299. The high proportion of remote
working has continued after the pandemic, bringing significant pressure to the city’s transit
viability.
San Franciscans are wealthy compared to most US cities. The household income in the
wider Bay Area is the highest of all US metro area – only Washington DC comes close to its
median household income300. The high income reflects the high educational level of the city: 60% of
San Franciscans have a Bachelor’s degree or higher, compared to a national average of 34%301
.
The cost of housing in San Francisco is also amongst the highest in the US. In March 2024,
Zumper put San Francisco as the third most expensive city to rent in the US after New York and
Jersey City, NJ302. Related is the cost of living in San Francisco which was ranked by the
296 SFMTA (2019) Travel Decisions Survey 2019 Summary Report. San Francisco Municipal Transportation Agency
(SFMTA) https://www.sfmta.com/sites/default/files/reports-anddocuments/2020/01/sfmta_travel_decision_survey_2019.pdf
297 Burrows, M. and Burd, C. (2024, February) Commuting in the United States: 2022. US Census Bureau.
https://www2.census.gov/library/publications/2024/demo/acsbr-018.pdf Retrieved 19 April 2024
298 https://mtc.ca.gov/tools-resources/data-tools/monthly-transportation-statistics Retrieved 19 April 2024
299 Burrows, M., Burd, C. and McKenzie, B. (2023, April) Home-based workers and the COVID-19 pandemic. US
Census Bureau https://www.census.gov/content/dam/Census/library/publications/2023/acs/acs-52.pdf
300 MTA (2023, January) Income. Vital Signs https://vitalsigns.mtc.ca.gov/indicators/income
301 US Census (2022) San Francisco Quick Facts. US Census Bureau.
https://www.census.gov/quickfacts/fact/table/sanfranciscocitycalifornia/HCN010217
302 Zumper (2024, November 21) Zumper National rent report. Zumper. https://www.zumper.com/blog/rentalprice-data/
139
Economist Intelligence Unit in 2022 as being the eighth highest in the world303. These costs are
driven by high demand, limited supply of housing, and a strong job market in high-paying sectors.
The presence of major tech companies like Google, Facebook, and Salesforce contributes to job
growth and high incomes, further fueling the cost of living.
The apparent wealth in terms of salaries, jobs and housing disguises a city of contrasts –
of rich and poor. In 2021, 10.3% of San Franciscans lived below the national poverty level304. As can
be expected, the cost of housing has pushed many out of the city to cheaper housing further out,
with consequent impacts on commuting patterns. The number of mega-commuters – people who
travel 50 or more miles and 90 minutes or more to get to work shows San Francisco to have the
highest mean travel time and mean distance in the country305 306
.
Ted Egan, San Francisco chief economist, has noted that the tech sector has transformed
the city’s economy since 2009-10, which was slow-growing before the tech surge. Egan has noted,
“San Francisco has had much faster economic growth than the country as a whole. For many
people in San Francisco, the growth of the tech industry led to broad benefits.” He noted that the
city’s GDP grew at an average rate of 7.1% a year between 2011 to 2019 (with a pandemic drop of
0.8% in 2020) and the share of its jobs in tech grew from 3.6% in 2006 to 18.7% in 2021. Tech’s
share of the private-sector payroll shows an even steeper trajectory increasing from 5.4% in 2006
to 32.8% in 2021. The tech sector has helped to boost the city’s general fund, collecting $1.27
303 The Economist (2022) Worldwide cost of living 2022. The Economist Intelligence Unit Ltd.
https://pages.eiu.com/rs/753-RIQ-438/images/worldwide-cost-of-living-2022-
summary.pdf?mkt_tok=NzUzLVJJUS00MzgAAAGSlahWHspWaS3QVAj5Vc2GkioMVMq_8NODpCunlU9_yFVenVzID28Vu6jY3kjzIPRiR5Y9h6ZuJFbao_V_bjQp63P6Li0CGscqkKh2zWlIwz_RQ
304 Vital Signs (2023, January) Poverty. Vital Signs. https://vitalsigns.mtc.ca.gov/indicators/poverty
305 US Census Bureau. (2013, March 5).Census Bureau Reports 265,000 Workers Commute into San Francisco
County, Calif., Each Day. US Census Bureau. https://www.census.gov/newsroom/archives/2013-pr/cb13-r22.html
306Rapino, M. and Fields, A. (2012, November). Mega commuting in the US. United States Census Bureau.
https://www.census.gov/content/dam/Census/library/working-papers/2013/demo/SEHSD-WP2013-03_poster.pdf
140
billion in business taxes in 2021-22, more than triple the $354 million in 2009-10307. With tech
companies recently moving out of the city and 37% office vacancy space308, the city undoubtedly
feels pressure to satisfy tech sector needs.
Transportation assets
San Francisco has an extensive public transit system that is operated by the San Francisco
Municipal Transportation Agency (SFMTA), for the City and County of San Francisco. It is the
seventh largest public transit system in the US309. The SFMTA is responsible for the management
of all city ground transportation, including oversight of the Municipal Railway (Muni), traffic,
walking, bicycling, paratransit, parking, and taxis. It is governed by a Board of Directors,
appointed by the Mayor and confirmed by the San Francisco Board of Supervisors.
The city’s transit includes:
● The San Francisco Municipal Railway (Muni) which operates an extensive network of 69
bus routes, a fleet of trolleybuses, several light rail lines (Muni Metro) operating above and
below ground. Muni also operates the city’s iconic cable cars and provides paratransit
services for individuals with disabilities.
● The Bay Area Rapid Transit (BART) system is the city’s rapid transit system, serving 50
stations along 131 miles of track across the five counties of San Francisco, Alameda,
Contra Costa, San Mateo, and Santa Clara.
● Caltrain: Caltrain provides commuter rail service between San Francisco and cities in San
Mateo and Santa Clara counties, including Palo Alto and San Jose.
307 Said, C. (2022, September 16). San Francisco Chronicle Turns out S.F. doesn’t blame the tech bros for ruining the
city, Chronicle poll finds. San Francisco Chronicle. https://www.sfchronicle.com/sf/article/sfnext-poll-tech17445059.php
308 CBRE (2024, July 9) Quarter 2, 2024 https://www.cbre.com/insights/figures/san-francisco-office-figures-q2-
2024
309 SFMTA letter to US DoT (2022, September 21). In response to NHTSA notice and request for public comment in
Docket 2022-0067.
141
The city has pedestrian-friendly neighborhoods and has made significant efforts to promote
cycling by providing shared and dedicated bike lanes and bike-sharing services. San Francisco's
road network includes major thoroughfares like the iconic Market Street and the Golden Gate
Bridge, connecting the city to Marin County and the North Bay. The city has 1,088 miles of streets,
946 miles of which are surface streets, and 59 miles of freeways310
.
Transportation Network Companies (TNCs) Uber and Lyft provide on-demand transportation
options alongside traditional taxi services. The city’s waterfront location enables a ferry service
connecting it to destinations including Alameda, Oakland, and Vallejo. For national and global
connections, San Francisco International Airport (SFO) serves as a major hub. In 2019, 5% of trips
in the city used TNCs and less than 0.5% used taxis311
.
The San Francisco County Transportation Authority (SFCTA) is the regional congestion
management agency responsible for public investments in the city’s transportation system,
ensuring safety, monitoring and analyzing travel activity and planning for the long-term.
San Francisco has a history of transportation protests. As long ago was 1896, five thousand
cyclists rode down Folsom Street calling for better street paving for safer riding312. Protests
received national coverage in the 20th century when the city became a national focus for the revolt
against the freeways that spawned after the Federal-Aid Highway Act of 1956. San Francisco was
an early leader in the revolts which vociferously opposed the perceived ugliness and intrusive
nature of the freeways that threatened the city’s livability and character313. In 1959 the same
310 SFMTA (2015) Fact Sheet. SFMTA.
https://www.sfmta.com/sites/default/files/reports/2016/2015%20SFMTA%20Transportation%20Factsheet.pdf
311 San Francisco Municipal Transportation Agency (SFMTA) Travel Decisions Survey 2019 Summary Report. SFMTA.
https://www.sfmta.com/sites/default/files/reports-anddocuments/2020/01/sfmta_travel_decision_survey_2019.pdf
312 Connect SF (2018, March) Appendix A: The history of transportation in San Francisco. https://connectsf.org/wpcontent/uploads/ConnectSF-Vision-Report_Appendix-A_The-History-of-Transportation-in-SF.pdf
313 City of San Francisco (n.d..) General Plan Introduction. City of San Francisco.
https://generalplan.sfplanning.org/I4_Transportation.htm
142
sentiment successfully halted the development of the Embarcadero and Central Freeways that
were already under construction. With freeway plans shelved, the San Franciscan government
looked to other ways to enable mobility in what is a very constrained land mass: a mass transit
solution was perhaps the only alternative314
.
The San Franciscans’ transportation activism and city government support has left a legacy of
transit strength, in contrast to what happened across the US. Bloom, notes that “citizen resistance
to redevelopment and highways, public support for transit, and elite devotion to the center city
helped balance the scales between automobiles and mass transit. More broadly, the combination
of timing, subsidies, density, and demographics proved crucial to leaving more transit in place in
San Francisco, and thus sustaining more ridership and political support, than a typical American
city315”. Indeed, Bloom notes, “The differences in San Francisco public opinion concerning the role
of transit are striking and speak eloquently to an alternative path other cities might have
followed”316
.
In 1973, the San Francisco City Planning Commission and Board of Supervisors took things a
significant stage further by adopting the ‘Transit First Policy’, giving priority to public transit
investments. This became the founding principle of the city's transportation strategy which used
street capacity and parking policies to discourage increases in vehicle traffic and sought instead
to encourage active travel317
.
314 Bloom, N. (2023). The Great American Transit Disaster: A Century of Austerity, Auto-Centric Planning, and
White Flight. Chicago: University of Chicago Press. p275
315 Bloom, N. (2023). The Great American Transit Disaster: A Century of Austerity, Auto-Centric Planning, and
White Flight. Chicago: University of Chicago Press. P259
316 Bloom, N. (2023). The Great American Transit Disaster: A Century of Austerity, Auto-Centric Planning, and
White Flight. Chicago: University of Chicago Press. P256
317 City of San Francisco (n.d.) General Plan History of Transportation in San Francisco. City of San Francisco.
https://generalplan.sfplanning.org/I4_Transportation.htm
143
Not long after the transit first policy’s introduction the city’s population began to grow fast –
rising 29% between 1980 and 2020318. This growth alongside the constraints of an historic city on a
peninsula has put pressure on the need for more efficient mobility319
.
The city’s location - an hour’s drive from Silicon Valley – has helped to power its tech industry
into the twenty-first century and become one of the trailblazers in smart city transportation
technology320. State regulators have sided with the tech companies when it comes to disputes
with the city. Uber, for example, promoted themselves as having the potential to disrupt the taxi
industry and "break the back of the taxi medallion evil empire” and the state enabled them to
create that disruption321. The California Public Utilities Commission (CPUC) allowed drivers to
remain commercially unlicensed, as Uber wanted, outsourced the responsibility for safety
regulations to the Transportation Network Companies (TNCs), and put no limitations on fare rates
or on how many vehicles could be in operation. In creating the new rules in 2013, the CPUC
emphasized that their priority wasn’t just in regulating to ensure public safety but also to support
innovation322
.
San Francisco has been an attractive launch pad for transportation technology because of
such explicit state support. The city government had been supportive in the past too; a notable
early example was San Francisco’s 2015 bid for federal funding through the Smart City Challenge
318 City and County of San Francisco. (n.d.) San Francisco population and migration. City and county of San
Francisco. https://www.sf.gov/data/san-francisco-population-and-migration
319 SFCTA and SFMTA (2020, January 21) Initial opening comments. Before the Public Utilities Commission of the
State of California. R.12-12-011. . https://www.sfmta.com/sites/default/files/reports-anddocuments/2020/01/49._initial_opening_comments_of_the_sfmta_and_sfcta_on_alj_ruling_ordering_parties_to_
comment_on_questions_regarding_cpuc_regulation_of_autonomous_vehicles.pdf p1
320 McLaren, Duncan, and Julian Agyeman. Sharing Cities : A Case for Truly Smart and Sustainable Cities. MIT Press,
2015. ProQuest Ebook Central (p21)
321 Dubal, V. B. (2017). The Drive to Precarity: A Political History of Work, Regulation, & Labor Advocacy in San
Francisco’s Taxi & Uber Economies. Berkeley Journal of Employment and Labor Law, 38(1), 73–135.
(p124)http://www.jstor.org/stable/26356922
322 Dubal, V. B. (2017). The Drive to Precarity: A Political History of Work, Regulation, & Labor Advocacy in San
Francisco’s Taxi & Uber Economies. Berkeley Journal of Employment and Labor Law, 38(1), 73–135. (p127)
http://www.jstor.org/stable/26356922
144
which demonstrated it was keen to harness AV technology for the city. This Challenge had asked
mid-sized cities to share their ideas for creating an integrated smart transportation system that
would use data, applications, and technology to help people and goods move faster, cheaper, and
more efficiently.
San Francisco’s bid, centered around growing the number of regional commuters that used
carpooling highlighted its objectives of improving affordability, increasing mobility and relieving
congestion on roads and transit323. Led by the San Francisco Metropolitan Transportation Agency
(SFMTA), the bid was about ‘linking shared mobility with CAV [Connected & Autonomous Vehicle]
technology and transit to improve mobility, equity, access, safety and affordability for all’.
The city was announced as one of seven finalists in the Challenge and although ultimately
unsuccessful, San Francisco worked to advance some of the work on AVs described in its bid so
that the city could benefit anyway from the perceived opportunities. One project that followed
was an autonomous shuttle service on Treasure Island which was free to the public. Called the
Loop, it operated for four months in August 2023 along a fixed route and was a partnership
project of the SFMTA and the Treasure Island Mobility Management Agency. It's second phase
was abandoned however because of changes to the road configuration along the route which
required significant funding and time to remap, obtain permits and test - making the project
infeasible324
.
Today, San Francisco does not include AVs in its comprehensive plan but has a policy for
making its streets safer – but AVs are not a part of it. The city’s Vision Zero policy and projects
include: 20mph zones, speed cameras, street redesign, and mode shift: the contribution of these
323USDOT (2016) Smart City Challenge. US Department of Transportation.
https://www.transportation.gov/sites/dot.gov/files/docs/Smart%20City%20Challenge%20Lessons%20Learned.pdf
324 Treasure Island Mobility Management Agency (2024 June). The Loop Final Evaluation Report. SFMTA.
https://www.sfcta.org/sites/default/files/2024-06/Loop_Final_Evaluation_Report_2024-06-25_0.pdf
145
interventions have been measured and quantified and proven to work to improve safety in a way
that AVs have not325
.
Governance
State Legislature
The governance of transportation in California involves federal, state, regional and city
governments326. The California Legislature sets the state’s overall transportation policies, its
priorities and budget. The main state transportation department is the California Department of
Transportation (Caltrans), which owns, operates, and maintains the state highway system.
Californian legislators have been supportive of AVs but have recognized the need for
regulation and oversight. Controls are managed at state, and not city or county, level . This
approach has been supported by the AV industry who have feared a city-level NIMBY approach
with a patchwork of regulations across city or county boundaries. The testing of AVs though is not
something that the SFMTA has resisted; “We’re glad to see autonomous vehicle testing in San
Francisco and we believe that in the future AVs may be safer than human drivers, but in the
meantime we need data to demonstrate their safety performance,” Jeffrey Tumlin, Director of
Transportation for SFMTA, told governing.com in June 2023327
.
325 Vision Zero San Francisco (2024) 10-Year report prioritizing street safety. SFMTA
https://www.sfmta.com/media/40114/download?inline Retrieved August 4, 2024
326 Taylor, M. (2018, June) California’s transportation system. Legislative Analyst’s Office.
https://lao.ca.gov/reports/2018/3860/californias-transportation-060418.pdf
327 Brey, J. (2023, June 13) Why San Francisco Transit Wants a Slower Rollout of Driverless Taxis.
Governing. https://www.governing.com/next/why-san-francisco-transit-wants-a-slower-rollout-of-driverless-taxis
146
State governance framework and regulatory history
California was one of the first US states to legislate on autonomous vehicles, enacting
legislation in September 2012328. The early key regulations are Senate Bill 1298, which authorized
the DMV to adopt regulations for the operation of AVs. This was followed by, in September 2014,
the first set of regulations governing how manufacturers could test AVs with a driver on California
roads. In April 2018 the DMV began approving applications for AV testing without a driver and for
deployment.
The state has set out its guiding principles: the environment (minimizing emissions, VMT
and promoting smart growth); improve affordability and convenience and eliminate the systemic
inequalities of the transportation system; require that jobs created are of a high quality; provide
improved access for people with disabilities and older people; work in authentic collaboration
with governments including tribal governments, labor and community stakeholders; that AV
operations are integrated as part of a multimodal system to enhance livability, public health and
safety; increase the safety of all road users; and that the economic benefits are shared for all
Californian workers329
.
Although California has preemptive power over AVs, not all members of the state
legislature have been happy with the balance of power. In June 2024 a bill went to the state
senate aimed at allowing cities with a population of 250,000 or greater to enact ordinances to
protect public health, safety, and welfare as it relates to AV services within their jurisdiction330
.
The bill was so watered down in response to industry opposition that Senator Dave Cortese, who
328 Anderson, J. M., Nidhi, K., Stanley, K. D., Sorensen, P., & Samaras, C. (2014). Autonomous Vehicle Technology: A
Guide for Policymakers (1st ed.). RAND Corporation. p41
329 CalSTA (2022, August) Driving the Future: Autonomous Vehicles Strategies Framework. Vision and Guiding
Principles. State of California. https://calsta.ca.gov/-/media/calstamedia/documents/final_avsf_visionguidingprinciples-a11y.pdf
330 Senate bill (SB 915 Cortese, As Amended, May 16, 2024) that had its hearing in the Assembly Committee on
Transportation 17 June, 2024. https://atrn.assembly.ca.gov/system/files/2024-06/sb-915-cortese.pdf
147
introduced the bill, withdrew it. The supporters and opponents to the bill (as set out at the end of
its summary) were largely the car makers, AV developers and chambers of commerce; those in
opposition are largely public sector bodies and their worker representative organizations. This
snapshot shows the ecosystem of interests operating at a national level are very much operating
at the California state level.
For an AV company to operate in the state, it must register with the DMV and is then
subject to its Regulations including annual reporting on test disengagements, mileage
information, and collision reports of people or property. The DMV website stipulates that
manufacturers must: certify they meet a number of safety, insurance and vehicle registration
requirements, including verifying that vehicles are capable of operating without a driver; that they
meet federal Motor Vehicle Safety Standards or have an exemption; confirm vehicles have been
tested under controlled conditions that simulate the planned area of operation; notify local
governments of planned testing in the area; and provide information to law enforcement and
other first responders on how to interact with test vehicles.331
The California governor, Gavin Newsom, has been a champion of the technology industry
since his days as San Francisco’s mayor. He sets the agenda for transportation policy in California
and his office has substantial influence over transportation funding through the state budgeting
process; he can propose transportation budgets, and works with the state legislature to allocate
funds. The Governor also has the power over appointment of key officials who oversee
transportation agencies and commissions, such as the Secretary of Transportation and members
of transportation-related boards including the California Public Utilities Commission.
331 DMV (n.d.) DMV authorizes Cruise to test driverless vehicles in San Francisco. DMV.
https://www.dmv.ca.gov/portal/news-and-media/dmv-authorizes-cruise-to-test-driverless-vehicles-in-sanfrancisco/
148
The governor’s support for the technology was demonstrated when he refused to sign a
bill (AB 316) that would have required a human safety operator to be present any time a selfdriving truck operated on public roads in the state. The state legislature approved the Bill in
September 2023 but Governor Newsom overturned the decision saying, that sufficient controls
were already in place332. In so doing, Newsom continued with his California brand –
entrepreneurial, innovative, leading-edge, forward-looking, and progressive. Echoing this
sentiment, on a visit to a China Tesla factory in November 2023 he told reporters, "With AI in
particular aiding this advancement, I think it's just going to explode and you're going to start
seeing driverless flying cars as well"333
.
San Francisco's metropolitan planning organization (MPO) is the Metropolitan
Transportation Commission (MTC), which is responsible for coordinating transportation planning,
funding, and policy decisions for the region. As an MPO, the MTC is responsible for long-term
Regional Transportation Plans to meet the needs of the area over a 20 year period, allocating
federal and state transportation funding, and prioritizing projects and programs.
The California Public Utilities Commission (CPUC), is California’s regulatory agency
responsible for overseeing public utility services, including transportation. Its role here primarily
involves regulating passenger carriers including limousines, shuttles, and transportation network
companies (TNCs) such as Uber and Lyft.
The CPUC sets rules and standards for these transportation services to ensure safety,
reliability, and fair business practices. It issues permits and licenses, establishes fare structures,
and enforces regulation compliance. The CPUC may investigate complaints and take enforcement
actions against companies that violate transportation regulations.
332 Letter from Governor Newsom to Member of the California State Assembly (2023, September 22)
https://drive.google.com/file/d/1FvplFybpMpNzgyzE7ZpJWpGzubAa7L4w/view?usp=sharing
333 Rosenhall, L. (2023, November 13) Driverless Car Growth Causes Political Tension in California. Govtech.
https://www.govtech.com/transportation/driverless-car-growth-causes-political-tension-in-california
149
The CPUC was a late entrant to the AV policy-making process, deciding in 2018 that it too
had a role in issuing or denying permits for ‘deployment’ – that is, to carry fare-paying
passengers. In launching the AV deployment programs in November 2020 it set out four goals, “1)
protect passenger safety; 2) expand the benefits of autonomous vehicle technologies to all of
California’s communities; 3) improve transportation options for all, particularly for disadvantaged
communities and low-income communities; and, 4) reduce greenhouse gas emissions and air
pollutants, particularly in disadvantaged communities334”.
On the assessment of passenger safety and where this takes place – my first interviewee,
Julia Friedlander, (Senior Manager, Automated Driving Policy at the SFMTA), told me that there’s
a prospective and retrospective approach. The prospective approach, she said, takes place at all
levels of governance from federal to local. The prospective approach sees a framework based on
a self-assessment and affirmation model; “regulators are not under the hood of code as a matter
of course”, she said. This model includes baseline safety requirements that aim to mitigate risk,
such as insurance requirements, background checks on the safety drivers, inspections of vehicles,
and a ‘passenger safety plan’ for driverless vehicles. On the retrospective side - regulators collect
data (mainly numerical data) from those companies charging fares and the DMV leads on
collision data. The city has been pushing for the expansion of data reporting requirements.
The DMV handles vehicle registration and administers driver's license - testing human
drivers, issuing and revoking them and enforces regulations related to vehicle operation and
ownership. It also administers the AV program and issues permits to manufacturers that test and
deploy them. Their concern is with the safety of AVs to operate on the public roads in contrast to
the CPUC that focuses on the safety of fare-paying passengers.
334 CPUC (2020, November 19). CPUC launches new autonomous vehicle programs. CPUC.
https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-launches-new-autonomous-vehicle-programs
150
The roles of the tiers of government in AV development can be expressed in a very basic
way: the federal government, through the NHTSA, is responsible for the vehicle; and the state
DMV is responsible for the safety of the driver; the CPUC for allowing operators to charge for
rides. This simple division is complicated as far as AVs are concerned: in a driverless ride, who is
the driver? Although humans have to pass a test in order to drive a vehicle, a driverless AV, by
definition, has no human driver to test.
The DMV and CPUC need to work together as they both have a regulatory role. On such
collaboration, Curtis told me that "the CPUC and DMV have a really great working relationship
that is governed by an interagency agreement established in recognition of their joint goals and
regulatory jurisdictions”.
Friedlander noted the challenges though, “There is some complexity in having two
agencies – but there is effort to dovetail better and an understanding that there could be better
alignment – but even coordinating data collection is quite challenging”.
The role of the City and County of San Francisco is formally to manage the competing
interests in a geographically constrained area by regulating the use of its streets and curbs, ‘while
maximizing public safety and other city and state climate, equity, economic, accessibility, and
environmental goals’335. All power over AVs lies with the state.
The consequence of the power at state level has left the city of San Francisco with very
little influence on the trials in their city even though it has been largely supportive of AV
development in theory. Reflecting this, Aaron Peskin, President of the San Francisco Board of
Supervisors, told the San Francisco Standard, “making illegal turns or rolling through a stop sign -
and we can't even ticket the car or send a ticket to the company because state law doesn't allow
335 SFMTA (2022, September 21) Letter from the SFMTA to NHTSA. SFMTA.
https://www.sfmta.com/sites/default/files/reports-anddocuments/2023/11/2022.09.21_filing_cruise_comment_letter_cruiseoriginvehicle_1.pdf
151
it”, he said. “All these people”, he continued, “are calling me and complaining about it and they
don't believe it when I tell them that I have as much authority over autonomous vehicles as I do
over the price of tires in Brazil”336. This sense of powerlessness of local people is a familiar theme
in the history of transportation technology.
The city recognizes the value in having state-wide regulation. Julia Friedlander, said that
she recognizes that to be viable, AVs need to operate across city and county boundaries and there
needs to be uniform rules as they do so.
In my interview with the director of urban planning, Rich Hillis, he noted that the
department had not had much involvement in the testing of AVs in the city – and certainly no
direct involvement, “the state is the regulator so the city can’t do much”. The transportation
element of its general plan hasn’t been updated in some decades but it is now in the process of
updating it. “The tech is moving fast”, said Hillis, “and we need to catch up”.
The focus has been on safety, which as Hillis pointed out, the state determines but with
the city “weighing in”. These more immediate safety considerations (such as getting in the way of
emergency vehicles on city streets) has necessarily deprioritized the broader analysis of the
contribution AVs might make or the impacts they might have on the city.
Hillis was optimistic though about the investment in the city’s time, “we'll get to a point
where they can operate safely. They're not perfect now but we'll get there and when we do we'll
have to figure out how it can add to our transportation system so that they don’t add to
congestion, they don’t compete with transit, that they contribute to more equitable mobility,
decreasing reliance on private ownership of vehicles”.
336 Koehn, J. (2023, August 24) Here’s How Much Robotaxi Companies Spent to Influence California Regulators. SF
Standard https://sfstandard.com/2023/08/23/cruise-waymo-spent-millions-to-lobby-for-free-rein-on-sanfrancisco-streets/
152
The governance of AVs in San Francisco is complex with power resting at state level. The
state government has acted as an incentivizer and enabler of AVs with the city and county having
very little involvement. This reflects the historic pattern of the role of state government
particularly on freeway-building and TNCs. Again the policies and regulations are detached from
what is happening on the streets – where the direct consequences are felt.
City Experience of AVs
The area covered by AVs in San Francisco has been extensive as shown in the figures
below. In August 2023, the San Francisco Chronicle reported that Cruise operated about 300
robotaxis at night and 100 during the day while Waymo had a fleet of 250 vehicles337
.
337 Cano, R. (2023, August 8) Cruise, Waymo reveal how many driverless cars they have in S.F. and how often they
stall on city streets. San Francisco Chronicle https://www.sfchronicle.com/sf/article/cruise-waymo-driverless-carsin-s-f-18282902.php
153
Figure 2: San Francisco coverage area for Waymo. Source: Waymo338
The city government, while supportive of AVs in theory, has expressed strong concerns for
how the program is working in practice. Much of the city’s objections came from the emergency
services who considered AVs, “not ready for prime time”. Secondarily, that the AV companies
were not providing adequate data on the testing and of a general lack of transparency in how
they were operating339. The AV companies’ response has been to say that the emergency services
need educating. While the city government has protested robotaxi operations it has not been able
to directly limit their use.
338 Waymo (2024) Where you can go. Waymo. https://waymo.com/waymo-one-san-francisco/
339 Nicolson, J. (2023, August) Letter to CPUC from S.F. Fire Chief
154
Figure 3: Cruise’s expanded service area in May 2022 Source: Cruise340
By November 10, 2023, the California DMV had received 673 collision reports involving AVs
from across the state341. Skepticism in the city appeared justified due to a number of wellpublicized incidents that have soured the perception of the technology and lost the trust of some
San Franciscans. Up to May 2023, the following categories of incidents had been reported:
● driving erratically
● making planned and unplanned stops in travel lanes that block traffic and interfere with
transit service
● interfering with emergency response operations and posing grave hazards to first
responders
340 Vogt, K. (2022, May 3) At @Cruise we recently expanded our driverless AV service area to nearly 70% of SF.
It goes online tonight. x.com. https://x.com/kvogt/status/1521554237037023232/photo/1
341 California DMV (n.d.) Autonomous vehicle collision reports. California Department of Motor Vehicles.
https://www.dmv.ca.gov/portal/vehicle-industry-services/autonomous-vehicles/autonomous-vehicle-collisionreports/
155
● making intrusions into construction zones marked with cones and signs in which City
employees are working in and under city streets
● intrusions into crime scenes and scenes with downed power lines and other hazards
marked with caution tape
● crashes involving non-AVs where AV driving was a contributing factor
● causing obstructions caused by driverless AV challenges with interpreting and responding
to direction given by human traffic control officers
● minimal risk condition failures in travel lanes that trap drivers of other vehicles parked
at the curb and preventing them from leaving the curb342
The problems with AVs had become far more apparent once safety drivers were taken out
of the vehicles in 2018 (Waymo) and 2020 (Cruise); since which, concerns grew significantly. The
CPUC meeting on 10 August 2023, that was to decide whether they should be allowed to charge
fares, became a focal point for opposition.
Pointing to Cruise, the ‘city’ (that is the San Francisco Municipal Transportation Agency
(SFMTA), San Francisco County Transportation Authority (SFCTA), and the Mayor’s Office of
Disability) protested Cruise’s expansion on the grounds that it was “unreasonable in light of the
Cruise AV performance record” and the lack of incrementalism, data transparency, and adequate
reporting and monitoring.343
’
342 SFMTA (2023, May 31) San Francisco’s Comments On The Draft Resolution Approving Authorization For Cruise
LLC’s Expanded Service In Autonomous Vehicle Passenger Service Phase I Driverless Deployment Program. SFMTA.
https://www.sfmta.com/sites/default/files/reports-anddocuments/2023/06/2023.05.31filing_cruise_comments_tl-19145.pdf
343 CPUC (2023, August 10) Resolution approving authorization for cruise LLC’s expanded service in autonomous
vehicle passenger service phase i driverless deployment program. California Public Utilities Commission.
https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M516/K812/516812218.PDF
156
Since then some significant incidents have been reported:
● A few days after the 10 August CPUC meeting that allowed expansion, Cruise was ordered to
halve its fleet following another incident with an emergency vehicle.
● On October 2, 2023 a Cruise vehicle hit a pedestrian who had just been knocked into its path
by a hit and run driver. The pedestrian fell under the Cruise vehicle and, while pulling to the
curb, dragged the woman about 20 feet. The media notices from Cruise emphasized the hit
and run but in sharing the footage with regulators, Cruise was not clear and open about the
dragging of the victim344. After this apparent dishonesty emerged, the DMV rescinded Cruise’s
license to operate. Vogt’s tweets ended the following month after resigning as CEO,
“Ultimately”, he tweeted, “we develop and deploy autonomous vehicles in an effort to save
lives.”
● Waymo set on fire by crowd in China Town in February 2024345
.
Since August 2023 there had been far fewer incidents involving emergency services; a
significant reason is that Cruise made up the large share of such incidents and their fleet had been
depleted346. There are other reasons too that are more about logistics than technology
development. First, both Cruise and Waymo agreed to allow emergency respondents to take over
the vehicles rather than wait for a technician to arrive; second, the Fire Department asked that the
344 This is reported in many news outlets including Reuters https://www.reuters.com/business/autostransportation/how-gms-cruise-robotaxi-tech-failures-led-it-drag-pedestrian-20-feet-2024-01-26/ and Wired
https://www.wired.com/story/robot-car-crash-investigation-cruise-disclose-key-information/
345 Jin, H., Sandoval, M. and Roy, A. (2024, February 11) Crowd sets Waymo self-driving vehicle ablaze in San
Francisco. Reuters https://www.reuters.com/world/us/crowd-sets-waymo-self-driving-car-ablaze-san-francisco2024-02-11/
346 Cano, R. (2024, February 23) Exclusive: Driverless robotaxis are causing less mayhem on S.F. streets. City
officials explain whySan Francisco Chronicle.
https://www.sfchronicle.com/sf/article/driverless-robotaxis-incidents-decrease-18672791.php
157
robotaxis do not stop outside fire houses; third, in event of an emergency event the city agencies
alert Waymo which then creates an ‘exclusion zone’ for its vehicles to avoid the area.
Data
Reflecting the pattern of data secrecy of TNCs, AV companies have been extremely
guarded about data release in the city, citing commercial confidentiality. The AV companies have
had to start reporting on accident rates though347. Waymo successfully filed a lawsuit against the
DMV over the latter’s request for data disclosure, which Waymo claimed amounted to making
public ‘trade secrets’348. In fact, it was the DMV that recommended Waymo seek an injunction
without which the DMV felt it had to disclose to a third party under a public access to information
request349. This close working between state entities and AV companies reflect the pattern of
governmental encouragement and support to the industry.
The lack of openness was reflected in my interview with the Chief Planner for the City of
San Francisco; who said that the AV companies “should be more candid about the future”.
Gaining more clarity would enable city planners to make sure the regulations shape the
technology, how they can “add positively to the transportation system – improve equity, reduce
the reliance on private vehicles, and not add to the congestion”. On how AVs might help achieve
these objectives, he felt they were still at the beginning of a journey and that having more
information would help. The city is certainly not yet in the position to update the Comprehensive
Plan, for example, in a way that sets out how AVs might contribute to city targets including
347 Zhang, L. (2023) Human Ridehail Crash Benchmark. Cruise.
https://www.getcruise.com/news/blog/2023/human-ridehail-crash-rate-benchmark/
348 Hawkins, A. (2022, January 28) Waymo sues California DMV to keep driverless crash data under wraps. The
Verge. https://www.theverge.com/2022/1/28/22906513/waymo-lawsuit-california-dmv-crash-data-foia
349 Bellan, R. (2022, February 22) Waymo to keep robotaxi safety details secret, court rules. Techcrunch.
https://techcrunch.com/2022/02/22/waymo-to-keep-robotaxi-safety-details-secret-court-rules/
158
around safety, equity and active travel. He was upbeat about the potential though – “I think we’ll
get to a point at which they operate safer than humans”.
Persuasion
The promotion of AVs in San Francisco is communicated at a national or global level
through news media, social media, academic and consultant publications, tech conference
interviews and publicity events - or a combination of some or all of these. The AV companies
themselves have of course been key promoters of the technology but there are many other
champions in the broader tech industry and business community who benefit from the broader
goals of automotive, AI or tech development and have helped in the promotion.
Both Cruise and Waymo, the companies dominating the AV market in San Francisco, have
invested heavily in their promotion, which is reflected in the amount spent on lobbying efforts,
logged by the Californian senate. Official records of the state legislature show Waymo spent
$3.8m on lobbying efforts in the legislative session 2023 through 2024 while Cruise spent $0.35m.
The previous year Waymo had spent $0.63m and Cruise $0.68m350
.
Of course it’s not just paid-for lobbying that counts: informal meetings, discussions and
relationship-building are also important in influencing legislation. As we saw with the roadbuilding program in the 20th century, it helps if people supportive to a cause are appointed to
relevant public organizations351 It’s worth noting that one of the five voting commissioners on the
CPUC, John Reynolds, who was appointed by Governor Newsom, was Managing Counsel at Cruise
LLC from 2019. He was one of the four commissions present for the 10 August meeting that
350 California State Government (n.d.) Lobbying Activity Cruise LLC. https://calaccess.sos.ca.gov/Lobbying/Employers/Detail.aspx?id=1405741&session=2021&view=activity
351 As noted earlier, the appointment of Logan Page, as Director of the Office of Public Roads in 1905 was helpful
in setting a data-led, ‘rational approach’ to road-building
159
enabled Cruise and Waymo’s expansion. It’s also worth noting that records show that Cruise gave
$52,400 to Newsom’s reelection campaign in 2022352. Appointing people favorable to the new
technology to top positions of power in governmental organizations is a well-used tactic – such as
when President Eisenhower appointed a committee to study the country’s highway requirements
and made Lucius Clay – a member of GM’s board of directors - its chairman.
Messaging about the robotaxi companies has come through the city and Bay Area media
providing positive coverage – but including the mishaps too. The biggest selling news media in
the city is the San Francisco Chronicle with a circulation of around 54,000. In an interview with
tech reporter of the Chronicle, Ricardo Cano, who covers the robotaxi stories, he made it clear that
he was extremely impressed with the robotaxis. He had just come back from a few days in
Phoenix and had just reported in the paper of his impressions and the conversations he’d had with
people there about them. Cano had covered many AV stories including the incidents that have
caused safety concerns and was well aware that in publishing such stories that it could give them
a disproportionate focus (the attention heuristic).
The influence of such traditional sources of news has declined considerably of course over
the last couple of decades being replaced by social media to a significant extent353 and these
sources have provided a wealth of publicity. Early adopters – those who have been able to access
rides in San Francisco - have tweeted, shared on LinkedIn or shared Instagram images and TikTok
videos about their experiences. These early adopters, ambassadors for the technology, will find
their posts re-tweeted or shared, especially by supporters of the technology including the AV
companies themselves. Social media has also been used showing AVs misbehaving and have been
352 Koehn, J. (2023, August 23) Here’s how much robotaxi companies spent to influence California regulators. SF
Standard. https://sfstandard.com/2023/08/23/cruise-waymo-spent-millions-to-lobby-for-free-rein-on-sanfrancisco-streets/
353 Statistica report that 50% of adults in the US get their news from social media often or semi-regularly
https://www.statista.com/topics/1640/news/#topicOverview
160
accepted as evidence by NHTSA as it cross-checks on completeness of the AV company incident
data returns.
From October 15, 2020 when safety drivers were no longer required in Cruise vehicles in
San Francisco, Cruise began running vehicles without them. Tweets from Cruise’s CEO Kyle Vogt
emphasized themes of fun, magic and the future – the feelings that appeared to be enjoyed by
the tech-savvy people of the city using Cruise, the experiences of whom were retweeted by Cruise.
This has echoes of the ‘pleasure car’ from the early history of the car and TNC advertising – and
Geddes’ Magic Motorways book of 1940 foreshadowing the Interstate Highway System.
The themes of the promotional material since reflect those at national level and historic
‘silver bullets’ for safety, congestion and environmental externalities. The companies have had to
pay more attention to safety though in San Francisco because of the significant concerns
expressed by emergency services and the well-documented incident there. Early in 2021, Waymo’s
messaging was leading on safety following consistently disappointing levels of trust in the city.
Their market research towards the end of the previous year had shown safety to be a key concern
in mobility issues in the city, on which they reported, “in Fall 2020, we surveyed nearly 1,000 San
Franciscans to understand their transportation needs. When asked to name factors making it hard
to get around the city, 63 percent of respondents pointed to dangerous drivers, 74 percent to
parking and 57 percent to stressful commutes. Worryingly, nearly a quarter didn't feel safe on San
Francisco’s roads at all”. The actual questions or choices given from which respondents could
choose is not shared (if an option is not given for ‘dangerous roads’ they may well pick ‘dangerous
drivers’).
Cruise too demonstrated a sharpening of its purpose, such as in a tweet on March 15, 2021
describing their, ‘mission to make transportation safer & more accessible’354. A similar mission
354 Vogt, K. (@kvogt). (2021, March 15). 1/ I’m pleased to welcome @OliverCameron & @Voyage to the
161
was announced by Waymo in an open letter to San Franciscans on 21 March 2021, “Our mission at
Waymo is to make it safer and easier for people and things to get where they’re going.355”
While at a national level AV companies talk of national crash numbers, they do not
translate into local numbers in the cities in which they operate. This is in part because of the
relatively small fleet size so hard to extrapolate but is also a rhetorical device: “numbers can be
and are used as a stylistic device, as a rhetorical means that contribute to a melodramatic world
picture356”. To focus on the 30 road deaths per year and 3,000 injuries357 that San Francisco
sustained in 2019, for example, would make the problem appear more manageable and less in
need of such a radical solution that AVs present especially if you bear in mind the impacts on VMT
and congestion. Indeed bringing those numbers down is precisely what SF Vision Zero is doing
through such work as mapping the crashes and injuries and assessing the road layout, traffic
signals and other factors that might be improved to make the city safer. A revolutionary change in
the city’s transportation may seem disproportionate alongside relatively small numbers, the
causes of which are easier to address and are being addressed.
San Francisco had a particularly high road death toll - 39 - in 2022 (half of which were
pedestrian deaths), a nearly doubling of the low of 20 in 2017358 in spite of increasing safety
features in cars and increasing automated features. Just over 9 million miles were traveled each
day in San Francisco in 2019 – with 29 deaths during that year that equates to over 300m miles
@Cruise team! Voyage is a nimble and highly capable company that shares our …
https://x.com/kvogt/status/1371484760028110850
355 Mawakana, T. (2022, March 21). To our fellow San Franciscans. Waymo. https://waymo.com/blog/2022/03/toour-fellow-san-franciscans/ Retrieved 20 February 2024
356 Roeh, I., & Feldman, S. (1984). The rhetoric of numbers in front-page Journalism: How numbers contribute to
the melodramatic in the popular press. Text & Talk, 4(4), 347.
357 Quoted by Jeffery Tumlin, in statement to House of Representatives Committee on Energy and Commerce
Subcommittee February 11, 2020 https://docs.house.gov/meetings/IF/IF17/20200211/110513/HHRG-116-IF17-
Wstate-TumlinJ-20200211.pdf page 2
358 SF.gov. (2023, October 3) Vision Zero SF: 2023 Update. SF.gov https://www.sf.gov/sites/default/files/2023-
09/2022%20VZ%20Fatality%20Report%20Out_2023-10-03%20Health%20Commission%20FINAL_v2023-09-
21_ca.pdf
162
between fatalities. This figure is worth bearing in mind when AV companies state the number of
miles they have driven so far. Only around 9m autonomous miles were driven in the state of
California over the year to December 2023 – yet every day over 9m miles are traveled in San
Francisco alone. To assess safety in San Francisco there would need to be multiples of 300m miles
to gain a true picture. This hasn’t been reflected in their promotion of safety which instead has
relied on such data as shown in Figures 4 and 5 below.
Aligning themselves with California’s safety program, in February 2021 Cruise had
tweeted, “Today, we’re proud to announce Cruise has signed on to support @VisionZeroSF and
the first-ever #ZeroTrafficDeaths goal, a campaign urging the Biden Administration to set and
advance the national goal of zero traffic-related fatalities by 2050.” It’s notable however that the
10-year review of vision zero in the city only mentions robotaxis in passing indicating a
Figure 4: Human Drivers vs Cruise AVs - Cruise Tweet from 28 April 2023 (and from 1m miles report?)
163
Figure 5: Updated Human Ride Hail Benchmark vs Cruise AVs in 1M (Cruise tweet from 27 Sept 2023)
preparedness to take on future challenges rather than a realistic, measurable contribution to the
strategy359
.
David Zipper, senior fellow at the MIT Mobility Initiative, has set out his assessment of the
focus on safety: “It is a cynical but savvy move for AV companies to focus on safety as their core
lobbying pitch, instead of on other options like convenience or access for those unable to drive”.
He points to the sort of measures in San Francisco’s Vision Zero: “Unlike more realistic roadsafety strategies like slowing down urban traffic, self-driving technology does not threaten the
primacy of the automobile in American life, which many public officials are wary of challenging. In
fact, overhyping the safety benefits of self-driving cars allows the auto industry to concurrently
fulfill two key objectives: It positions car companies as a solution to a American safety crisis they
themselves helped create, and it serves as a distraction from proven tactics (like road diets or
transit expansions) that make their cars and tech less useful in urban areas. And they have little
359 Vision Zero San Francisco (2024) 10-Year report prioritizing street safety. SFMTA
https://www.sfmta.com/media/40114/download?inline Retrieved August 4, 2024
164
to lose by exaggerating AV benefits; past promises of car-dominated utopias have repeatedly
come to naught without inspiring a regulatory smackdown or popular backlash.360” Zipper’s
assessment is entirely realistic – AV companies and the ecosystem that surrounds them are there
to sell cars and encourage vehicle use and very clearly they sidestep the fact that San Francisco
already knows how it needs to cut those road deaths and their working through their Vision Zero
strategy to do so.
The safety theme is echoed in a video interview carried out by another Cruise staffer with
former mayor of San Francisco, Willie Brown, retweeted by Vogt in March of 2021. In the interview
Brown points to the disadvantages of transportation when he was mayor, ‘unreliable cabs’,
‘transit iffy from a cost standpoint’ contrasting to life with AVs, ‘no more running red lights, drunk
driving… will be far safer’. It’s worth adding that Brown was taken on as a paid advisor by
lobbyists David Ho (Proverb Strategy Advisors) and Ramneek Saini (S2 Partners) who were hired
by Cruise from January 2022361. The ecosystem is growing and careers and money to be made by
championing the AV industry, reflecting the interest groups supporting the growth of automobiles
and freeway-building of the past.
Concerns about AV safety grew after 2020 and especially in the summer of 2023, around
whether the technology was in fact ready to be on city streets without a safety driver. This
emerged as noted after a series of mishaps involving emergency response vehicles and the city’s
fire chief, SFFD Chief Jeanine Nicholson, speaking out against AVs, supported by the
chief of police. She expressed her frustration not just at the technology but also the
response from the AV companies: “What we have gotten from them is: ‘Here is how you
360 Zipper, D. (2023, July 17) The Safety Dance: Autonomous vehicle companies claim that “humans are terrible
drivers” and their tech is needed to save lives. Don’t buy it. Slate. https://slate.com/technology/2023/07/cruiseautonomous-vehicles-safety-waymo-self-driving-cars-ad-new-york-times.html Retrieved 24 April 2024
361 Koehn, J. (2023, August 23) Here’s how much robotaxi companies spent to influence California regulators. SF
Standard. https://sfstandard.com/2023/08/23/cruise-waymo-spent-millions-to-lobby-for-free-rein-on-sanfrancisco-streets/
165
should interact with our vehicles.’ So just sort of that hubris,” she told Forbes. “Let’s turn that
around a second. Let’s talk about how you should interact with our vehicles.362” But Waymo and
Cruise insisted otherwise with Waymo tweeting on 25 July 2023: ‘“Autonomous vehicle technology
isn’t a panacea, but it’s now mature enough to make a meaningful impact on road safety”.
On 10 August 2023 the CPUC considered the decision to grant Cruise and Waymo licenses
to carry fare-paying customers. Stepping up the communication on safety before the crucial vote,
Waymo Co-CEO, Tekedra Mawakana tweeted, “Why do we need AVs? Look no further than the
stats: Since 2017, nearly 250,000 people have lost their lives on U.S. roads. We can’t accept the
status quo when improved road safety can benefit everyone. Learn more about our focus on
safety: http://waymo.com/safety”.
During that meeting on 10 August, only four of the five CPUC commissioners were present
for the hearings and vote, and the decision was approved 3 to 1, with one Commissioner absent.
The opposition that was expressed during the meeting was dismissed by Commissioner Reynolds
at the end, saying that the decision to be made was about collecting fares and that all other
issues raised by speakers (such as on safety and jobs) belong elsewhere. As noted, Commissioner
John Reynolds had been Cruise’s Managing Counsel from 2019 until he was appointed
Commissioner by Governor Gavin Newsom in 2021363– in spite of this he did not recuse himself
from the vote.
On the issue of conflicts of interest, Cruise and Waymo had been giving free rides to
people up to this point – including to disability groups and to the students at the University of San
Francisco thanks to Professor Billie Riggs, a keen supporter of AVs in the city, and author of the
362 Forbes (2023, December 5) The Mystery Around A Robotaxi, The Fire Department And A Death In San Francisco.
Forbes https://www.forbes.com/sites/cyrusfarivar/2023/08/30/cruise-robotaxis-waymo-san-franciscofirefighters/?sh=1f1ce96e1fea
363 CPUC (n.d.) John Reynolds, Commissioner. https://www.cpuc.ca.gov/about-cpuc/commissioners/pagecontent/profile-list/commissioner-john-reynolds
166
APA guidance on AVs published in 2018364. That guidance – which predicted a 43% or more
reduction in private car ownership, aimed to help planners prepare for the implications and
changes that AVs would bring which were particularly around less need for parking and the
changing use of curb space. Professor Riggs could therefore be seen to have led the approach
nationally for urban planners – an approach of acquiescence and accommodation – an approach
that urban planners of the past were encouraged to take for automobiles and freeway-building.
Transportation Equity
The American Council of the Blind and the National Federation of the Blind have lent their
support to AVs as they campaign on behalf of their members’ mobility365 – this was clear from
speakers at the CPUC meeting on 10 August 2023 and in statements in news and social media. AV
companies have worked with disability groups, promoting this relationship; Cruise, for example,
set up an Accessibility Council made up of professionals working in that area from other
organizations. The promotion of disability support brings a moral imperative and halo effect
which reflects historic promotion of transportation technologies (such as freeways reducing road
deaths and TNCs keeping women safe at night). The disability community however wasn’t unified
in support for the expansion at the CPUC meeting in August 2023. The disability community do not
all act as one and some have expressed opposition to robotaxis. Rebecca Miller, who had visual
impairment, urged the commission to vote ‘no’, citing concerns about pedestrians and a need for
more affordable transportation like transit366”. Indeed with many people with disabilities
dependent on transit the impact of AVs on San Francisco’s transit system could be very important
364 APA Jeremy Crute, J., Riggs, W., Chapin, T. and Stevens, L. (2018) Planning for Autonomous Mobility. American
Planning Association. PAS Report 592.
365 Autonomous Vehicle Industry Association (n.d.) Member spotlight: Cruise. AVIA
https://theavindustry.org/resources/blog/member-spotlight-cruise
366 Hawkins, A. (2023, August 10) Robotaxis score a huge victory in California with approval to operate 24/7. The
Verge. https://www.theverge.com/2023/8/10/23827790/waymo-cruise-cpuc-vote-robotaxi-san-francisco
167
for them; if, as with TNCs, AVs cream-off more wealthy riders then the viability of transit services
could be in doubt, decreasing mobility for riders with disabilities.
Already featured in previous tweets, issues of accessibility featured more prominently
after the October 3 collision. On Sept 14 2023 Kyle Vogt had tweeted, ‘The transportation status
quo is not only unsafe, it is inaccessible. Over 41 million Americans with disabilities deserve better
transportation options. I’m excited to introduce the @Cruise WAV, the world's first self-driving,
wheelchair accessible vehicle’. The wheelchair accessible vehicle (WAV) was due to be launched
in 2024 but has since been long-grassed. In spite of the talk of support for people with disabilities,
neither Waymo nor Cruise have been able to offer a single ride to a wheelchair user in San
Francisco - ever. As with transportation technologies of the past, vulnerable people face threats
to their existing mobility.
Silver bullet for congestion
Other concerns raised on 10 August included congestion, something that seemed far more
imaginable in San Francisco – at this micro level – than in the broad national messaging and
debate on AVs.
Although AV companies have pushed national messaging on congestion, it’s not
something they have pushed in San Francisco and there’s good reason for that. It wasn’t long ago
that TNCs were introduced with promises – including fewer cars – but the outcome has been very
different. Ricardo Cano, Tech Reporter with the San Francisco Chronicle, in my interview with him
suggested people in the city felt this new tech - with its similar assurances - is watched in the city
as “a familiar movie”. The experience of TNCs weigh heavy in the city and with AVs it could be
worse; as Julia Friedlander told me, “TNCs promised that they would reduce congestion, but
research by the California Air Resources Board, the San Francisco Transportation Authority and
168
independent academics demonstrated that TNC service - delivered in small passenger vehicles -
increased congestion in San Francisco. Many factors contributed to this result, but a key factor is
that the business model includes a high volume of driving (“vehicle miles travelled”) with no
passenger on the board (i.e. no “passenger miles travelled”). Transit service provides a much
higher volume of passenger miles traveled for every vehicle miles traveled.”
While Cano understood the concerns of the public sector in relation to TNCs he wondered if
their objections might be more fundamentally about their lack of control over the process; after
all, until the disruptive technologies of the 21st century, city government had been used to
controlling vehicles within its boundary. But the concerns have been justified by research. Key
findings from the SFMTA’s TNCs & Congestion Report of October 2018367 included that collectively
the ride-hail services accounted for:
● 51% of the increase in daily vehicle hours of delay between 2010 and 2016;
● 47% of the increase in vehicle miles traveled during that same time period; and
● 55% of the average speed decline on roadways during that same time period.
● On an absolute basis, TNCs comprise an estimated 25% of total vehicle congestion (as
measured by vehicle hours of delay) citywide and 36% of delay in the downtown core. 368
367 The report notes, ‘The report utilizes a unique TNC trip dataset provided to the Transportation Authority by
researchers from Northeastern University in late 2016, as well as INRIX data, a commercial dataset which combines
several real-time GPS monitoring sources with data from highway performance monitoring systems. These data
are augmented with information on network changes, population changes, and employment changes provided by
local and regional planning agencies, which are used as input to the Transportation Authority’s activity-based
regional travel demand model SF-CHAMP’. https://www.sfcta.org/sites/default/files/2019-
05/TNCs_Congestion_Report_181015_Finals.pdf
368 SFCTA (2018, October) TNCs & Congestion. SFCTA. https://www.sfcta.org/sites/default/files/2019-
05/TNCs_Congestion_Report_181015_Finals.pdf
169
With TNCs being the main competitor to robotaxis there’s natural concern that the similar
increases in congestion would follow AV use; and that’s not the only source of potential
congestion. There’s concern too around the impact of AVs on VMT and reflected in research on
connected and autonomous vehicles (CAVs). For example, research by UCLA in December 2022
concluded that for Southern California, ‘The vehicle miles traveled (VMT) and greenhouse gas
(GHG) emissions increase significantly with the deployment of CAVs. Compared with the base
model, the travel demand predicted by the CAV model would lead to a 9.1% increase in VMT and
around a 9.6% to 10.4% increase in GHG emissions369’. There seems little reason to believe there’d
be a better outcome in northern California.
Silver bullet for the environment
This UCLA research that raises concerns about GHG emissions contrasts with the public
messaging of the AV companies that have made much of their low emission vehicles. On October
15, 2020, for example, Cruise tweeted, “Before the end of the year, we’ll send cars out onto the
streets of SF — without gasoline & without anyone at the wheel. Because safely removing the
driver is the true benchmark of a self-driving car & because burning fossil fuels is no way to build
the future of transportation.” And in April 2021 “100% all-electric. Powered by 100% renewable
energy. We're choosing a cleaner future now, as we develop our technology, and we challenge
others in the AV industry to do the same. #EarthDay2021”. Cruise’s 2022 Impact assessment
describes its achievements – “As an all-electric, renewable-powered AV fleet, a trip with Cruise
offers our customers a chance to travel emissions-free, regardless of their ability to purchase an
electric vehicle”.
369 Jiang, Q., Yueshuai He, B., & Ma, J. (2022). Connected automated vehicle impacts in Southern California part-II:
VMT, emissions, and equity. Transportation Research. Part D, Transport and Environment, 109, 103381-.
170
Of course legislation is in place for all new cars in California to be electric by 2030 so these
claims are increasingly losing their green edge; you don’t need to have AVs to have zero emission
vehicles. There’s also an issue of electricity source – if fossil fuels are being burnt to produce the
electricity then that’s not very green. Cruise though had an answer for that. In summer 2021,
Cruise launched a ‘Farm to Fleet’ program, whereby it purchased renewable energy credits
directly from two family-owned farms in California’s Central Valley ‘creating a pathway for farms
generating on-site renewable power to participate in and financially benefit from transportation
electrification’. This was presented as a way to help the local economy as well as the
environment.
Silver bullet for the economy
For California, it is the tech sector that has driven the economic growth of the state and
Governor Newsom and other senior state politicians have championed the industry for that
reason. It has brought jobs and with the multiplier effect370 has benefitted, not just those who
work for tech companies, but businesses that provide services and products to those workers.
Newsom’s tech-friendly approach is a long-standing one from earlier in his career. In 2011 when
he was lieutenant governor Newsom set up his San Francisco office in Founder’s Den – a home for
start-ups.371 On a visit to Shanghai in November 2023, Newsom talked of the sector’s continued
importance, as the LA Times reported, “Newsom made it clear that he’s committed to keeping
California the global leader in the development of autonomous technology and said the state
shouldn’t “cede the future” to other countries or states”372
.
370 Moretti, E. (2012). The new geography of jobs. Houghton Mifflin Harcourt.
371 Siegler, M.G. (2011, February 28) Meet The Newest Founders Den Tenant: Lieutenant Governor Gavin Newsom.
Tech Crunch. https://techcrunch.com/2011/02/28/founders-den-gavin-newsom/ Retrieved 16 July 2024
372 Rosenhall, L, (2023, November 12) Gavin Newsom is mesmerized by the growth of driverless cars. Other
California Democrats, not so much. Los Angeles Times. https://www.latimes.com/california/story/2023-11-
12/gavin-newsom-driverless-cars-autonomous-vehicles-california-democrats Retrieved 16 July 2024
171
The CA Chamber of Commerce has also been a cheerleader for the industry and for AVs in
particular. It has stated in policy briefings that, “encouraging the development and providing
incentives for deployment in the commercial sector as well as the passenger vehicle space will be
essential to continue California's place as a technological leader”373. The Chamber, though says,
“the perception of the safety of AVs has been hampered by some high- profile accidents... Despite
these fears, experts predict that AVs will end up much safer than human-control ones”. They
advise, “policymakers must balance safety, avoid conflicting regulations with other States and the
federal government, and avoid overly prescriptive and burdensome regulations that impede the
continued safe testing and deployment of AV technology’. The outcomes will decide “whether
California can be a leader in this technology, or whether companies will choose other states with
less restrictive policies for larger-scale deployment.” Their position in summary is that regulation
should not be “overly burdensome”. This pro-tech position is not surprising from the Chamber of
Commerce which has a strong history of supporting transportation technology and increased
private mobility; but, as with supporting freeway-building to enable suburbanites to favour doing
business in the city core, they don’t have a perfect track record in predicting what is best for
cities’ long-term interests.
The views of business diverge from those of city government in San Francisco. In my
interview with Consultant, Barry Einstig, who was part of the team that produced CALSta’s
principles for AV development, he pointed to the division in perceptions between the city
government and city business. He said that the City of San Francisco has seen AVs as, “a burden”
and that he didn’t think city leaders understand the potential or economic opportunity of AVs,
benefits that have been championed by business organizations in the city.
373 CA Chamber of Commerce (n.d.) Autonomous Vehicles. California Chamber of Commerce.
https://advocacy.calchamber.com/policy/issues/autonomous-vehicles
172
While the perception may be that California over-regulates, this is what businesses always
say. While California may appear to be on the high end of regulation that perhaps reflects the
looseness of other states’ regulation. Reflecting this, Riccardo Cano, Tech Reporter at the San
Francisco Chronicle told me, “Is there too much regulation? No one really is checking to see how
it’s going; there’s incredible latitude on the part of companies. There are arguments over
regulations which I think are a way of officials trying to get some sense of control”. This looseness
is also reflected in Professor Koopman’s description to a congressional committee on the
processes, noted earlier.
As in the past with cars and freeways being a tool to ‘decongest cities’, the AV industry
has sought to shape messaging around the political priorities of the day. Promotional messages
from tech companies have more recently included a sense of being community-based and
supportive of the community. This messaging aligned with the objectives of politicians and the
public sector organizations that the tech and auto companies aim to influence and perhaps are
messages that have come through in their market research. Talking up the benefits to the
community is important as a promotional tool.
The old narrative around tech entrepreneurs being college dropouts and young mavericks
creating things in their garages are no longer foundational to how people see tech companies.
Tech is ‘Big Tech’ and public perception is perhaps more influenced by the misuse of personal
data by Facebook and Cambridge Analytica; Bezos’s wealth, presiding over a behemoth; with
Amazon’s profound negative impacts on downtowns; and Uber’s apparent exploitation of lowpaid workers.
Perhaps in an attempt to overcome these perceptions, AV company messaging focuses on
how they have tried to engage with the communities in which they operate in very visible ways.
Cruise for example had used its vehicles to send meals to homeless people during the
173
pandemic374. Another clear example is the jersey patch used as Cruise sponsorship of the SF
Giants in August 2023. Announcing the sponsorship in a tweet shared by Cruise, the Giants
voiceover refers to the Cruise designed jersey patch saying, “This patch. It's more than thread and
stitching. It's a symbol of our city. An emblem of conviction. Of progress. The giants and Cruise
together driving our city towards a brighter future.” The meshing of company and community
contrasts with the perception of big auto companies (such as GM) and wealthy global tech
companies that appear somewhat community-less. Of course these don’t hide the fact that the
AV companies chose San Francisco, the ‘community’ did not choose them.
Cruise also appeared to have developed its brand as something of a community-rooted
organization, working with community groups and helping groups in need or partnering events
such as Afrotech Conference 2022, SF Pride, Women’s History Month, Black History Month and the
Chinese New Year. Both companies have supported community causes notably Cruise’s work
making food deliveries from SF-Marin Food Bank that started during the pandemic, Kyle Vogt
posted a tweet in September 2023 saying, ‘Together with our community partners, we've now
delivered 2,500,000 meals to those in need throughout SF.’ Waymo’s work has included donating
thousands of pounds of water and other goods to the Salvation Army of Texas to support disaster
response efforts in the summer of 2021 and thousands of backpacks to children in Chandler in the
summer of 2021. These references to community, to giving food to the homeless and books and
backpacks to children can be seen as a way of detaching themselves from Big Tech and meeting
the objectives of the regulators and politicians in the cities in which they operate and appealing to
a section of consumers more likely to use their services.
The community focus has also been reflected in an all-staff message to Cruise staff in
December 2023 post-dragging incident. Forbes quotes new President and CTO, Mo Elshenawy,
374 Hawkins, A. (2020) Cruise redeploys some of its self-driving cars to make food deliveries in San Francisco
The Verge. https://www.theverge.com/2020/4/29/21241122/cruise-self-driving-car-deliveries-food-banks-sf
174
"We remain focused on commercializing a fully driverless L4 service, relaunching ridehail in one
city to start ... Our priority from day one will be to launch with communities, not at them.375”
To increase the perception of concern for equity, Waymo has also used X (Twitter) to
celebrate Black History Month 4 February 2021, National Reading Month March 2021, Global
Accessibility Awareness Day 20 May 2021, International Day of Women and Girls in Science on 11
February 2022, and International Women in Engineering Day 23 June 2022.
Public Response to AVs – and signs of winners and losers
An important actor in the development of technology is of course the consumer; their
response is vital in achieving a breakthrough and securing momentum. San Franciscans haven’t
been asked for their views on AVs by governmental organizations – and while the tech companies
have asked some, they have not shared any meaningful findings (except in a very headline,
positive way). The public though has had the opportunity to express views in meetings and
workshops of the CPUC, although those doing so represent a tiny fraction of the city population.
In the tradition of San Francisco’s transportation protests, there have been demonstrations
against AVs – notably the ‘coning’ by Safe Street Rebel. Here a pro-active travel organization,
that organizes regular bicycle rides in the city, protested the presence of AVs by placing traffic
cones on their hoods (when there is no one inside the vehicle) which has the effect of disabling the
vehicle. In a more violent protest (not a Safe Street Rebel one) a Waymo was graffitied and set
alight in China Town in February 2024 during a street festival.
While there have been somewhat isolated protests, including at the CPUC and other
meetings, there has not been a public uprising or any mass demonstrations and widespread
375 Forbes (2023, December 5) Cruise will re-launch robotaxis with communities, not at them. Forbes.
https://www.forbes.com/sites/cyrusfarivar/2023/12/05/cruise-will-re-launch-robotaxis-with-communities-not-atthem/?sh=592e8e894c5a
175
negative sentiment does not appear to be evident. In my interview with Julia Friedlander she
noted that, “There’s a very wide range of opinion. Some fear them – there’s no possibility of eye
contact to assure safety - what do parents tell their children – it used to be ‘always look the
driver in the eye’. Some are terrified by AVs and feel bullied by them. On the other hand, some
people – including pedestrians and cyclists – find AV driving to be more consistent and
predictable and this increases their perception of safety”.
Ricardo Cano, the tech reporter for the San Francisco Chronicle, agreed it wasn’t clear how
the public generally perceived the robotaxis. Of the negative sentiment he told me, “It’s hard to
know if it’s widespread. Do people have faith in the tech? It’s hard to say - hard to tease out.
Some people don’t want it to succeed but some like it. Skepticism has been amplified because of
news coverage. Folks most vocal have been city officials trying to get the state government to
understand their position”.
Cano pointed to a source of some of the distrust that springs from recent city history,
“Some of it is historic context – they remember what happened with Uber and Lyft, which started
in SF. There weren’t many guardrails, oversight and regulations. It rankled the City guard and they
don’t want that to happen again. The city is still dealing with the effects – the congestion
downtown is the result of ridehail and they feel, ‘this is a familiar movie; let’s not go through that
again’”.
I suggested to Cano that some of the opposition to AVs might be a display of what could
be seen to be a culture of protest against car domination in the city, as seen in the Freeway
Revolts. He agreed, noting the vibrant community and vocal group of residents who are prowalking and bike infrastructure. “It’s the same groups of folks who oppose highway widening and
favor pedestrianization who come out and oppose AVs”, he said. Even this group was divided
though according to Cano, “some people who are in this camp also think AVs are a lesser evil to
human driven cars”.
176
Although the AV companies have not shared their surveys in the city, J.D. Power has
carried out surveys to gain an understanding of consumer viewpoints. They sampled 408 vehicle
owners residing in Phoenix and San Francisco (175 of whom lived in San Francisco). Of those 408,
154 were riders of the city’s robotaxis and 254 were non-riders (had not been in a robotaxi). Of
the non-riders in San Francisco the mixed views are apparent in their views on sharing the road
and interacting with robotaxis (Figure 8). Both cities had similar levels of comfort in spite of the
different AV experiences. 28% of non-riders in Phoenix felt not at all or not very comfortable
compared with 29% in San Francisco.
J.D. Power’s findings on the riders however were particularly interesting (Figure 9). J.D.
Power found 53% of riders were very or extremely comfortable with AV technology being tested on
streets and highways near them, so even with this self-selecting group, almost half (47%) were
only somewhat comfortable, not very comfortable or not at all comfortable.
Figure 6: Comfort level sharing the road and interacting with fully automated, self-driving
robotaxis: non-rider. Source J.D. Power 2023 U.S. Robotaxi Experience Study, SM. Used by
permission.
177
Figure 7: Comfort Level With AV Technology Being Tested On Streets and Highways Near You376
Source: .D. Power 2023 U.S. Robotaxi Experience Study,SM. Used by permission.
Terra Curtis, my interviewee from the CPUC, thought the general public were broadly
responding with ‘curiosity’, but acknowledged the complex relationship between the Bay Area and
the technology industry, which has left some feeling displaced by it while others benefit. This
seems a very plausible explanation for some of the animosity in the city towards AVs.
Figure 8 shows the changes in median household income for the top 10% highest income
households in the Bay Area had increased by 87% between 2010 and 2019. Figure 9 shows this
growing divergence between the rich and the poor in San Francisco – a growing gap in income
demonstrated by a ratio growing from 1:22 in 2010 to 1:28 in 2022.
376 J.D. Power (2023, July) White paper, following survey
178
Figure 8: Income of highest and lowest earning households in the Bay Area, California and the
United States as a whole between 2010 and 2019377
.
Figure 9: Increasing ratio between the highest and lowest 20% of earners in San Francisco,
demonstrating increasing income inequality378
377 Bay Area Council Economic Institute (n.d.) How has Bay Area economic inequality changed over the past
decade? Bay Area Council. https://www.bayareaeconomy.org/report/bay-area-income-inequality/
378 Federal Reserve Bank of St Louis. (2024, December 12) Income Inequality in San Francisco County, CA. Fred.
https://fred.stlouisfed.org/series/2020RATIO006075#
179
Beneath these figures other dynamics can’t be easily seen, notably the number of people
who have been forced out because of the increasing cost of living in the city. The received wisdom
is that growth benefits everyone – Enrico Moretti’s the New Economy of Jobs for example
indicates that for every one tech job a further eight service jobs are created379. While growth in
tech industry has a multiplier effect, other jobs are much lower paid; as those at the top increase
demand and cause prices to increase, those at the bottom are priced out.
The city’s public view of the contribution of tech companies to the city is reflected in a
survey in the San Francisco Chronicle of a random sample of 1,653 residents, reflective of the city
demographics, in the summer of 2022. This San Francisco/ SF Next poll found that overall, while
41% of the surveyed residents said that tech companies and workers are a good thing, 25% said
they were a bad thing and 34% thought them neither good nor bad. So a quarter of the sample
viewed it negatively. It shows wealthy residents (earning over $200,000 pa) are half as likely to
see tech sector growth as a bad thing compared to those earning less than $50,000. There is also
a marked racial difference too with Hispanic people being twice as likely as Asian people to see
the growth in tech companies as a bad thing. Older residents (those 65 and older) also show a
marked difference in views – being significantly less negative about the sector’s growth compared
to younger cohorts (perhaps reflecting an age group that is more likely to have bought a home
and seen its value significantly increase).
My interviewees in San Francisco very clearly connected the experience of AVs with the
wider context of the experience of the tech sector in the city. Ricardo Cano pointed to the sense
that AVs revisit the same scenarios – specifically the TNC experiment - San Franciscans have
experienced in the past and it didn’t end well for the city. This context has impacted on how AVs
379 Moretti, E. (2012). The new geography of jobs. Houghton Mifflin Harcourt
180
are perceived and people can perhaps be forgiven for thinking the impact of AVs would be a
continuation of past patterns of winners and losers. Julia Friedlander noted, “Most people who
travel in San Francisco don't use TNC's and some are annoyed by the perception that Uber and
Lyft increase traffic. This history may affect how they feel about AVs”.
These tensions have been played out visibly in the attacks on tech worker buses some ten
years ago. In 2014 there were a number of attacks on buses taking tech workers from San
Francisco to Silicon Valley. Small projectiles were thrown, bus windows broken, tech workers left
scared or at the very least embarrassed. Protesters used an Instagram-friendly banner reading,
‘Gentrification and Eviction Technologies’ in Google colors and font - the image went viral380. By
2014 protests had stepped up with protesters throwing rocks at the Google buses in Oakland.
The free bus service for staff at Apple and Google were condemned for making it easier to
live in the city and commute out. Google’s view was it was replacing many thousands of VMT that
the employees would otherwise have spent on the freeways. The Guardian claimed, ‘The core
grievance is one keenly felt by almost everyone in San Francisco: the way the tech sector has
pushed up housing prices in the city and made it all but unaffordable for anyone without a sixfigure salary. Almost no San Francisco police officer lives in the city any more, and neither do most
restaurant workers or healthcare workers’381
.
The anger centered on the use of city bus stops to transport workers to the various tech
campuses and the insufficient response from the city government who simply charged the
companies $1 per bus to use a bus stop and road. The Verge reported, ‘Though protesters do take
issue with the buses directly because they use the city's bus stops without paying for the privilege,
they're mostly seen as a prominent symbol of growing inequality in the Bay. Today's protests
380 Rushkoff, D (2016) Throwing Rocks at the Google Bus. Portfolio/ Penguin (New York). p2
381 Gumbel, A. (2014, January 25). San Francisco's guerrilla protest at Google buses swells into revolt. The
Guardian. https://www.theguardian.com/world/2014/jan/25/google-bus-protest-swells-to-revolt-san-francisco
Retrieved 19 Jan 2024
181
centered on low-income tenants evicted from their homes as a result of the area's housing
situation, a situation some blame on the high-income individuals employed by tech companies —
who have been bidding up housing prices in the area382’.
In Throwing Rocks at the Google Bus, Douglas Rushkoff explains that ‘Growth is good – at
least for those doing the growing. But the influx of Googlers to San Francisco's most historic
neighborhoods also raised rents, forcing out longtime residents and small businesses that were
not participating in all this growth. Google’s air conditioned tour buses epitomized the seeming
invasion - like space transports taking aliens to and from their mother ship. Adding insult to injury,
Google was now using publicly funded bus stops as loading stations for its very private
transportation system. Rents close to those bus stops were 20 percent higher than those in
comparable areas, which were themselves doubling every few years to accommodate not only
Google's employees but those of Facebook, Twitter, and other Silicon Valley darlings.’383
.
A sense of powerlessness is reflected in findings in research in the city by Maharawal, “At
a protest, Patty, a Mission resident fighting her eviction, said to me, “It’s as if tech thinks they can
do whatever they want, and the city just looks the other way! No one else gets to act like that ...
but we had to show them that!” 384. Maharawal adds, “to bolster the claims that the buses were
connected to displacement the Anti-Eviction Mapping Project (AEMP) produced a map that
showed a 69% increase in no-fault evictions between 2011 and 2013 for apartments near tech
shuttle stops …. In addition, a 2013 study found that the tech shuttle stops coincided with higher
rents: “rental prices within a walkable distance of the shuttle stops appear to be increasing at a
faster rate than rental prices outside the walkable distance” in a phenomenon that real estate
382 Hollister, S. (2013, December 20) Protesters block Silicon Valley shuttles, smash Google bus window. The Verge.
https://www.theverge.com/2013/12/20/5231758/protesters-target-silicon-valley-shuttles-smash-google-buswindow Retrieved 19 Jan 2024
383 Rushkoff, D (2016) Throwing Rocks at the Google Bus. Portfolio/ Penguin (New York). p1-2
384 Maharawal MM. Infrastructural Activism: Google Bus Blockades, Affective Politics, and Environmental
Gentrification in San Francisco. Antipode. 2023;55(5):1454-1478.
182
agents called the “shuttle effect” (Goldman2013).” From the historic, city perspective, robotaxis
have arrived at the end of a long line of changes that have enriched a part of the city but pushed
out or left out a great many others. There’s an important equity point here too; people of color
have been displaced at disproportionately higher rates than white San Franciscans385
.
While AVs have been promoted as beneficial in freeing people from the driving task in San
Francisco this feature would benefit some more than others. Transit riders won’t be affected
because they can work on the go anyway – but lower-paid manual or service workers can’t work
while on the move – their roles are very much in-person and in their place of work. The winners in
alleviating the burden of driving in an AV are likely to be white collar workers who can start their
working day once they get into the AV: cleaners can’t start to clean; bus drivers can’t start their
shift; care workers can’t bring their clients down for breakfast while in a robotaxi.
These issues brought to the surface a seemingly underlying discontent between the San
Franciscans and the tech sector which is perhaps still there. The ‘coning’ of AVs can be seen as a
continuation of active protest against the sector; these actions weren’t drunken high jinks – they
were a thoughtful protest, chiefly instigated by Safe Street Rebel whose modus operandi is more
around organizing safe cycling events across the city than inciting revolution.
Transit
San Francisco has, as noted, a strong culture of transit use but that has declined
significantly, especially since the pandemic. BART ridership in March 2019 was 409,515; in March
385 Maharawal M (2017a) San Francisco’s tech-led gentrification: Public space protest and the urban commons. In J
Hou and K Knierbein (eds) City Unsilenced: Urban Resistance and Public Space in the Age of Shrinking Democracy.
(pp 30–43). New York: Routledge
183
2024 it was 162,459386. This decline in transit use is a grave concern for the city and AVs could
lead to a further peeling away of ridership with a consequent negative impact on the service and
users.
An important group who are at risk of losing out if AVs do develop and become widely
adopted are those who are transit dependent. The lowest income groups in the city are the
primary frequent Muni users and frequent users are much less likely to be White387. AVs could
make transit dependent people more vulnerable especially if the more wealthy people no longer
use transit because of the availability of AVs that, it is claimed, would be cheaper than a TNC. If
you take away from transit all the people who can afford to buy a monthly pass, that’s an
existential threat in San Francisco. There are issues of equity but also issues of impact – the
viability of a functioning transit system – and this has equity results. Ricardo Cano recognized this
problem saying he couldn’t see how transit would be helped in San Francisco. He noted that there
is opposition or skepticism beyond city officials that “once the technology is ‘fully fledged’ it’ll
perpetuate what we have (single use vehicles). There are flowery ambitions but I can’t see how
transit will be helped”.
Working from homes and more flexible working schedules have had a significant effect on
ridership since the pandemic. In a March 2023 webinar was hosted by the Bay Area Council, a
business and civic organization, which pointed to the problem of higher costs, lower ridership and
rider safety concerns. The session showed that leaders of San Francisco-area public transit
agencies were worried about their ability to continue operating at current levels, without sources
of additional funding, let alone address rider concerns. “Bay Area Rapid Transit will run out of
386 Metropolitan Transportation Commission, Monthly Transportation Statistics, Update April 11, 2024
https://mtc.ca.gov/tools-resources/data-tools/monthly-transportation-statistics
387 SF.GOV (2024) San Francisco City Survey Data 2024. City and County of San Francisco.
https://www.sf.gov/sites/default/files/2023-04/City%20Survey%202023%20Summary%20Report.pdf p.8 Retrieved
August 4, 2024
184
federal emergency funds in 2025”, said Alicia Trost, BART chief communications officer. “We need
to find new funding so that it’s not on the back of our riders. If we can’t get funding, we’re going
to have to close early. There will probably be no weekend service. We’re going to have to cut
weekday service. We may have to close stations. ….If these transit systems have to cut service or
raise fares, the burden will fall harder on people of color. Black workers relied on transit in the San
Francisco-Oakland region to get to work at twice the rate of White workers in 2021, according to
TransitCenter, a foundation focused on urban public transit. Latino workers relied on transit 50%
more than White workers, it reports”388. This of course has significant impacts on the ability of
people to reach places of work, limiting employment opportunities with knock-on effects on
household income and life opportunities.
The situation risks a spiral of decline as explained by ratings agency, Moody’s: “Citing the
San Francisco Bay Area Rapid Transit District as a case in point, Moody’s noted that service cuts
contemplated by BART “illustrate some of the decisions facing transit systems when considering
service levels,” adding that such changes “risk exacerbating the ridership decline and, in turn, can
make it more challenging to secure needed government funding.”389
So the risks to transit are significant and my interviewees were also conscious of the
history of winners and losers in the recent past with TNCs and the benefits of ridehailing have not
been distributed equitably. As Boarnet, Shao and Pilgram note, “Several studies found that
ridehailing users tend to be younger, better educated, earn higher incomes than the general
population, are more likely to be childless, and tend to live in dense urban areas”. The low usage
388 Zukowski, D. (2023). San Francisco transit agencies face “worst-case scenarios” for revenue, potential service
cuts, leaders say. Smart Cities Dive, Retrieved from
http://libproxy.usc.edu/login?url=https://www.proquest.com/trade-journals/san-francisco-transit-agencies-faceworst-case/docview/2786645753/se-2
389 Zukowski, D. (2023). Transit agencies’ financial woes continue. Smart Cities Dive, Retrieved from
http://libproxy.usc.edu/login?url=https://www.proquest.com/trade-journals/transit-agencies-financial-woescontinue/docview/2894579385/se-2
185
amongst low income is very apparent from their data from the National Household Travel Survey
(2017), the American Community Survey (2015-19) and San Francisco’s own MTC Travel Diary
Survey (2019). The latter shows that for San Franciscans the beneficiaries of TNCs have been
Non-Hispanic white (59% of the city’s users) and Asians (24% of users) – a race profile that’s likely
to be mirrored with AVs in the city.
When I suggested AVs were a means of mobility for people who are relatively affluent and
that they won't help low income people; Terra Curtis from the CPUC thought this, ‘a reasonable
assessment’ for the current state of the industry, but hoped it might help others too. Both Curtis
and Freedlander were positive about the potential of the technology and clearly hoped it would
succeed in delivering on the expectations of the technology companies; but there was a
recognition that this hopeful vision is not guaranteed.
In my interview with Timothy Mathews, Research Director for Teamsters 856, who spoke
at the 10 August CPUC meeting, his view was that the use case of robotaxis was not the ultimate
goal of the tech and auto companies. He was convinced that the goal was logistics and deliveries
because this is where more money could be made. Taking the driver out of the equation would
bring more significant cost reductions without the overheads that may come with robotaxis (such
as trash left in the vehicle, drink spills, downtime waiting for riders). Although the shortages of
some truck drivers has been a problem, there aren’t such concerns with delivery drivers for
groceries and other retail purchases and this could be an important area of job loss.
Who these people are and whose jobs are at risk is important. These are not well-paid
people for the most part – especially delivery drivers – and they are more likely to be People of
Color. While the US White (alone) population was 61.6% in the 2020 census, White taxi drivers
accounted for only 52.5% of taxi drivers; in contrast the US Black (alone) population was 12.4% but
they made up double that (24.9%) as a proportion of taxi drivers; Hispanic or Latinos made up
186
18.7% of the US population but again, over-represented in taxi jobs – making up 23.9% of the
workforce390
.
Who will be the winners and who will be the losers in all this? I asked Mathews; “The
winners, a bunch of rich white guys”, he said, “Elon Musk, the angel investors who threw money at
Cruise, private equity, those folks. The losers will be TNC drivers, delivery drivers, and people who
live in the city who would have to deal with the congestion, induced demand, slower transit, air
pollution and degraded road maintenance”. It seems clear from the data that the people who
would bear the negative impacts in this case are low income and People of Color – as has been
the case with transportation technologies of the past, especially the highways that displaced and
divided Black communities.
This raises an important point – with AVs still malleable they may develop into a middle
class alternative to transit in San Francisco, a cheaper delivery method, and the new addition to a
household’s fleet in addition to a manually driven car. It’s here where the existential threat to
transit lies and the risk of more vehicles in the city and more of them on city streets.
Conclusion
The case study of San Francisco shows that there are strong echoes of the past and of the
national picture playing out in the city. We see government – in San Francisco’s case, the state
government – acting as an enabler to the development of the technology. We also see an
increasing number of interests gathering in support of the technology – the local business
community and tech industries. We see urban planners having a passive role at city government
390 US Census Bureau (2021) Improved race ethnicity measures reveal United States population much more
multiracial. https://www.census.gov/library/stories/2021/08/improved-race-ethnicity-measures-reveal-unitedstates-population-much-more-multiracial.html Retrieved 20 September 2024 and US Bureau of Labor Statistics
https://www.bls.gov/cps/cpsaat11.htm Retrieved 20 September 2024
187
level and a leading academic at San Francisco State University encouraging the development with
little challenge and setting standards nationally for planners to follow.
On the promotion we see the familiar silver bullets – that AVs will solve the safety,
congestion and environmental problems. And newer claims to be working towards equity and
community value. We see the use of a moral imperative – just as freeways and TNCs were
promoted – saving lives. This goes further with AVs with the industry promising San Francisco a
mobility to those with disabilities in spite of the fact they have never once provided a ride to a
wheelchair user. While noble use cases have been proposed and even trialed in the early days of
transportation technologies (such as TNC used as connectors to transit), the reality has been that
they operate where most profit can be made and no rules have ever been set in place to ensure
TNCs or other new transport tech sacrifice profit in the long-term to benefit the low-income and
those with fewer mobility options.
As with previous technologies too we see the business community and state government
looking for loose regulation to encourage technology development with the expectation that it will
lead to jobs and local economic growth. This may be true but as we’ve seen in the past there are
other consequences that this need to be weighed against.
For a profitable business we might expect that this interregnum will be seen as a period of
what a coevolutionist would see as a period when momentum would be created. It’s a vital part
of a successful transportation technology development – once momentum has been achieved it
becomes impossible for governments or the public to control the technology. Key to this will be
the public’s willingness to embrace the technology and in San Francisco, apart from isolated
protests and concerns expressed by the city and some politicians at state-level, the public has
largely acquiesced to the addition to the city’s transportation offer. It may be that the equity
issues around job loss of TNC and delivery workers would dampen public support.
188
We can see from the experience of TNCs that in spite of what was portrayed as a seismic
shift in how people would travel, this does not describe the impact they have had so far. Although
the use of TNCs has increased at the expense of traditional taxis and transit, they did not lead to
the sharing culture and societal shifts in transportation mode that were predicted. The same may
be true for autonomous vehicles. There seems to be little reason to think that given people have
been unwilling to share rides in TNCs that they would be any more willing to do so in an AV.
While in theory the costs of running a vehicle without a driver should be cheaper, we don't
know the additional costs associated with people in an unsupervised vehicle, the cost of
congestion charges to manage additional congestion or the cost of having technicians at hand to
handle problems. For San Franciscans this means we cannot know if the benefits of AV rides
would widen or reduce the mobility of those on low income but it would appear they could have a
significant impact on transit.
We also don't yet know how strong the culture of transit is in San Francisco but we know
that some have already been encouraged to forego transit in favor of TNCs. That experience has
shown that when a transportation technology arrives that is trusted, easy to use and at a
reasonable price then San Franciscans will use it, with consequent impacts on other modes
(notably transit) and on congestion and livability.
189
Case Study of Phoenix, AZ
“The tone in Phoenix is a couple of octaves lower; it seems more normalized there – it feels the
technology is further along than people may have thought. There are hopes that it will make
Phoenix more walkable”.
Ricardo Cano, Reporter, San Francisco Chronicle - impressions from his Phoenix visit
Background
Why Phoenix was chosen
The Phoenix metro area has hosted AV trials in the city of Phoenix and other suburban
cities, notably Chandler and Mesa. Phoenix (by which I mean the metro area) presents as a strong
case study because in many ways it serves as a counterpoint to San Francisco. First, it has not
had the very negative publicity and incidents experienced in San Francisco. Second, the street
layout is very different being wide with a grid-structure, as a result of being a city developed
much later and around the needs of the modern car driver. Third, the city has been more
welcoming of the technology.
The weather certainly helps AV developers and contrasts with the less predictable weather
in San Francisco. The city has a hot desert climate that’s easier for AV companies to develop their
technology than cities that endure rain, fog and snow.
190
Waymo, one of two companies operating in the area, points to some of the quirks that
also add appeal: “We want to learn how sensors react to the dust in the air, and other area quirks
— unique vegetation and golf cart crossings, for example. There’s a lot of enthusiasm for tech in
Chandler and residents are supportive of self-driving cars being on the road.”
Cruise has offered this summary of the appeal of the city: “We began testing our selfdriving technology in the Phoenix metro area because the driving environment offered valuable
insight early on that was utilized in our other markets. By comparison, Phoenix’s traffic and road
conditions are more forgiving than our other locations, which gave us more latitude when testing
new features and hardware.391”
Waymo’s boasts that its trial area in Phoenix, at 315 square miles, is the largest
autonomous, ride-hailing territory in the US392. The scale of the trial makes for a useful case study
in seeing robotaxis operate across a wider area. The city has a relatively long history of AV testing
and regulation, with the state’s governor signing one of the first pieces of enabling legislation in
the US in 2015. This longer history enables more scope to examine and understand the power and
processes at work.
Importantly, the regulatory approach of Phoenix contrasts greatly with that of San
Francisco; in some ways they are at either end of the regulatory oversight spectrum. California
regulates its AVs more strongly thanArizona. . Governor Doug Ducey, who enabled the launch of
the AV trials, emphasized in his State of the State speech in 2017 that Arizona’s approach would
be “the opposite approach” to California, which he said had moved “backwards with more nutty
391 Greater Phoenix Economic Council (2023, April) Autonomous and Electric Vehicle Ecosystem. GPEC.
https://www.gpec.org/wp-content/uploads/2021/12/Autonomous-Electric-Vehicles-Use-Case.pdf
392 The Waymo Team (2024, June 5) Largest Autonomous Ride-Hail Territory in US Now Even Larger. Waymo.
https://waymo.com/blog/2024/06/largest-autonomous-ride-hail-territory-in-us-now-even-larger/
191
ideas”. With a sophisticated use of mixed metaphor, he added: “We will move forward by rolling
up our sleeves and rolling back more regulations that are standing in the way of job growth.393
”
City profile
The Phoenix metro region is the census-defined urbanized area of Phoenix and Mesa in
Maricopa County, Arizona with a population of 4.85m. Of the seven cities that make up the metro
region, Phoenix has the largest population (1.6 million residents), followed by Mesa (at about half
a million) and Chandler, Scottsdale, Glendale, and Gilbert (about a quarter of a million residents
each) while Tempe has a little fewer.
The city of Phoenix is at the heart of the region and contrasts demographically with San
Francisco. The average household income in Phoenix is $93,870, compared to San Francisco’s
$126,187; and 21% of the adult population have a bachelor’s degree or higher compared to San
Francisco’s 60%. Housing costs are much lower too – Phoenix’s rents are the 52nd highest in the
country, while San Francisco comes in as the third highest394. Wages reflect these low costs and
there isn’t the intensity of pressure of high rental costs experienced by San Franciscans.
Unlike San Francisco, which has the same political party represented at state and city
level, for a long time Phoenix was a swing city but recently trended Democrat. Phoenix’s state
legislature is dominated by Republicans and its seat is within this Democrat-leaning city.
Phoenix is experiencing continued population growth in contrast to San Francisco which
has been declining. Indeed Phoenix has been the fastest growing city in the US over the past
decade according to the 2020 census. Its history shows that it was a late developer - its
393 Harris, M. (2018, March 28). Exclusive: Arizona governor and Uber kept self-driving program secret, emails
reveal. The Guardian. https://www.theguardian.com/technology/2018/mar/28/uber-arizona-secret-self-drivingprogram-governor-doug-ducey
394 Zumper National Rent Report (2024, November 21). Zumper. https://www.zumper.com/blog/rental-price-data/
192
population expanded rapidly in the 1950s as car culture took hold of the US and has seen
continued growth since, although easing off in the last couple of decades.
Historically, the economy of Phoenix was based on farming and natural resources
including cotton and mining. Today the economy is dominated by education and health services
with construction a rather distant third395. Phoenix aspires to develop it’s tech sector, working on a
general plan update that moves it towards being a ‘tech-forward city’396
. According to the
Greater Phoenix Economic Council, jobs in the tech industry increased more than 15,000 from
2016-21, an 18.3% growth rate. The change has been the result of an effort by successive state
governors to diversify the region’s economy and improve its resiliency397
.
Transportation history, assets and culture
In contrast to the tightly-packed city in the Bay Area, Phoenix sprawls across 517 square
miles and has a relatively flat, twentieth-century street pattern - a classic grid pattern northsouth and east-west. The streets are generally uniform with major arterial streets one mile apart
and sub-divided into smaller blocks approximately every 1⁄8 mile. The roads are wide, built for the
car rather than for pedestrians. The metro area was ranked 23 out of 101 largest US Metro areas
in terms of the annual average fatalities per 100,000 people between 2018 and 2022. San
Francisco – Oakland – Berkeley ranks 63 having nearly half the number of fatalities per 100,000398
.
395 U.S. BLS (2024, May 6). Phoenix Area Economic Summary. US Bureau of Labor Statistics.
https://www.bls.gov/regions/west/summary/blssummary_phoenix.pdf
396 PLANPHX Summit (2023, December 14) https://www.phoenix.gov/pddsite/Documents/121423_GP%20-
%20Tech-Forward%20summit%20combined.pdf#search=autonomous
397 GPEC (2023, 13 March) Ambassador Event: Greater Phoenix Tech Story. Greater Phoenix Economic Council.
https://www.gpec.org/blog/ambassador-event-greater-phoenix-tech-story/
398 Smart Growth America (2022) Ranking the Most Dangerous Metro Areas.
https://smartgrowthamerica.org/dangerous-by-design/
193
Phoenix likes its roads. It prides itself with having a high quality road network and bridges
and suffers from less congestion than many cities. Inrix places Phoenix 31st in their list of most
congested US cities in the 2022 Global Traffic Scorecard while San Francisco is in position 7.
Phoenix has a history of keenly supporting road-building, which has benefited from continued
local rather than federal funding – a 0.5c general sales tax approved in 1985 that continues to
fund the expansion of the freeway system399
.
Built out during the mid-twentieth century, the city of Phoenix has not inherited the strong
transit system enjoyed by San Franciscans. Valley Metro is the regional public transportation
agency providing transit services to residents of metro Phoenix. Services include regional local,
express and rapid bus services, a Streetcar service, paratransit and light rail. The latter began
operations only in 2008 with 20 miles of track – eight more miles were added between 2015 and
2019. Valley Metro opened Streetcar in Tempe in May 2022, a hybrid battery and off-wire rail
service. Lines continue to expand with a goal of completing a 50-mile rail system by 2030400
.
While the transit system is growing it’s not a well-used system; census data from 2019 show the
metro area to have the third lowest percentage of workers who use public transportation out of
the 25 largest US metro areas401 but can be expected to grow as the network expands.
Unsurprising given the age and density of Phoenix, it has a strongly car dependent population.
Daily VMT is 209m miles – an interesting comparator to the 22m miles that Waymo has boasted
its AVs have driven so far402
.
399 Regional Development Plan, ‘Momentum 2050’ (2021) Maricopa Association of Governments
https://azmag.gov/Portals/0/Transportation/RTP/2022/RTP-Momentum-2050-v2.pdf
400 Valley Metro (2023, Spring) Valley Metro System Factsheet https://vulcanproduction.nyc3.cdn.digitaloceanspaces.com/pages/downloads/2300321051---fy22-agency-fact-sheetupdate_v1_ada_sb.pdf Retrieved 18 June 2024
401 Burrows, M., Burd, C. and McKenzie, B. (2021, April) Commuting by Public Transportation in the United States:
2019. American Community Survey Reports. US Census Bureau.
https://www.census.gov/content/dam/Census/library/publications/2021/acs/acs-48.pdf
402 Waymo (2024, September 5) New Data Hub Shows How Waymo Improves Road Safety. Waymo.
https://waymo.com/blog/2024/09/safety-data-hub/
194
The state hasn’t had a consistent history of supporting transportation technology, notably
for TNC services in the early years of their rollout. In April 2014, the then-governor Jan Brewer
vetoed legislation that would have enabled TNCs to avoid insurance regulations imposed on
traditional taxis. Uber and Lyft continued to operate, risking fines to do so. Once Ducey replaced
Brewer though he instructed officials not to pursue TNC drivers over taxi licensing rules. This
encouragement of technology development and loosening of regulations was a key feature of the
Ducey governorship.
Governance
The AZ Governor and State Government
As has been seen in the history of transportation technology, state government has taken
a leading role. This makes sense for strategic projects because it enables a seamless transition
between cities. Arizona state governors have been emphatically supportive of the AV technology,
largely because it is perceived to have an important role in driving economic growth in the state.
In an article for the New York Times in 2017403 Cecelia Kang described Governor Doug
Ducey’s approach to Uber and Lyft back in 2015, which was a harbinger of the approach to
transportation technology that was to come. Kang reports that in January 2015 Phoenix area was
about to host the Super Bowl and Ducey heard that a local regulator was planning to shut down
Uber and Lyft ride-hailing services for operating illegally. Ducey was furious, saying it was “the
exact opposite message we should have been sending,” and that, “We needed our message to
403 Kang, C. (2017, November 11). Where self-driving cars go to learn. New York Times
Company. https://www.nytimes.com/2017/11/11/technology/arizona-tech-industry-favorite-self-driving-hub.html
Retrieved 18 July, 2024
195
Uber, Lyft and other entrepreneurs in Silicon Valley to be that Arizona was open to new ideas.”
Ducey fired the regulator and - without the necessary time to weigh the decision or measure the
impact - shut down the entire agency, the Department of Weights and Measures. Three months
later, Arizona had legalized ride-sharing.
Regulation
The testing of AVs in Arizona is regulated mainly through Executive Order 2018-04, which
requires them to meet criteria prior to being operated or tested on public roads. Under the
Executive Order, autonomous vehicles are required to be: 1) in compliance with all applicable
federal motor safety standards; 2) able to achieve minimal risk conditions in the event of an
automated driving system failure; 3) capable of complying with all applicable traffic and motor
safety laws; and 4) in compliance with all applicable certificate, title registration, licensing and
insurance requirements404
.
As an early advocate for AVs, Governor Ducey had put his support into practice by signing
an Executive Order 2015-09 in 2015 “Self-Driving Vehicle Testing and Piloting in the State of
Arizona: Self-Driving Vehicle Oversight Committee” (EO 2015). An Arizona Self-Driving Vehicle
Oversight Committee was set up alongside the legislation to act as a forum for the Arizona DOT to
work with public safety and policy experts to decide how they could best operate in the state,
although there are doubts on how effectively it operated405
.
404 Arizona State Senate (2021, March 4) Fifty-Fifth Legislature, First Regular Session, Fact Sheet For H.B. 2813,
Autonomous Vehicles
405 Harris, M (2018, March 28) Exclusive: Arizona governor and Uber kept self-driving program secret, emails
reveal. The Guardian. https://www.theguardian.com/technology/2018/mar/28/uber-arizona-secret-self-drivingprogram-governor-doug-ducey Retrieved July 20,2024
196
It’s clear from these early days and the explicit framing of the EO that support was rooted
in growth, jobs and educational opportunities: “the State believes that the development of selfdriving vehicle technology will promote economic growth, bring new jobs, provide research
opportunities for the State’s academic institutions and their students and faculty, and allow the
State to host the emergence of new technologies”. Like transportation technologies of the past,
the Arizona state government structure has encouraged and enabled the development of AVs
driven by the perception of future economic growth and jobs.
Safety also featured in the EO however which notes, “the State has a shared vision that
the future of transportation and commerce relies on innovative technologies that could result in
more passenger and pedestrian safety, increase mobility options, and foster economic
productivity” (EO 2015, p. 5). Following EO of 2015, Uber and Waymo began testing in Phoenix
with the latter establishing a testing center in Chandler, a southeast suburb of the city. Most of
the rhetoric however was about economic competitiveness.
In late 2016, when Uber started testing self-driving cars in California without the required
permits, the DMV there said it would revoke the cars’ registrations so Uber pulled them out of the
state and shifted instead to Arizona406. Governor Doug Ducey was delighted to have snaffled a big
tech company from Silicon Valley, tweeting, ‘This is what over-regulation looks like
#ditchcalifornia’ and ‘Here in AZ we WELCOME this kind of technology & innovation
#ditchcalifornia’ #AZmeansBIZ’. He proclaimed, "Arizona welcomes Uber self-driving cars with
open arms and wide open roads," and that, "While California puts the brakes on innovation and
change with more bureaucracy and more regulation, Arizona is paving the way for new technology
406 Said, C (2018, March 27) Uber puts the brakes on testing robot cars in California after Arizona fatality. San
Francisco Chronicle.
https://www.sfchronicle.com/business/article/Uber-pulls-out-of-all-self-driving-car-testing-in-12785490.php
197
and new businesses"407. This was fun for the politics of the situation as Ducey worked to burnish
his anti-regulation credentials but there was a deeper motive – Ducey was keen to diversify the
state’s economy and encourage growth in the tech sector.
While there can be seen to be a growing ecosystem around tech growth there appears to
have been dynamics that are not about mobility but about horse-trading. McCarroll and Cugurullo
state that in 2017, Governor Ducey awarded a no-bid contract worth over $24 million to Waymo’s
parent company Google to provide the state government with new email and communication
accounts408. Such arrangements are indicative of what Aoyama and Leon have described as
partnerships between the public, private and non-profit sectors409. Such partnerships speak more
to government making deals that relax regulation in exchange for jobs and the prestige of the
growing tech economy.
Unbeknown to the people of Phoenix, Uber began testing its AV technology in the city in
August 2016, according to emails between the then state Governor, Doug Ducey, and Uber
obtained by The Guardian newspaper through public record requests410. From these emails, The
Guardian reported that in June 2015, Uber opened a customer support center in Phoenix bringing
300 jobs to the city and in August, Ducey held a joint press conference with Uber to announce the
latter’s $25,000 gift to the University of Arizona’s College of Optical Sciences. Uber committed to
basing a fleet of mapping cars at the university, and collaborating with academics on laser407 Neuman, S. (2018, March 20) Arizona Governor Helped Make State 'Wild West' For Driverless Cars. NPR.
https://www.npr.org/sections/thetwo-way/2018/03/20/595115055/arizona-governor-helped-make-state-wildwest-for-driverless-cars Retrieved 20 April 2024
408 McCarroll C and Cugurullo F (2022) No city on the horizon: Autonomous cars, artificial intelligence, and the
absence of urbanism. Front. Sustain. Cities 4:937933. doi: 10.3389/frsc.2022.937933 p03
409 Aoyama, Y., & Alvarez Leon, L. F. (2021). Urban governance and autonomous vehicles. Cities, 119, 103410-.
https://doi.org/10.1016/j.cities.2021.103410
410 Harris, M (2018, March 28) The Guardian, https://www.theguardian.com/technology/2018/mar/28/uberarizona-secret-self-driving-program-governor-doug-ducey
198
ranging lidars. The same day, the Guardian reported, Ducey issued an executive order clearing the
way for the public testing and operation of autonomous cars.
In March 2018, Executive Order 2018-04, “Advancing Autonomous Vehicle Testing and
Operating: Prioritizing Public Safety”, enabled further AV development, including testing without a
driver, much of which was codified into state law by HB 2813. The framing of EO 2018-04
reinforced the thinking behind the technology development, referring to the economic gains
brought to the state: “the business friendly and low regulatory environment has led to increased
investment and economic development throughout the state”.
The EO of 2018 also set out safety standards for testing driverless vehicles, instructing the
Arizona Department of Public Safety to issue a law enforcement interaction guide and required AV
companies to notify it at the start of driverless testing. The Self-Driving Vehicle Oversight
Committee was tasked to find out “how best to advance the testing and operation of self-driving
vehicles on public roads” (EO 2015-09, p. 11) reflecting its lack of neutrality and the historic role
of government as incentivizer and enabler of transportation technology. It has since been replaced
with a public-private partnership with more clout, the Institute of Automated Mobility (IAM)
which serves as an advisory board under the Commerce Authority.
In 2018, Chandler adapted its zoning code to accommodate AVs to encourage new
developments to plan for drop-off and pick-up areas for people using AVs or TNCs; a change that
was recommended by the American Planning Association’s guidance on AVs, published that
year411. Alongside this change a relaxation of a development’s parking requirement could be
reduced by up to 40 percent412
.
411 APA Jeremy Crute, J., Riggs, W., Chapin, T. and Stevens, L. (2018) Planning for Autonomous Mobility. American
Planning Association. PAS Report 592.
412 MAG (2019) Strategic Transportation Safety Plan (updated). Maricopa Association of Governments.
https://azmag.gov/Portals/0/Documents/Transportation/Task-7_Review-Safety-Implications-of-EmergingTechnologies.pdf
199
The AZDOT set out the timeline of the legislation413:
● 2015 — Governor Doug Ducey signs an Executive Order 2015-09 outlining Arizona’s
process for the safe development and testing of autonomous and connected vehicle
technologies.
● 2017 — HB 2159 passes allowing for the demonstration of truck platooning technology on
Arizona's highways.
● 2018 — In March, the Governor issued Executive Order 2018-04 to reflect advancements in
technology and testing.
● 2018 — HB 2422 passes allowing for “personal delivery devices” to operate.
● 2019 — HB 2132 passes allowing for the operation of “personal mobile cargo carrying
devices.”
● 2021 - HB 2813 passes codifying much of Executive Order 2018-04 into state law.
● Title 28 Chapter 32 (Arizona Revised Statutes)
● 2022 - SB 1333 passes that creates a new vehicle type of Neighborhood Occupantless
Electric Vehicles (NOEV). These goods delivery vehicles are low speed vehicles with no
passengers or driver.
By 2022 there had been three bills passed by the Arizona State legislature and signed by the
Governor. House Bill 2813 (2021), set the legislative framework for testing on public roads and
two further bills the following year were passed enabling driverless testing (SB 1333) and
regulations enabling robotaxis to operate (2273) that meet vehicle safety and state emission
standards.
413 ADOT (n.d.) Autonomous Vehicles Testing and Operating in the State of Arizona. Arizona Department of
Transportation. https://azdot.gov/mvd/services/professional-services/autonomous-vehicles-testing-andoperating-state-arizona Retrieved June 27, 2024
200
Arizona Department of Transport (ADoT)
The Arizona Department of Transport has regulatory and authorization responsibility for
AVs. Unlike California, the Public Utilities Commission has no responsibility or involvement; there
is no mention of AVs on their website and this was confirmed with me in my interview with
Amanda Kindle, Stakeholder Manager for ADoT. Kindle’s stakeholders include the Secretary of
State, the Governor’s Office and law enforcement – and her role entailed mediating between
them. She outlined the process, noting there was no formal paid-for permitting process, instead
there was an initial meeting with the company involved and the need for a self-certification letter
saying that it will meet safety standards and comply with the rules of the road. If the AV company
wanted to test without a driver then they would have to present a law enforcement interaction
plan too. AV testing is a simple, straightforward process.
The Arizona DoT has had a hands-off approach to regulation. Under Ducey, the Director of
the ADoT, John Halikowski, demonstrated the Department’s approach that aligned with that of
the Governor, “We shouldn’t be getting in the way by prescribing regulations”, he said, “when we
really don’t know how the equipment will perform414”. This is echoed at the city level too – Kang
notes of the start of AV testing in her city, “Chandler’s officials sealed the deal with further
promises to stay out of the way. ‘We don’t think it’s our role to be engaged in regulation’, said
Micah Miranda, Chandler’s director of economic development.415”.
414 Kang, C. (2017, 11 November). Where self-driving cars go to learn. New York Times
Company. https://www.nytimes.com/2017/11/11/technology/arizona-tech-industry-favorite-self-drivinghub.html
415 Kang, C. (2017, 11 November). Where self-driving cars go to learn. New York Times
Company. https://www.nytimes.com/2017/11/11/technology/arizona-tech-industry-favorite-self-drivinghub.html
201
DMV and other agencies
The Maricopa Association of Governments (MAG) is the designated metropolitan planning
organization (MPO) and council of governments (COG) for transportation planning in the Phoenix
metro area. It provides a forum for local governments to work together on regional-level issues.
Their strategy document, ‘Regional Development Plan, Momentum 2050’ serves as a long-term
blueprint for the metro area, guiding investments into its transportation system. The Plan
mentions AVs only once to say, “The emergence of electric, connected and autonomous vehicles
and technologies will continue to change how people make travel decisions”416. Interestingly, it
doesn’t mention them under the chapter on safety – as with San Francisco a city is not able to
factor-in safety improvements on the basis of the unproven evidence so far. This reflects the
various pronouncements and EOs – AVs are seen as an economic tool, not a mobility tool.
The Arizona Commerce Authority (ACA)
The Arizona Commerce Authority (ACA), is the state’s leading economic development
organization. It leads on advancing an economic development strategy focused on the idea that
“while technological advances are the keys to prosperity, the support of an innovation ecosystem
that includes private industry, state government and academia, is crucial”417. In practical terms
they act as the starting point for a business looking to set up in the state and for AV companies
they provide a welcome and can help with practical issues including introducing them to city
councils and helping to iron-out any wrinkles that might get in the way. It is notable that the
advance party comes from economic development and not mobility.
416 MAG (2021, December 1) Regional Transportation Plan: Momentum 2050. Maricopa Association of
Government. https://azmag.gov/Portals/0/Transportation/RTP/2022/RTP-Momentum-2050-v2.pdf
417 HBR (2018, February 15) How Arizona Is Leading the Autonomous Vehicle Revolution (Sponsored content from
Arizona Commerce Authority). Harvard Business Review. https://hbr.org/sponsored/2018/02/how-arizona-isleading-the-autonomous-vehicle-revolution Retrieved June 28, 2024
202
The ACA act as a facilitator and problem-solver and works in an accommodating way
with AV companies with the understanding that “Where they test, they invest418”, as described in
an interview with Marisa Walker, Executive Vice President, Infrastructure of the ACA. The ACA is
an important business voice providing very practical support to tech companies such as on
workforce and footprint issues and incentives for businesses to grow. It reviews applications from
AV companies, organizes discussions with them about the support they need and helps connect
them to city councils where they want to test.
The Commerce Authority makes it clear why it supports the industry. Their website points
to familiar silver bullets - that AVs “promise to make transportation safer, cheaper and faster. They
spark the opportunity to reduce road congestion and fuel consumption and ultimately increase
productivity… Automated vehicles also pave the way for economic opportunity. The “passenger
economy” represents a $7 trillion global opportunity by 2050 and by 2023, companies will
spend $255 billion in capital expenditures for electronic vehicle R&D. This large-scale market and
societal disruption demands industry-driven validation of safety standards, policies and technology
solutions”.
Walker is also Executive Director of the Institute of Automated Mobility, leading work of
this public private partnership that collaborates ‘to help prepare the state and the region and our
communities around preparing for advanced mobility’419
.
418 Brodie, M. (2023, August 22) Arizona is a hub for driverless cars. Here's why — and what's next for autonomous
vehicles. KJZZ. https://www.kjzz.org/2023-08-22/content-1855481-arizona-hub-driverless-cars-heres-why-andwhats-next-autonomous-vehicles
419 Arizona Commerce Authority (2024, September) YouTube. www.youtube.com/watch?v=9VguYqbza6S
203
City Experience of AVs
Twelve companies have submitted applications to test autonomous vehicles in the state420
with Cruise and Waymo having had the most vehicles and both operating without a driver. Cruise
had provided ride-hailing services and a food delivery service in Scottsdale and had partnered
with Walmart for self-driving delivery in Chandler. Waymo operates across Greater Phoenix
including the suburbs of Chandler, Gilbert, Mesa, Phoenix, and Tempe. Before the Covid outbreak,
Waymo’s robotaxis were offering riders between 1,000 and 2,000 rides a week, with 5-10 per
cent being driverless421. The maps of the areas of service for Waymo and Cruise respectively are
shown in Figures 12 and 13 below.
Figure 10: Waymo’s operating area in Phoenix Metro. Source: Waymo422
420 ADOT (n.d.) Autonomous Vehicles Testing and Operating in the State of Arizona. Arizona Department of
Transportation. https://azdot.gov/mvd/services/professional-services/autonomous-vehicles-testing-andoperating-state-arizona
421 FT (2020, July 19) Self-driving Industry takes to the Highway after Robotaxi Failure. The Financial Times Ltd.
https://www.ft.com/content/96d3eeff-7f52-46e3-a8a8-aeb668472034
422 Waymo (2024) Where you can go. Waymo. https://waymo.com/waymo-one-phoenix/
204
Figure 11: Cruise’s Operating area in Phoenix Metro in 2023. Source: Cruise423
Of course the most significant AV story in Phoenix was the first AV death – that of Elaine
Herzberg on March 18, 2018 in Tempe. She was hit by an Uber test vehicle with a distracted
safety driver at the wheel while the vehicle was in autonomous mode. As with the Cruise dragging
incident of October 2023, finger-pointing the bad actors only emerged after a serious incident. On
20 March 2018, NPR reported that the non-profit group Consumer Watchdog described Arizona as
423 Vogt, K. @kvogt (2023, August 8) As we deploy new foundational AV capabilities, geofences will exponentially
expand. As an example, today our Phoenix service area expands by 20x…. X.
https://x.com/kvogt/status/1688943208125149184
205
"the wild west of robot car testing.", saying that there is "no regulation in place," in the state.
"That's why Uber and Waymo test there. When there's no sheriff in town, people get killed.424" It
took the Governor, who’d been instrumental in allowing a relaxed regulatory regime, eight days to
rescind Uber’s permit to test.
Reporting on the crash, the National Transportation Safety Board criticized the state for its
lax safety regime: “Arizona’s lack of a safety-focused application-approval process for
automated driving system (ADS) testing at the time of the crash, and its inaction in developing
such a process since the crash, demonstrate the state’s shortcomings in improving the safety of
ADS testing and safeguarding the public”425. The NTSB went on to recommend that all states
require developers to submit an application that, at minimum, outline how they plan to
incorporate risk management into their testing.
Although much of the testing has involved robotaxis, Phoenix has also experimented with
AV technology as a complement to transit. In 2018 Valley Metro, the city’s public transit agency,
went into partnership with Waymo to test a first mile-last mile service to transit stops and park
and ride lots with the stated aim of enabling better transit access for people who don’t live close
enough to reasonably walk to transit. Phase I of the pilot program was made available exclusively
to Valley Metro employees residing in Chandler, facilitated by Waymo's mobile app. Building upon
Phase I feedback, Phase II targeted underserved communities enrolled in Valley Metro's
RideChoice program which offered subsidized curb-to-curb ride-hailing for mobility
disadvantaged people, including people over 65 years of age, people with disabilities, or those
with low incomes. Feedback from Phase II (carried out by the University of Arizona and sponsored
424 Neuman, S (March 20, 2018) Arizona Governor Helped Make State 'Wild West' For Driverless Cars. NPR.
https://www.npr.org/sections/thetwo-way/2018/03/20/595115055/arizona-governor-helped-make-state-wildwest-for-driverless-cars Retrieved 20 April 2024
425 National Transportation Safety Board (2019, November 19) Accident Report NTSB/ HAR-19/03: 58 NTSB
https://www.ntsb.gov/investigations/AccidentReports/Reports/HAR1903.pdf
206
by the DoT) found a preference for the AV service over alternative RideChoice options of taxis,
buses, paratransit, and transportation networking companies. Research, led by Stopher et al also
found a positive perception of robotaxis replacing personal vehicles, especially among older
people who were unable to drive.
Stopher et al found AV rides had higher ratings than conventional ride-hailing rides for
cost, comfort, convenience, and travel and wait times in Phoenix426. The feedback noted, “Key
findings were that participants felt safe, found the AV services more convenient than typical
RideChoice options, and engaged in more out-of-home activities (i.e., made new trips) as a result
of the AV option. Participants indicated a willingness to ride alone in AVs and to ride with family
or friends. Riding with strangers in an AV mobility future was the least desirable option. Their
ratings of wait time, travel time, convenience, and comfort of the AV option were in all cases
higher than for traditional options available through RideChoice. A majority of participants
expressed positive feelings about the introduction of AVs, both for RideChoice services and more
generally on the roads.427” What this reveals isn’t very illuminating - people prefer not to share,
they like not having to wait, and put this together they are likely to make more trips. It’s
important to note that a safety driver was on board so the issues of trust and vulnerability - and
indeed traveling alone – are not fully taken into account.
The current Phoenix mayor, Kate Gallego, shares the former governor’s enthusiasm for the
technology, reflecting her commitment to driving economic growth too. When announcing the
Waymo expansion to include Phoenix Sky Harbor Airport, for example, she pledged her support for
the technology with a positivity that the San Francisco mayor had not expressed. “Phoenix leads
426 Stopher, P.R., Magassy, T.B., Pendyala, R., McAslan, D., Arevalo, F.N., Miller, T. (2021) An evaluation of the
Valley Metro-Waymo automated vehicle RideChoice mobility on demand demonstration. (FTA Report No. 0198)
Federal Transit Administration, United States Department of Transportation.
427 Federal Transit Administration (2021) An Evaluation of the Valley Metro–Waymo Automated Vehicle RideChoice
Mobility on Demand Demonstration, Final Report. US Department of Transportation.
https://www.transit.dot.gov/sites/fta.dot.gov/files/2021-09/FTA-Report-No-0198%20REVISED.pdf
207
the nation in demonstrating autonomous vehicle technology, science, and safety," said Mayor
Gallego, the “future of travel is here, and Waymo One service to Phoenix Sky Harbor International
Airport marks our city's commitment to innovation and technological advancements that will
impact the world."428 The service to and from the airport is an important one as airport rides
generate much money and there’s a prestige to being allowed to do so. The rides are an
important commercial asset. At the state level too the Ducey approach of more relaxed regulation
has continued: Arizona did not require Cruise to stop their trials after the October 2023 San
Francisco dragging incident.
Safety
Theoretically, the safety of one AV is the same as the safety of the rest of the fleet
because they all learn as one. What differs is the environment in which they operate and the
scenarios and situations that this presents; it’s here that Phoenix differs markedly from San
Francisco. While it’s still too early to say how safe AVs are, we do know that they haven’t caused
as many incidents as in San Francisco.
Speaking with the reporter on the Arizona Republic, Corina Vanek, in December 2023, she
said that she didn’t know of any serious incidents apart from the death of Elaine Herzberg five
years before – the trial of which she had reported on for her paper. She told me she had seen
public records that showed no concerning incidents involving police and firefighters and had had
this confirmed with the relevant departments. Vanek pointed to the differences in road width – as
well as number of pedestrians, density – that might account for the difference in experience in
428 City of Phoenix (2022, November 1) Waymo Autonomous Vehicles Arrive at PHX — Phoenix Sky Harbor is OnTrack to be the First Airport in World to Offer Waymo Rider-Only Autonomous Vehicle Service. City of Phoenix.
https://www.phoenix.gov/newsroom/aviation/2548
208
San Francisco and Phoenix, noting that Phoenix drivers don’t stop for pedestrians as much as they
might in San Francisco because the road layout is so different and car-centric.
Vanek did know of crashes – two minor ones – from the police reports involving Waymos.
In an exchange some months after our interview (September 2024) she noted that there had been
more incidents since and that NHTSA had listed several that included Waymos colliding with
stationary objects, such as a pole or curb, and that rear-ending of Waymos is still the most
common crash they experience you can stay there. Arguably of course if you’re a rider in a
robotaxi the question of fault matters little compared to the extent of your injuries. When I asked,
“Are they safe?’ she responded that, “the data we've seen pretty overwhelmingly says ‘yes’.
Looking at the data does show they're safe”, she said.
Vanek pointed to the attention heuristic – that it was natural that because they’re new
that people won’t trust them because they don’t understand them. She said people would have a
different view if they were used to them, as she was. “I don’t see fear when people are used to
them. It’s a matter of adjustment”, she said and saw no reason for people to be more cautious
than with human drivers.
When I asked why Phoenix was supporting the testing she pointed to Governor Ducey’s
approach of being welcoming to different types of technology. She noted too the Valley Metro
transit trial. I asked whether she thought people’s behavior might need to change if AVs became
more commonplace (noting the inability to make eye contact with a driver when you wish to cross
a road) and she thought it ‘not unreasonable for pedestrians to be extra wary crossing in front of
an AV’ given the inability to communicate with a driver but added that she thought the
technology is developing to understand gestures. As with the reporter in San Francisco there is
much optimism and belief in the tech companies' pitch on the part of the local media.
The confident pitch of the AV companies appears to diminish up close. Alicia Radatz’s
research at Arizona State University focused on public meetings, events and interviews with
209
people more directly involved in the trials. She noted from the public meetings that by far, the
most common and fully developed purpose for the technology that was conveyed in these public
forums was safety which was advanced by commenting on the unsatisfactory status quo and
suggesting autonomous vehicles were the solution. She contrasts this though with the more
private conversations, “In interviews, there was a rampant acknowledgment that there is a lot of
work needed to make autonomous vehicles safe, whereas in most public events, there was
general abstract optimism about autonomous vehicle futures and the improvements that they
bring to the status quo429.”
Like San Francisco and the national messaging about AVs, in Phoenix the focus has been
on safety when it’s a public-facing audience. Radatz continues, “When discussing social futures, a
safer future was the predominant vision for the future across interviews and events. The power of
safety as the propelling value for the technology was conveyed consistently throughout public
events, but was not consistently the focus of every conversation.430” At the governance level
though the message in Phoenix has been consistently a dominating theme of strengthening the
local economy. Like most public affairs professionals you adapt the message depending on the
audience.
The positive approach to AV companies has framed the relationship with first responders
in Arizona. Whereas in San Francisco the relationship is marked by animosity, Phoenix brings
together Arizona public safety as a partner of the IAM together with the other partners including
the Maricopa Association of Government, Maricopa County government, Arizona State University,
Northern Arizona University, ADOT, intel, and Cox Automotive Mobility. Arizona’s emergency
429 Radatz, A (2021) Living in the Arizona Test Bed. PhD Dissertation, Arizona State University p107
430 Radatz, A (2021) Living in the Arizona Test Bed. PhD Dissertation, Arizona State University p109
210
responders also collaborated with Waymo in 2019 to produce an instructional video on how their
profession should interact with Waymo AVs431
.
The Economy
This chapter has already discussed in detail the way in which Arizona has supported AVs
for economic reasons. The Executive Order of 2015 goes straight to the heart of the priority of
Arizona in encouraging AVs: “the state believes that the development of self-driving vehicle
technology will promote economic growth, bring new jobs, provide research opportunities for the
state’s academic institutions and their students and faculty, and allow the state to host the
emergence of new technologies.”
As noted, The Guardian newspaper has reported on emails from 2007-2015. They also
revealed the relationship between the then-Governor, Doug Ducey, and Uber which appeared
close and verging on the inappropriate. The Guardian reported that the emails “reveal how Uber
offered workspace for Ducey’s staff in San Francisco, praised the governor lavishly, and promised
to bring money and jobs to his state. Ducey, meanwhile, helped Uber deal with other officials in
Arizona, issued decrees that were friendly to the company, tweeted out an advert at the
company’s request, and even seems to have been open to wearing an Uber T-shirt at an official
event432”. Some of the activity may be seen as par for the course of public affairs but sponsorship
of politicians crosses a line.
431 Waymo (2019) Waymo First Responders Training. Waymo. www.youtube.com/watch?v=7G0a1PRAtaM&t=79s
Accessed December 29, 2024
432 Harris, M. (2018, March 28) Exclusive: Arizona governor and Uber kept self-driving program secret, emails
reveal. The Guardian https://www.theguardian.com/technology/2018/mar/28/uber-arizona-secret-self-drivingprogram-governor-doug-ducey Retrieved December 22, 2024
211
Looking at the city level, the view is echoed by Chandler, the first city in Phoenix metro to
host AVs. McAslan et al report on an interview with the city where they speak directly to the goal
of economic development: “we’re looking at it from an economic development perspective. How
can our support of this immature, growing industry, help advance the future economic
competitiveness of the city? Automotive tech is one of our target industry clusters, so it was a
natural fit to see how this can elevate Chandler’s reputation within the state”433
.
The Arizona Chamber Foundation is an educational and research organization that
supports the Arizona Chamber of Commerce & Industry to accelerate the priorities of the business
community. The Foundation hired consultants Rounds Consulting Group Inc, an Arizona-based
economic consulting firm, to look at the impact of autonomous vehicles on the state's economy:
they reported back in January 2020. Their report reveals why the state has continued to support
the AV market; their forecast is for tens of thousands of new jobs and hundreds of millions in tax
revenue. The report states: “By 2026, $6.1BN in estimated statewide R&D would create 39,000
direct jobs throughout the state and generate $4.3B in economic output. The spinoff (indirect and
induced) effects would produce an additional 35,000 jobs and $4.9B in economic output. State
and local governments would collect more than $350M in tax revenues from a conservative
assumed capture rate of the national market arriving in Arizona”434. The unequivocal forecast
makes it clear why business and the government have been supportive of the AV industry in
Arizona.
Speaking with the Director of Planning at Phoenix City Council, Joshua Bednarek, he
pointed to the contribution robotaxis were making to a changing narrative about the city. While in
433 McAslan, D., Najar Arevalo, F., King, D. A., & Miller, T. R. (2021). Pilot project purgatory? Assessing automated
vehicle pilot projects in U.S. cities. Humanities & Social Sciences Communications, 8(1), 1–16.
434 Rounds Consulting Group (2020, January) Economic Impacts Of Advancing Arizona’s Competitive Position In The
Autonomous Vehicle Industry (Prepared for Arizona Chamber Foundation). Rounds Consulting Group Inc.
https://www.azchamberfoundation.org/wp-content/uploads/2020/01/January-2020-Foundation-AV-ImpactStudy.pdf Retrieved 29 June, 2024
212
the past, he said, Phoenix was thought of as rather sleepy, suburban town for retirees, there’s a
changing narrative that says, “Phoenix is open to trying new and innovative things and has an
expanding tech and university system, you don’t have to drive, and there’s an expanding AV
network”.
It's early days as far as Bednarek is concerned and hard to tell the impact of the robotaxis
on the city but he didn’t anticipate the sort of infrastructure needs that historic transport changes
brought in the past, such as acquisition or demolishing of homes or road-widening that came with
freeways. “They’re seen as kind of fun, they’re still being discovered”, he said, “we want to allow
them to happen but we want to keep people safe”, he said.
Bednarek said that his department hadn’t been involved in the trials but are planning
around scenarios where AVs may be more prevalent. An important part of their thinking is around
the reduced need for parking with AVs which he thought was an exciting proposition. Noting that
the city council wants more homes built and that developers press for more flexibility in parking
minimum requirements they are looking favorably on loosening such requirements in anticipation
of the potential for future reduced need with AVs.
Asked what success looks like, Bednarek pointed to equity issues: “I don’t think the
robotaxis go to many transit dependent population areas just yet – and expanding there is where
it could be classified as a success”. An important question is, “Can they expand to areas and keep
cost low enough so that they are accessible to one or no-car household who can be provided with
an additional option to bikes and transit - so that car ownership isn’t a requirement”. There’s an
openness and curiosity about the possibilities in Phoenix as far as Bednarek is concerned, “how
they’ll change city shape is something that interests me”, he said, and that “some city engineers
talk of them as an inevitable part of the future – but not all”.
213
Congestion
The DoT manager I interviewed, Amanda Kindle, thought the impact of AVs on the city had
not been significantly negative or positive – and certainly from a congestion standpoint she
hadn’t seen a difference.
Public Response to AVs
In my interview with Corina Vanek, reporter on the Arizona Republic, in December 2023,
she came across as very knowledgeable about the AV trials and had been keeping track of
developments and public sentiment. She explained that although some people may feel
uncomfortable with AVs there hasn’t been any organized opposition or protest as had happened
in San Francisco. She did note the potential for job loss though pointing not just to TNC drivers but
also delivery drivers as Cruise had for a time been providing deliveries with their AVs.
In my interview with Amanda Kindle, the Relationship Manager for the DoT, she echoed
the cordial relationships between the AV companies and the community. She said she felt that the
public were either accepting of the technology or not really expressing an opinion. She said she’d
had the AV brief for nearly a year and that maybe only two or three residents had made contact
with her – and none of them had felt very strongly one way or another, she said. Asked about who
might benefit most, she suggested people with mobility problems – who can’t drive on their own;
the losers, she suggested, would be the section of the population who won’t embrace the
technology.
The city has had protests such as those reported in the New York Times on New Year’s Eve
2018 435 (well before Vanek started in the news team at the Arizona Republic). Here a list of
435 Romero, S. (2018, December 31) Wielding Rocks and Knives, Arizonans Attack Self-Driving Cars. New York
Times. https://www.nytimes.com/2018/12/31/us/waymo-self-driving-cars-arizona-attacks.html Retrieved
December 22, 2024.
214
attacks were outlined from police reports including that two dozen vehicles had their tires slashed,
someone waved a .22 caliber revolver at a safety driver, people pelting Waymo vehicles with
rocks while other drivers have repeatedly tried to run the vehicles off the road.
In total at least 21 incidents had been reported by the end of 2018 but the NYT reports, “In
some of their reports, police officers also said Waymo was often unwilling to provide video of the
attacks. In one case, a Waymo employee told the police they would need a warrant to obtain
video recorded by the company’s vehicles”. Waymo also refused to press charges against
perpetrators. This may be because they don’t want to be involved in the bureaucracy of reporting
or it may be that Waymo doesn’t want to publicize the attacks because it could encourage copycat incidents or put off their investors. The NYT suggested it was because they just didn’t want
the bad publicity. The company did what perhaps any public relations officer would advise, frame
a response in terms of the core message: safety. Waymo responded, “Safety is the core of
everything we do, which means that keeping our drivers, our riders, and the public safe is our top
priority,” Alexis Georgeson, the Waymo spokeswoman told the NYT, adding that, “Over the past
two years, we've found Arizonans to be welcoming and excited by the potential of this technology
to make our roads safer.436”
Vanek said that although some are very wary of AVs and some will say they’d never go in
one, she thought that that fear dissipates as people become more familiar with them, seeing them
around the city. I asked whether she thought it was inevitable that the technology would roll-out
and she said that she had heard that and indeed had recently been in a panel discussion with real
estate and economic development professionals who very much thought that AVs were inevitable
and that the industry in the city needed to adapt their development projects with them in mind –
436 Romero, S. (2018, December 31) Wielding Rocks and Knives, Arizonans Attack Self-Driving Cars. New York
Times. https://www.nytimes.com/2018/12/31/us/waymo-self-driving-cars-arizona-attacks.html Retrieved
December 22, 2024.
215
such as by loosening the need for parking. She thought that those who were more ‘tuned in’ to the
AV issues did think the technology was inevitable.
Amanda Kindle at the DoT said she too had heard people speak of AVs as inevitable –
even referring to a meeting she’d had the morning I interviewed her as an example. She said she
hears the sentiment from people in the industry as well as people socially. Again, seeing the
vehicles around the city was seen as an important driver in making people more accepting of
them”.
Winner and Losers
For now the winners in Phoenix are those who have benefitted from the economic
opportunities that have emerged – precisely as they were intended. An ecosystem has developed
between the university gaining research funding, the AV companies gaining a good testing ground,
the governance being associated with an industry that’s leading edge and brings investment and
jobs, and of course the people who take those jobs. If the people gaining the jobs are people from
out of state – incomers from Silicon Valley for example – then that’s not jobs for local talent, even
though they are likely to help the economy grow indirectly.
The Tech Reporter on the San Francisco Chronicle, who had just come back from a visit to
Phoenix, spoke of a hope people felt that AVs could help the city become more walkable. This
seems highly unlikely however because the default is private car ownership. Yes, the metro area is
investing in transit and yes they ran a pilot with Waymo to connect transit to riders – but there
seems little reason to believe the culture will change.
As far as the Valley Metro experiment is concerned, finding a way for city councils to offload the cost of Dial-a-Ride services is appealing – and there have been plenty of champions
from disability groups who have welcomed AVs. But the Valley Metro survey identified a barrier -
216
the need for assistance in boarding and alighting and in loading and unloading packages, which
services with no human present would be unable to assist. There is a risk that public services take
advantage of the support from some quarters to cut back on the expense of Dial-a-Ride; the
losers would be those who need physical assistance getting in and out of a vehicle and with
baggage.
Conclusion
The AV experience in Phoenix that it is led by the economy. This may bode well for jobs,
for growth and for the prestige that comes with being associated with the tech industry but it has
implications elsewhere. By not looking more strategically, Arizona is neglecting the region’s
transportation system, especially the fledgling transit system. It’s a system that is being invested
in and expanded but it won’t be able to draw in the ridership needed to change the narrative of
the region being car-centric. Making transit appealing is only half the equation; the other half is
making it harder to drive.
Unlike San Francisco it seems the state government and the city governments are all
pulling in the same direction – pro-AVs as pro-growth. Their close and supportive relationship
with the AV companies and others in the AV ecosystem (including ASU in particular) help create a
smooth ride for AVs in the state.
While there may be talk of safety to a public-facing audience, AVs in Phoenix hasn’t had
to be a hard-sell, especially with an ecosystem that’s self-supporting and with few detractors.
As in the history of transportation technology there’s a complex ecosystem that is working
hard to drive AV development forward. There is little sense of the public shaping how the AVs are
used and although there was a pilot using AVs for transit first and last mile this doesn’t not seem
217
to have been a very illuminating case study. It reflects the early ‘trials’ of using TNCs to connect
people to transit. It’s great public relations but no private sector organization will pursue a line of
business in the long-term if it can make a stronger profit from another.
Like much of the history of transportation technology, consumers respond positively to
something that gets them somewhere faster or in more comfort – and ideally both; and
manufacturers will shape the technology around those demands. In this way, there’s an
interpretative flexibility now whereby use cases will shape the final AV product and it’s not certain
yet whether that will be a robotaxi, a pure AV, a dual-mode vehicle that could be conventionally
driven in cities (to get ahead faster) and autonomous vehicles for long journeys. Given that GM
has been clear that its endgame is private ownership and given that GM is ending Cruise’s
robotaxi business it suggests this won’t be the use case in the long-term.
218
Chapter 7: Analysis
‘Science and technology revolutionize our lives, but memory, tradition and myth frame our
response. Expelled from individual consciousness by the rush of change, history finds its revenge
by stamping the collective unconscious with habits, values, expectations, dreams’.
Arthur Schlesinger437
This chapter will draw together the previous chapters on the history of transportation
technologies (including that of autonomous vehicles) with the findings from the case studies. It
will make clear the role of technological determinism (it’s a promotional tool), why early use
cases should not be assumed to be the final transportational artifact once closure has occurred,
and the reasons why the ‘silver bullets’ messages do not appear sincere. Those same promotional
messages have been used in the history of transportation technologies, the same people involved
in its development (engineers, politicians, and the auto ecosystem), and so it seems quite possible
we should expect the same outcomes.
Motives other than for mobility
The findings from the case studies provide a more micro-level picture of the current AV
trials. The AV trials in the Phoenix metro area have played out very differently to those in San
437 By Arthur M. Schlesinger Jr. (1986, July 27) The challenge of change. New York Times (1923) Retrieved from
http://libproxy.usc.edu/login?url=https://www.proquest.com/historical-newspapers/challengechange/docview/111108536/se-2
219
Francisco, having a much smoother ride in the former. Although the case study cities and their
experiences are very different they align with the overall picture of AV development in the US. It is
the result of a combination of actors working together within an ecosystem that drives forward a
program of development and the rollout of the technology. This ecosystem is composed notably of
tech and auto companies, their suppliers, the infrastructure industry, and politicians – the same
ecosystem that has produced road transportation over the past century and beyond. There are
huge amounts of money being invested and indeed many jobs created in the development of the
technology which creates a wider pool of champions hoping to find employment. This has
heightened the hype and helped some see AVs as inevitable.
We can see from the history of transportation technologies that consumers will buy into
those transportation technologies that make their travel easier. The history also shows that the
growing ecosystem around automobility will respond to these customer wants and politicians and
businesses will support this for economic and political reasons.
This research has examined the processes and consequences of historic transportation
technological developments and compared them to those of AVs today. What we see in the past
is politically-motivated and politically-led changes that are inspired by and supported by
engineers and the car industry. That political motivation can come from a desire to gain the
support of businesses, to attract jobs or develop a perception of a strengthening local economy or
to more directly advance political careers.
Naturally tech and auto companies want to make money, and politicians want evidence of
delivering on their electoral commitments and on the needs and wants of the public. But
politicians work in electoral cycles and can be swayed towards what looks or sounds good rather
than what is good for cities in the long-term. If we treat AVs as an addition to mobility options
rather than a political device or engineering problem then the cast of players involved in
developing them would look different and transit would feature more prominently.
220
A notable contrast between San Francisco and Phoenix is that with the latter, there are
shared values and assumptions at the city and state level: both are in agreement on the value of
AV testing and both see its positive economic benefits. The approach to AVs is led by an economic
rather than transportation perspective; this comes through in the legislation, the public messaging
from the governments and their agencies and from my interviewees. AVs are not a mobility
solution for Arizona, they are a complement to the very active role of government to create a
thriving economy and bring jobs to Arizonians, prospering from the trickle-down monetary
benefits and – at least to some extent – scoring points over their perceived rival for tech dollars
and jobs, California.
The use of technological determinism as a promotional strategy
Autonomous vehicles (AVs) have been described as having the potential to have a massive
impact on cities and a silver bullet to cure a number of urban problems. The narrative around their
development has drawn on technological determinist theory – as if their arrival is unstoppable.
Such narratives disempower the public sector, governments and the public by implying a force
greater than they will dictate the future of mobility. Even though there is clear evidence of societal
power in transportation technology, the narrative seems to have an insidious intent. Technological
determinism has been used to sap the power of opposition to AVs but it's only a marketing tool
and does not describe the true dynamic at work. What we see in the history of transportation
technologies - including that of AVs (and in the case studies here) - is that they are subject to a
coevolutionary process.
221
Coevolution and its impacts
Coevolution theory sees government as having an involved and supportive role –
establishing laws, policies and standards; funding research; building and maintaining necessary
infrastructure; promoting education and skill development to support the technology; and to
address social and ethical implications. An ecosystem of interests help drive forward a technology
and in the case of transportation technologies this has included auto companies and the public.
Klein and Pinch talk of interpretive flexibility - the early stage in the development of technology
where it could take on different forms. While the use cases for AVs are for robotaxis currently, we
know from history that the early ‘use cases’ are not necessarily how the technology develops.
What eventually emerges will depend on the players involved and the receptiveness of a consumer
market. Car companies exist to sell cars. Tech companies exist to sell tech. There may be other
objectives on the way but if they conflict with their primary purpose then they will not be pursued.
Robotaxis can be seen as a Trojan Horse for the technology. They have been an
acceptable way of allowing the technology to develop and finding the most profitable use case.
It’s clear that what’s profitable is to take the driver out of the cost base – and this is part of the
business case for robotaxis and for future use of AVs as delivery vehicles. The impacts of the latter
on lower income people was something I raised in my interview with a Teamster representative in
San Francisco. Yes, this would affect jobs but it could have a significant impact on VMT, livability
and downtowns. Deliveries would be cheaper making it less attractive to use local shops and
businesses although it does require cooperation at the delivery end to ensure a human can receive
the goods. For deliveries, it won’t matter too much if there’s congestion delay as there would be
no person the vehicle to be inconvenienced (although the good recipient may be) but it would
have an impact on those sharing the road who suffer from the added congestion. For riders in
robotaxis though they would suffer the inconvenience shared by other road users - and perhaps
222
more so if AVs continue to be programmed to be more of a stickler for road rules and more
courteous than human drivers.
Because consumers have been seen to have a role in the development of other
transportation technologies (such as the car, road-building and TNCs) we should assume they
will do so with AVs too as markets respond to the needs and wants of consumers. In the past
we’ve seen that it is ease and convenience that has informed consumer shaping of transportation
technology, so we should expect the same with AVs. A key question is, ‘what will make my trip
easier and more convenient’, and that is likely to mean a continuation of private ownership – and
that will mean private ownership of AVs. Indeed the GM closure of Cruise’s robotaxi development
in December 2024 can be seen as a step towards that goal of private ownership for a company
that is the largest automaker in the United States. Research too reflects that private ownership of
AVs will prevail after a transition period438 and it has been made explicit in statements from Mary
Barra, Chair and Chief Executive of GM. However, there is a possibility that even if the technology
does develop fully that people might find AVs too slow or too expensive and they remain confined
to the robotaxi market where the use case and trade-offs (labor cost savings) are more clear.
GM CEO, Barra, speaking about how the future might look in 20 years at the SXSW
Conference in 2023, talked of ‘choice’ – implying a choice of AVs as ridehail and privately owned
– and that GM would be ‘following what the customer wants’. From what we have seen with
TNCs, the customer doesn’t want to share; the customer wants their own vehicle and will want
more trips if the cost goes down.
In an interview in December 2023 with David Rubenstein, chairman of the Economic Club
of Washington, D.C., Mary Barra said, “one of the greatest opportunities within 10 years is the
438 Kaplan, S., Gordon, B., El Zarwi, F., Walker, J. L., & Zilberman, D. (2019). The Future of Autonomous
Vehicles: Lessons from the Literature on Technology Adoption. Applied Economic Perspectives and Policy, 41(4),
583–597. https://doi.org/10.1093/aepp/ppz005
223
vehicle having the opportunity to be – to be a personal autonomous vehicle. So, yes, you love to
drive, but if, you know, if you’re gone to dinner with a friend and you’ve had, you know, a couple
glasses of wine, the vehicle can drive you home.439” There are two significant issues here. First, the
reference to a future of personal AVs, which so many commentators have feared because of its
impact on VMT and sprawl. This is Robin Chase’s ‘hell’ scenario – the opposite of what the former
CEO of Zipcar described as the ‘heaven’ future in which shared vehicles are the norm440. Second
she looks to a future in which a vehicle can switch between autonomous and human-driven, which
has safety and other implications. It suggests that any safety advantages that AVs might
eventually be proven to bring could be watered down as a higher proportion of vehicles can be
expected not to be driving in autonomous mode. It also takes away the guarantee that people
who purchase autonomous capability won’t be able to drive drunk or distracted – these will
remain possibilities, especially as people tend to believe they are better drivers than they in fact
are441
.
Barra’s position has echoes of the discussions in Congress in the 1960s about safety where
the manufacturers maintained that free market factors should decide what goes into the cars they
build and that the public should be given exactly what it demands442. GM says it will follow the
customer, and the federal government, according to its 2020 guidance, says people will decide. So
it can be expected that this is what will happen with consequent increases in congestion and
decreases in livability.
439 The Economic Club of Washington DC (2023, December 13) Transcript. Signature Event. Speaker, Mary Barra.
Moderator, David Rubenstein. https://www.economicclub.org/sites/default/files/transcripts/economic_club_-
_mary_barra_transcript.pdf Retrieved July 23, 2024
440 USDOT Volpe Center (2016, October 18) Robin Chase and Anthony Townsend on automated vehicles and urban
mobility. US Department of Transportation, Volpe Center. https://www.volpe.dot.gov/news/robin-chase-andanthony-townsend-automated-vehicles-and-urban-mobility Retrieved 17 December 2024
441 Roy, M. M., Liersch, M. J. (2014). I am a better driver than you think: examining self-enhancement for driving
ability. Journal of Applied Social Psychology, 43(8), 1648–1659. DOI: 10.1111/jasp.12117
442 Congressional Record, Senate (1965) July 26) 18207 Vol. 111, Part 13 89th Congress - 1st Session
https://www.congress.gov/89/crecb/1965/07/26/GPO-CRECB-1965-pt13-6-2.pdf
224
Claims that high costs of AV technology would preclude large-scale private ownership are
weak. As with all technologies the costs will come down dramatically over time – as happened
with cars and also electronic goods including televisions and computers443. We can see this
already with some of the AV hardware with some lidar producers including Velodyne, Aeva and
Luminar partnering with carmakers to build units at scale for less than $1,000 each. AlixPartners,
the consultancy, estimates that a fully autonomous “hardware stack” comprising lidar, cameras,
sensors, radar and electronics will cost as little as $7,000 by 2025444
.
Car companies that have moved towards a focus on automated features in place of full
autonomy is a means of making tech development cheaper and lowering risks, which will
encourage development. In this way, AV technology gets broken down into components as ‘addons’ to a vehicle that can be purchased (or not) by a broader range of consumers at different price
points - as has been happening for many years already with, for example, automated braking and
lane-keeping, which are now common features of newer cars. Rolling out individual automated
features incrementally allows companies to test and refine their technology in real-world
conditions without the regulatory hurdles. It also spreads research and development costs and
minimizes the financial risks associated with launching an all-or-nothing fully autonomous
vehicle. The risks associated with high profile crashes and AV computer errors are also
significantly reduced; currently one error can have a significant impact on AV reputation. Gradual
automation also avoids the problem of trust: yes, consumers will be hesitant if they’re not used to
something but it’s easier for them to accept small incremental changes than the radical change of
‘driverless’. So autonomous vehicles could emerge through stealth rather than through the
443
BLS (2015) Long-term price trends for computers, TVs, and related items. Bureau of Labor Statistics
https://www.bls.gov/opub/ted/2015/long-term-price-trends-for-computers-tvs-and-related-items.htm Accessed 27
July 2024
444 FT (2020, July 19) Self-driving Industry takes to the Highway after Robotaxi Failure. Financial Times Ltd.
https://www.ft.com/content/96d3eeff-7f52-46e3-a8a8-aeb668472034
225
proposition of whether the public wants AVs on their roads with the potential consequences on
safety (which are yet to be proven possible) and with the impacts on livability and the
environment too. Gradual automation however comes with its own separate risks.
Studies have shown that semi-automated vehicles can encourage driver complacency. It
can take time for a driver to take control of a vehicle when sent a take over request by a semiautomated vehicle and this time is affected by such things as road density, road geometry and
curves445. Tesla and many other vehicle manufacturers have been testing and developing such
technology and their impacts have been hard to discern largely because of data availability.
Koopman and Widen, for example, note that a number of Tesla crashes have involves the
‘Autopilot’ technology being shut off less than one second before a crash impact and although it is
unclear whether this is a deliberate strategy on the part of Tesla to avoid blame on the technology
- or a side-effect of reasonable design decisions – it does point to a potential area of abuse to
shed blame onto the human driver446. The history of transportation technologies, as we have seen,
has consistently sought ways of blaming drivers rather than the vehicle manufacturers. So
automated features have an additional advantage to auto makers.
A Tipping Point
For a transportation technology to gain traction in the marketplace it needs to be
sufficiently developed to be trusted by the public, for a powerful ecosystem to have developed,
and a sufficient proportion of the public willing for it to succeed. For the latter, there needs to be a
445 Morales-Alvarez, W., Sipele, O., Léberon, R., Tadjine, H. H., & Olaverri-Monreal, C. (2020). Automated Driving: A
Literature Review of the Take over Request in Conditional Automation. Electronics, 9(12), 2087.
https://doi.org/10.3390/electronics9122087
446 Widen, W. H. and Koopman, P. (2023) The awkward middle for automated vehicles; Liability attribution rules
when humans and computers share driving responsibilities. 64 JURYMETRICS J 61
226
combination of perception of trust and reliability, a perceived economic benefit, and improved
perceived personal convenience at a reasonable price. The tipping point for technology can be
seen to be achieved once there are prominent respected champions of the technology, a cultural
shift that taps into an aspirational vision and active federal financial government support. This
has been the pattern for cars, national highways and TNCs.
More recently we can see how this tipping point occurred with electric cars (EVs). While
the technology had been developing for some years with small electric vehicles, Elon Musk helped
create a tipping point for the mainstreaming of EVs by providing a public charging network and
an attractive and reliable vehicle at a price point that - when combined with a generous federal
tax credit - was reasonably competitive with gas-powered vehicles (enabled by the significant
decrease in cost of lithium-ion batteries). This combined with a cultural shift of climate
consciousness influenced by international standards (such as the debates leading up to the 2023
decision by the European Union to phase out sales of new gas cars by 2035) and US federal
government support (tax credits) and state-sponsored incentives that reduced EV up-front costs.
Tesla played a very important role in changing the mold of electric vehicles which had until that
point largely been that of slow, functional, especially micro-cars used for short urban trips that
reflected their small battery and range (such as Smart USA’s Fortwo coupe). Tesla changed this
mold, with the Model S (2012), into the mold of an electric car that was high-performance,
desirable, and a luxury good. Like Uber, it started at the luxury end of the market and moved to
the middle classes, broadening the appeal of Teslas in 2017 with the Model 3 that was affordable
to middle America.
The personality of Elon Musk and his political outlook were also important to the success
of EVs in the US and can be expected to play an important role in the future of AVs. The
environmentally-conscious Californian liberals were early adopters of EVs and provided something
of a Democrat mobility choice in an increasingly polarized country. The Californian state
227
government’s response to the pandemic and workers’ rights changed the apparent political nature
of Teslas into one that embraced Republican sentiments too, broadening the appeal, as the
headquarters moved from California to Austin, Texas. With other auto manufacturers then turning
their attention to EVs they were in effect mainstreamed capturing a broader consumer acceptance
with costs coming close to gas cars. The tipping point was the coming together of technology
development, political support, government policy and financial support, market dynamics,
international standards, which combined to produce a cultural shift.
The example of EVs provides a useful comparator for AVs, especially as Elon Musk
appears to have the ear of the President Elect. Musk has announced his intention for Tesla to enter
the robotaxi market which may have influenced GM’s decision to leave it. If Musk applied the
same methods employed to the EV market he may bring AVs to the tipping point that would
enable mass deployment. Already it appears his business interests might be furthered by the
President Elect (although the latter can be fickle) and we might expect federal regulation to
loosen and so enable an expansion of AV development. As with EVs - and past transportation
technologies - the combination of technological development, government policy, political
support, international developments could combine to produce a momentum for change. As with
all technology the price can be expected to drop significantly over time and with availability more
widespread we can expect opinion leaders and influencers to shift towards a more positive
cultural perception. We’re not yet at the tipping point for AVs but an incoming Trump presidency
could enable this point to be reached.
The continued use of silver bullets – and the city problems of the day
There has been a pattern of transportation technologies being promoted as silver bullets
to solve the mobility problems of the day. In the past this has been safety, congestion and
228
environmental impacts and these are still used today with AVs. Added to these silver bullets has
been a solution to the transportation problems of the day – such as decongesting cities. For AVs
this has been a solution that’s been important to regulators and governments: transportation
equity. The silver bullet claims of the past have been unjustified as they failed to resolve longstanding problems but are these similar claims around AVs justified? Looking at the evidence, the
claims made do not stand up to examination.
Safety
Safety claims have been the chief selling point of AVs to the public. In April 2023, three
weeks before the CPUC meeting that was to allow for paid-for rides, a full-page advertisement
paid for by Cruise appeared in the New York Times, The San Francisco Chronicle, the Sacramento
Bee and the Los Angeles Times. Headed, ‘Humans are Terrible Drivers’ (Figure 10). The
advertisement was a clear message to San Francisco, the Bay Area, Los Angeles and New York –
but not to Arizona - that safety was Cruise’s main concern. As shown in Figure 14, the company,
which had never taken out such advertising before, stated: ‘You might be a good driver, many of
us aren’t. People cause millions of accidents every year in the US. Cruise driverless cars are
designed to save lives. Our cars were involved in 92% fewer collisions as the primary contributor. *
They also never drive distracted, drowsy or drunk. Learn more at get cruise.com/safety ‘*When
benchmarked against human drivers in a comparable driving environment – from the Cruise
1,000,000 mile safety report. Fatality statistics estimated by NHTSA’.
229
Figure 12: Cruise advertisement April 2023
The 92% figure used is one that originated with the federal government and has been used
extensively, including in academia. In their systematic review of academic literature on AV safety,
Faisal et al note a central issue of contention that AVs might not make human errors but they do
make computer errors: “These studies do not consider a myriad of issues that might cause an AV
to be involved in a crash such as software failure, factors that are not included with the within the
AV’s artificial intelligence, failure to recognize a new street layout pattern, and so on”447
.
The Cruise advertisement created controversy within AV safety circles too, with Missy
Cummings (senior advisor to NHTSA and Professor at George Mason University) pointing out that
the 92% figure was taken out of context. Speaking of the source she said, “As the NHTSA itself
noted in that report, the driver’s error was “the last event in the crash causal chain…. It is not
intended to be interpreted as the cause of the crash.” She points out there were many other
possible causes, such as poor lighting and bad road design. Importantly too, she notes, “anyone
447 Faisal, A., Kamruzzaman, M., Yigitcanlar, T., & Currie, G. (2019). Understanding autonomous vehicles: A
systematic literature review on capability, impact, planning and policy. Journal of Transport and Land Use, 12(1),
p63.
230
who has ever worked in software development knows all too well: that software code is incredibly
error-prone, and the problem only grows as the systems become more complex….. AI has not
ended the role of human error in road accidents. That role has merely shifted from the end of a
chain of events to the beginning—to the coding of the AI itself. Because such errors are latent,
they are far harder to mitigate 448”.
Professor Philip Koopman commented on the advertisement: “As imperfect as humans are,
they are impressive at handling chaotic, unexpected situations. They drive on average up to 100
million miles between fatal crashes, and if you leave out the drunk and distracted drivers it is
much better than that. Computer drivers are terrible at the unexpected, and it might take a new
type of technology to fully address that issue. We don’t know yet whether computers or human
drivers will be better, and we certainly cannot assume computers will automatically be better.449”
Koopman has noted that at 100 million miles or more between human driver fatalities, it's
another 97 million or more miles before we might confirm AVs are safer – assuming there are zero
fatalities before then. Koopman summarized the AV companies’ approach as a, “Fake it until you
make it” one, which aligns with the deliberate blurring of technological advancement.
Adding to the outrage over the advertisement, Joan Claybrook, a lawyer who served as
head of NHTSA from 1977 to 1981 (and president of consumer advocacy group Public Citizen from
1982 to 2009) told TechCrunch, “Using the pain and suffering of those deaths for self-promotion
of an unproven and unsafe product is unscrupulous”450. Pointing to the 1 million miles already
448 Cummings, M. (2023, 30 July) What Self-driving cars tell us about AI risks. IEEE Spectrum.
https://spectrum.ieee.org/self-driving-cars-2662494269 Retrieved 26 July 2024
449 Koopman, P. (2024, July 26) Evidence to Congress, AV Hearing.
https://d1dth6e84htgma.cloudfront.net/IDC_Philip_Koopman_Ph_D_Testimony_Self_Driving_Cars_AV_Hearing_2
023_07_26_c2ebaa103f.pdf
450 Bellan, R. (2023, July 18) Former NHTSA Head Blasts Cruise’s ‘Humans Are Terrible Drivers’ ad. Techcrunch
https://techcrunch.com/2023/07/18/former-nhtsa-head-blasts-cruises-humans-are-terrible-driversad/?guccounter=1&guce_referrer=aHR0cHM6Ly9kdWNrZHVja2dvLmNvbS8&guce_referrer_sig=AQAAAG9pZwlzJ1i
napZlF2FL3k_Yt-
231
driven by AVs at the time, Claybrook noted that this is a tiny number compared to annual miles
driven in the US or even California: the figure, she said, “at first glance may seem like a
substantial amount but is less than 0.00003 percent of the more than three trillion miles driven
annually on U.S. roads,” in a statement released by Advocates for Highway and Auto Safety.
“Furthermore, it is infinitesimal compared to the 310 billion miles driven in California each year.
By comparison, in one work week, human ride-hailing drivers in San Francisco were nearly tripling
the one million miles Cruise took a year to accumulate.”
Writing for slate.com in July 2023, David Zipper (Senior Fellow at the MIT Mobility
Initiative) also criticized the advertisement, saying its claims are ones that the industry has made
for years and “should be treated as the self-serving deceptions that they are”451
.
What the AV companies choose not to acknowledge is that we already know how to
reduce road fatalities and injuries – the problem is the willingness to act on this. Zipper suggests,
“look across the Atlantic. With nary a self-driving car in sight, the European Union has seen a
steady decline in roadway deaths. Last fall I visited Helsinki, a city with so few annual traffic
deaths that one can count them on a single hand (for comparison, Portland, Oregon, a city with an
equal number of residents, had 63 such fatalities last year). I asked a Helsinki official what role
technology had played in his city’s safety success: “Zero,” he told me. “We simply slowed down
the cars.””452. For all the stated objectives of safety, it has been speeding traffic up through car
engineering development in the early twentieth century and highway-building soon after that has
driven transportation technology as the auto ecosystem and governments have responded to the
perceived needs or demands of consumers.
vhNZ9GNpi9eAxqWPprcBQA2BND1_Y0RwKgCLoagSFBHzwxDxTnZF8QOmK3pyaAwBr1DG8XUfBIwaSaj1qqtqSvdzfk
zrXO_5Gbyg3K7FoeW_lBtWUKJ3b5eZ_E-NRG46pJPXT5n97upSWN4Fdd
451 Zipper, D. (2023, July 17) The Safety Dance. Slate. https://slate.com/technology/2023/07/cruise-autonomousvehicles-safety-waymo-self-driving-cars-ad-new-york-times.html Retrieved July 28, 2024
452 Zipper, D. (2023, July 17) The Safety Dance. Slate. https://slate.com/technology/2023/07/cruise-autonomousvehicles-safety-waymo-self-driving-cars-ad-new-york-times.html Retrieved July 28, 2024
232
The lack of very active championing of other safety measures questions the sincerity of
AVs companies’ claims around having a safety mission. The car industry pushed back on safety
measures in the past including on airbags and seatbelts and similarly neither GM nor other car
companies have been championing safety features as standard on their vehicles including
governors to limit vehicle speed or technology to detect and disable the ignition of DUI drivers. So
one might be forgiven to think that talk of ‘missions’ on safety is not sincere.
Another deception on the part of AV companies is on the nature of ‘driverless’. It became
clear after the San Francisco dragging incident that robotaxis were dependent on a team of
remote operators and they were intervening often. There were in fact 1.5 workers dedicated to
each vehicle and they were intervening every 2.5-5 miles although none of the AV companies have
formally said how many staff they employ for such tasks and how much they intervene. These
higher-paid workers could be seen to be set to replace lower-paid jobs of TNC driving and
delivery driving today, which could in fact negate some of the cost savings from automating the
driving task. Reporting for the New York Times having visited one of the remote operator centers,
Cade Metz, pointed to motives, “By creating the illusion of complete autonomy, companies can
fuel interest in their technology and raise the billions of dollars they need to build a viable robot
taxi service”453. For tech companies success is measured in how much funding you can attract and
how much you can sell your business for – so creating an illusion of technological advancement is
very important. Gary Marcus has described this as a Wizard of Oz situation and a dark secret at
the heart of the entire driverless car industry454
.
453 Metz, C. (2024, September 11) When Self-Driving Cars Don’t Actually Drive Themselves. New York Times.
https://www.nytimes.com/2024/09/11/insider/when-self-driving-cars-dont-actually-drivethemselves.html#:~:text=Like%20other%20robot%20taxis%2C%20the,taxi%20hundreds%20of%20miles%20away
Retrieved September 22, 2024
454 Marcus, G. (2023, November 4) Could Cruise be the Theranos of AI? And is there a dark secret at the heart of
the driverless car industry? Substack. https://garymarcus.substack.com/p/could-cruise-be-the-theranos-of-ai
233
The environment
Looking at the claims on the environment one could be forgiven for being equally skeptical
- most obviously in California and other states where conventional vehicles are being phased out
anyway. Although electric cars won’t impact air quality in the same way as combustion engines,
the tires and braking systems throw off particulate matter with consequent effect on health455
.
There are other health impacts too. Having, in effect, a chauffeur on hand 24/7 will not
provide walking opportunities even if it’s just to your parked vehicle. Pourrahmani et al found
health disbenefits from the increase of AV use in the SF Bay Area, concluding that a shift of 11%
walk/bike trips to single occupancy vehicle would generate annual 773 increased annual deaths
and disability-adjusted life years (DALYs) - and 37 additional deaths due to physical activityrelated diseases and an additional 915 DALYs and 67 deaths in combination with PM2.5 effects
(mostly for cardiovascular disease)456
.
As with historic transportation technologies the interest groups that want the industry to
flourish will choose the definition of ‘environment’ to suit what they can offer. The American
Planning Association guidance on AVs emphasizes the electric power used by AVs and also what it
calls the opportunities for urban greening. The latter is a result, they say, of roads not needing to
be as wide because AVs don’t weave as human-driven cars do, they can move closer together.
They would also require less parking they say – all producing idealist images of the future that
plays down other environmental issues such as induced demand, that not all vehicles will be AV
and funding available for infrastructure changes. The research process seems flawed: it amounts
to finding a link to connect the outcomes planners want with a new technology – and perhaps
455 Air Quality Expert Group (2019) Non-Exhaust Emissions from Road Traffic. Department of the Environment,
Food and Rural Affairs, UK Government. Retrieved July 25, 2024.
456 Pourrahmani, Jaller, M., Maizlish, N., & Rodier, C. (2020). Health Impact Assessment of Connected and
Autonomous Vehicles in San Francisco, Bay Area. Transportation Research Record, 2674(10), 898–916
234
helps explain why planners ended up feeling positive about so much of the past when they have
had a significant damaging potential. The alternative is to look at the history of transportation
technology and consider the dynamics at play and then make assessments of the probable
outcomes and what might need to happen to change that course of events.
There’s another reason not to boast of environmental impacts though: AVs use an
enormous amount of electricity. Friedlander noted that SFMTA is aware of the research showing
that the amount of electricity AVs take from the grid is significantly more than for a conventional
vehicle. Speaking with Barry Einsig, a consultant who helped Caltrans write their AV strategy, AVs
hold a tremendous amount of compute power in the trunk – and using electricity for mobility too
means they need a powerful cooler too, adding to the electricity usage. Indeed, one of the
drawbacks of Cruise vehicles was that they could not capture the important airport ridership
because their Chevvy Bolts did not have space in the trunk because it was taken up by the battery
and compute systems.
Research at MIT has looked in detail at this issue of compute power. Sudhakar et al note
concern in industry and academia around emissions from data centers worldwide but the impacts
of AV compute power has been overlooked457. Sudhaker et al noted that in 2018 data centers
collectively consumed an estimated 205 TWh or 1% of the world’s electricity and contributed to
about 0.3% of the world’s emissions. The researchers found that if AVs are widely adopted, the
overall computing workload would be comparable and may even exceed current data centers’
workloads. So if there is wide scale deployment of AVs there could be significant environmental
impact.
In research led by Ashley Nunes at Harvard, they found that while electric vehicles have
lower emissions than conventional ones, deploying electric robotaxis en masse on America’s
457 Sudhakar S, Sze V, Karaman S. (2023) Data Centers on Wheels: Emissions From Computing Onboard
Autonomous Vehicles. IEEE MICRO. 2023;43(1):29-39. doi:10.1109/MM.2022.3219803
235
streets would increase the number of trips and miles driven and hence increase the overall
emissions458. The research focused on San Francisco data and concluded that the convenience of
ubiquitous fleets of robotaxis, increasing demand for rides, generating more trips and more
vehicle miles traveled would erase the environmental benefits of having an electric vehicle.
Reflecting on the importance of financing arrangements that fueled car growth in the
previous century, it’s possible different funding models could further exacerbate such problems; a
subscription service to a robotaxi company for example could further increase VMT – as could
privately owned AV models.
While much has been made of the freeing up of one’s time by not having to drive, Jevons
Paradox is relevant here. While technological progress increases the efficiency with which a
resource is used, the falling cost induces an increase in demand. The paradox being that the
increased efficiency in use of time could lead to increased consumption – i.e. more VMT and
congestion.
It’s worth also bearing in mind that the creation of the AVs themselves would have a
significant energy drain; the manufacturing of private cars and taxi and ride-hailing accounts for
around a sixth of its lifecycle carbon footprint459. It’s quite possible that AVs could lead to
additional cars per household as they may be a preferred choice for long journeys while
conventional vehicles are quicker (because they are less likely to cede road space) for intra-city
trips. Like the transition from phones to smartphones there could even be a more radical change
in the appearance of an AV. Without the need for vehicles to be shaped around the driving task
(with steering wheel, pedals and seats facing forward) a set-up more like a living room
458 Ashley Nunes et al (2021) Estimating the energy impact of electric, autonomous taxis: evidence from a select
market. Environ. Res. Lett. 16 094036. https://iopscience.iop.org/article/10.1088/1748-9326/ac1bd9/pdf
459 de Bortoli, A., & Christoforou, Z. (2020). Consequential LCA for territorial and multimodal transportation
policies: Method and application to the free-floating e-scooter disruption in Paris. Journal of Cleaner Production,
273, 122898.
236
experience could fundamentally change conception of a car. This would make it easier and more
enjoyable to travel in one – and so encourage more usage.
Equity
Big Tech, that is the five biggest tech companies (Microsoft, Apple, Amazon, Meta and
Alphabet), have seen decreasing popularity in recent years460. In part this is because of their
perceived power and the way in which they wield that power. AV companies are exposed to this
unpopularity (indeed Waymo is an Alphabet company) and one way to disassociate themselves
from the rich and powerful Big Tech is by being visible and concerned with community and equity
issues, rather than imposing themselves on communities. AV companies have used their social
media feeds to celebrate various social causes, pledge community affiliation, and ‘celebrate’ the
diverse communities of a city or their workforce. The messages appear to be not much deeper
than a tweet and so more like virtue signaling than a real commitment that prioritizes a
community or equity over company expansion and growth.
Is the broader concern for communities sincere or is it a means to get regulators and
government on side? Is it actually performative activism - publicly demonstrating a commitment
to equity but lacking genuine intentional effort to create substantial change? If AVs proliferate
with the impacts on VMT and congestion as described it seems likely that there will be a need to
introduce congestion pricing to manage the traffic in urban areas and congestion pricing is a
regressive taxation, impacting particularly on people who have to drive for a living (such as selfemployed trades people) and people who have to drive a lot precisely because they can no longer
afford to live in the city.
460 Brenan, M. (2021, February 18) Views of Big Tech Worsen; Public Wants More Regulation. Gallup.
https://news.gallup.com/poll/329666/views-big-tech-worsen-public-wants-regulation.aspx Retrieved 27 July 2024
237
Introducing congestion charges is extremely difficult because of public resistance. A more
publicly palatable way of introducing such a charge is by creating a situation where drivers
actually demand the charge. This has been the experience with residents’ parking zones in London
for example – as the zones grow, the neighboring areas demand one themselves as they become
burdened by the excess vehicles parking in their neighbourhood. But the only people likely to
demand this are those who can afford to pay for others not to be on the road.
Another future charge impacting equity may be on infrastructure. There is little deep
discussion publicly about connected vehicles – especially connected to infrastructure. We may
find that the technology doesn’t deliver on the Level 5 automation and the industry insists it will
but that the government needs to invest in infrastructure capable of connecting with vehicles to
improve congestion and safety. This cost could fall on users but from past experience of gas tax it
may be that the costs fall on general taxation and everyone pays even those who are transit
reliant.
Speaking rather vaguely and hesitantly on this Mary Barra has said, “there's already been
discussions - that at some point in the proper city boundaries - that it will limit what vehicles can
be there. I think that's where rideshare or personal autonomous vehicles will be a big piece of
that”. Such ‘limiting’ could be done by pricing – congestion charging or could be a banning of, for
example, private unshared vehicles. Or she may have meant to ban non-AVs. Whatever she
meant, these are important areas for discussion, not just between tiers of governance but with the
AV industry itself.
Congestion is also likely to impact the more major roads, on which lower income people
are more likely to live with consequent impacts on health as described and on livability of
neighborhoods. As with historic transportation technologies such impacts would mean it's the
same populations who suffer - those on low income, which disproportionately are people of color.
238
Congestion
An important driver of congestion of course is private vehicle ownership. The AV trials in
San Francisco and Phoenix have been based on the technology as a private sector driven
transport, but not entirely. San Francisco rather quietly hosted an AV shuttle service, The Loop, on
Treasure Island, as mentioned earlier. Open to the public, the Level 3 AV served seven stops on a
one and a half mile route, running every 30 minutes from 9am to 6pm with an attendant on board.
The vehicles were shared pods capable of holding up to ten passengers and served residential
areas and businesses on the north side of the island which is accessed from San Francisco bay by
a ferry service. It operated for the second half of 2023 and ran into a number of technical
problems soon after starting, the service was unreliable and the trial was forced to finish early461
.
A survey of users found that the people who used the service would have walked had the Loop not
been available.
The Phoenix State University evaluation of Valley Metro’s first mile-last mile shuttle
service for RideShare users was a subsidized curb-to-curb service using taxi or ride-hail for
people certified under the Americans with Disabilities Act and people over 65 in the greater
Phoenix areas. The vehicles were the standard Pacificas that Waymo were using in their standard
rides in the city at the time. The users were positive about the trial compared to the services they
were used to as part of the RideChoice program (they were also made cheaper for longer miles);
their preference though was that they would not share. To be enrolled in the RideChoice program
participants had to be able to enter and exit the vehicle by themselves but even so a quarter of
respondents still valued having a driver to assist with getting into and out of the vehicle and
461 Treasure Island Mobility Management Agency (2024, June) The Loop: Final Evaluation Report. SFCTA.
https://www.sfcta.org/sites/default/files/2024-06/Loop_Final_Evaluation_Report_2024-06-25_0.pdf Retrieved
December 17, 2024.
239
helping with packages, suggesting that some level of immobility may exist for these users462
.
Interestingly, the findings shows that most users of the service agreed that they’d make more
trips – almost 60% strongly agreed or somewhat agreed (only 8% disagreed).
Neither trial involved replacing single car use with AV use; never mind replacing single
occupancy vehicles with shared AVs. While the Valley Metro trial did help the mobility
disadvantaged take more trips - and we might focus on and applaud the equity implication - but
there is a concern on the ‘take more trips’ aspect if the consequent VMT, environmental and
congestion impacts went beyond the mobility disadvantaged.
The history of public utility regulation of transportation in the U.S. reveals a shift
from public-interest oversight to a more market-driven, deregulated framework. This transition
was shaped by evolving economic, social, and technological changes, leading to different
approaches in the regulation of transportation systems such as streetcars, buses, and later ridehailing and autonomous vehicles463
.
An ideal complement to transit could be an AV that would collect passengers and take
them to transit stops, replacing car journeys and lessening the need for cars. This could be a bus
with a flexible route that could collect people, who have sent a request through an app, with a
cap on the number of people to collect so that the journey time is reasonably predictable. While
this would be prohibitively expensive for a private sector concern it is something that could be
funded as a public service, subsidized with public funds in the same way that the automobile has
been in the past. The latter might be defined based on how long a direct car trip might take or
linked to the time of a train departure. Such a vehicle would take a different route depending on
462 Stopher, P., Magassy, T.B., Pendyala, R.M., McAslan, D., Arevalo, F.N., & Miller, T.R. (2021). An Evaluation of the
Valley Metro–Waymo Automated Vehicle RideChoice Mobility on Demand Demonstration, Final Report. (FTA
Report No. 0198) Federal Transit Administration, United States Department of Transportation.
463 Mallach, S. (1979). The Origins of the Decline of Urban Mass Transportation in the United States, 1890-
1930. Urbanism Past & Present, 8(8), 1–17.
240
the location of the people requesting a ride (within a broad geographical area). Similar trials have
taken place on a small scale but have not been followed up with the expansion and scale
necessary to provide a realistic alternative to car use.
The main use case for AVs currently is robotaxis and the principle of sharing with the
expectation of reducing car ownership makes the use case appealing to city governments.
However, it’s likely that car companies will continue to want to sell as many cars as possible and
it’s likely that people will continue to want to buy them. The so-called sharing economy and the
rise of TNCs did little to dent the public’s ownership of cars – taking only 4% of the mobility
market464. Behavior science suggests that people will continue to buy cars because of habit,
convenience, the familiarity heuristic and because whatever the rational arguments for giving up
cars they are reluctant to do so even if there are transit alternatives.
In the case of GM, the investment in Cruise may well have been because they saw it as a
means to sell more cars and more expensive ones (and not just selling mobility). As with any
private sector company, they are highly influenced by the demands of their shareholders and
there’s a great deal of money at stake. Reuters reported that GM had been losing $2 billion a year
on Cruise but was anticipating turning this into a $50 billion a year revenue stream each year by
2030465
.
This investment has been made in anticipation of future profit of course. While Cruise and
indeed any AV company will talk of other goals (sustainability and safety) if they are reliant on an
auto manufacturer, they can perhaps be forgiven for acting in a way that prioritizes the sale of
cars or the growth of VMT leading to more congestion.
464 Jiang, J. (2019, January 4) More Americans are using ride-hailing apps. Pew Research Center.
https://www.pewresearch.org/short-reads/2019/01/04/more-americans-are-using-ride-hailing-apps/
465 White, J. and Tiwary, S. (2023, June 2) GM chief Barra stands by her bet on autonomous vehicles. Reuters
https://www.reuters.com/business/autos-transportation/gm-ceo-stands-by-her-bet-autonomous-vehicles-2023-06-02/
Retrieved 21 April 2024
241
Importantly, the future is not likely to be shared nor robotaxi. The CEO of General Motors,
Mary Barra, told a South by Southwest tech conference audience in 2022 that she anticipated
consumers would be able to purchase their own private autonomous GM-branded AV by 2025466;
in 2023 she repeated the prediction467. Former CEO of Cruise, Kyle Vogt, speaking at TechCrunch
Disrupt in September 2023, called the rise of personal self-driving cars “inevitable”468
.
GM and other auto manufacturers exist to sell cars, not to save lives, mitigate climate
change, cut congestion or help resolve transportation inequity (except in so far as expanding their
customer base) even if their proclaimed vision is zero crashes, zero emissions and zero congestion.
Similarly, it would be unreasonable to expect venture capitalists to be investing in anything other
than that which can create a significant profit. We know from the experience of introducing safety
measures (especially seatbelts) that car companies will resist any change that could impact their
profit margins unless they are compelled to do so by legislation.
Same cast of players
The automotive and highway engineers of the 20th century have to some extent been
replaced by the software engineers of the 21st century. The data, metrics and the rationalizing by
both types of engineers are rhetoric tools that are deployed to provide an illusion of scientific,
evidence-based facts that are indisputable.
466 Bote, J. (2023, December 1) GM is pulling away from Cruise. What happens to the robotaxi company now? SF
Standard. https://sfstandard.com/2023/12/01/cruise-future-gm-pullback/ Retrieved December 22, 2024
467
White, J. and Tiwary, S. (2023, June 2) GM chief Barra stands by her bet on autonomous vehicles. Reuters
https://www.reuters.com/business/autos-transportation/gm-ceo-stands-by-her-bet-autonomous-vehicles-2023-06-02/
Retrieved 21 April 2024
468 Bote, J. (2023, December 1) GM is pulling away from Cruise. What happens to the robotaxi company now? SF
Standard https://sfstandard.com/2023/12/01/cruise-future-gm-pullback/ Retrieved 21 April 2024.
242
Similar to the historic power balance, engineers have much influence; today, it’s the
software engineers. In part this is because the AV industry has been reasonably successful in
creating a narrative around safety – or at least Waymo had and Cruise to much less extent. The
reporter I spoke to from the San Francisco Chronicle believed in the safety of the Waymo vehicles,
having visited Phoenix and met with a Waymo representative there to discuss the technology. The
impression he had (and that was shared by the reporter at the Arizona Republic I later spoke to in
Phoenix) was that the technology had been proven.
While the sentiment has been expressed earlier and elsewhere, Abraham Maslow has said,
“it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail”. If
you ask an engineer to solve the problem of vehicle safety they will find an engineering solution;
ask a police officer or lawyer and they are likely to talk of speeding fines, more patrols,
crackdowns on DUIs. And ask an urban planner and they are likely to talk of road layout, traffic
calming and public realm design. The experience of cars and road-building shows how important
it is to have a breadth of professions helping to shape how a technology should or shouldn’t be
allowed to work. What’s needed is a more strategic, independent body to listen and create a
future plan that reflects the responsibilities of a range of professions in finding mobility solutions.
If engineers create without reference to a democratically agreed strategic plan – then their work
would, “dictate what the plan shall be”469
.
Left to the coalition of state engineers, politicians, and the auto ecosystem, we are likely
to find a repeat of previous outcomes. The same processes, motives and interests have enabled a
continuation of the same urban problems with the same sections of the population benefitting or
469 Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of Tennessee
press. p66
243
losing from the technology. From what has been seen so far, the same outcomes can be expected
from AVs.
Winners and Losers
Part of the historic experience of the Interstate Highways was a playing out of the racist
policies of governments. It was the poor and, especially, communities of color that were more
likely to lose their homes and have their communities divided or broken up in the process of
creating the strategic highway infrastructure. There’s two important aspects of AVs that could
again hit communities of color more than other demographics.
First, people on lower incomes are more likely to drive older cars without more advanced
safety features. Also, consumers are likely to want autonomous personal vehicles that favor the
safety of the occupants rather than other road users. This has been the case with historic car
development and we can see it today with the increasing popularity of larger vehicles, notable
SUVs that, being heavier, are less likely to result in occupant deaths in multiple-vehicle crashes
but are more dangerous to people they hit.
The second aspect is the issue of the impact on jobs. The most obvious and immediate
change is the displacement of TNC and taxi drivers who are disproportionately people of color470
and delivery drivers too.
Looking at the winners in cities with AVs, there’s good reason to believe they will be higher
income households as reflected in their disproportionate interest in the technology and more likely
to afford the technology. A survey in 2019 of people in the Bay Area found people with high
incomes more likely to be interested in AV adoption, ‘high household incomes continue to serve as
a signal of adoption interest: belonging to the second-highest income quartile confers a 12
470 Osman, T. and Maury-Holmes, S. (2022, February) An Analysis of App-Based Drivers in California. UC
Riverside School of Business, Center for Economic Forecasting and Development.
244
percentage point higher interest in fully automated AVs relative to those in the lowest income
quartile, while membership in the highest income quartile yields an 11–19 percentage point higher
interest in any of the AV technologies’471
.
While in theory the technology should make taxi trips cheaper (because you don’t have to
pay a driver) this may not translate into rides cheap enough to compete with transit for all but the
more wealthy. We still don’t know the true costs of robotaxi fleets, the cost of the hardware and
software and the impact of not having someone in the vehicle to affect behavior. If there
continues to be a need for remote intervention then that’s a staff cost (and they may as well be at
the wheel).
San Francisco is in the enviable position of attracting transit riders across income brackets
and class; the danger is that AVs peel off the more wealthy into AVs. This could present two
problems: first, that farebox income drops and threatens the future of some routes and the
affordability of maintenance; second, few riders and poor maintenance can create fear of crime –
and crime itself. The losers here would be those who cannot afford to switch modes – people on
low income with fewer transit choices.
471
Spurlock, Sears, J., Wong-Parodi, G., Walker, V., Jin, L., Taylor, M., Duvall, A., Gopal, A., & Todd, A. (2019).
Describing the users: Understanding adoption of and interest in shared, electrified, and automated transportation in
the San Francisco Bay Area. Transportation Research. Part D, Transport and Environment, 71, 283–301.
245
Chapter 8: Conclusion
“So we beat on, boats against the current, borne back ceaselessly into the past.472
”
F. Scott Fitzgerald, The Great Gatsby, 1925
This dissertation has found that the history of transportation technology shows a track
record of promises to resolve the problems caused by the impacts of automobiles but that it fails
to deliver. The same messages, promises and actor involvement that have been features of past
technologies are features in the development of AVs. The motives are the same: sell technology
and sell cars. So there seems no reason to believe the outcomes will be any different.
This final chapter concludes with a potential way forward and a call to action for planners.
It also raises limitations of the research in this dissertation and suggests areas for further
research.
The need for honest discussion
The public has been led to believe it is entirely possible to have safe, affordable
congestion-free car travel. The promises of the series of transportation technologies has kept this
myth alive and AVs are the last in this long line. People are susceptible to the images of the future
and the advertisements and persuasive techniques of tech companies and auto manufacturers
472 Fitzgerald, F. S. (2005) The great Gatsby. The Folio Society.
246
that make them think, ‘if only the road was wider’, ‘if only the car ran on electricity’, ‘if only cars
could be programmed to be safe’. But ‘if you build it they will come’, is a truer and more fitting
phrase to describe technologies that make it easier, quicker or more comfortable to drive; it is
also a more publicly understandable concept than ‘induced demand’, which could have been
invented by someone bent on keeping the concept hidden!
The actors involved in the development of AV technology have been disingenuous to say
the least. This was demonstrated in their lack of clarity around human intervention in AV testing
that was revealed after the Cruise pedestrian dragging incident in October 2023 and their
continued reticence in sharing data.
Their statements on ethics and concerns for safety are deeply questionable. While they
tweet about such things as celebrating National Reading Week, International Day of Women and
Girls in Science, or Black History Month these claims to social consciousness appear shallow
because at the same time they bypass the very important equity issues inherent in their
technology. Women and people of color are more likely to be transit riders and the low income
(more likely to be people of color) are less likely to be able to afford AV rides. Both Cruise and
Waymo tweeted their celebration of Black History Month when transportation history shows that
the technology and engineering of the interstate highways did so much damage to Black
communities and surely it doesn’t take much deep research into Black history to connect the
threat of new technology to that community. It seems the same communities would be negatively
impacted by the rollout of AVs - notably the transit dependent and those who drive for a living,
who are more likely to be people of color. It seems the same people would benefit from an AV
world as has happened in the past and that’s more likely to be higher income earners.
While it is true that it is the vulnerable who are more likely to be victims of road violence
today and that AVs might reduce impact on them, it remains the case that there just isn’t the
evidence to demonstrate that. And while in theory a fully developed AV may have such an impact
247
there’s no evidence that owners of AV technology would enable its use to protect vulnerable road
users - especially if a driver can switch between AV and non-AV use or software can protect
vehicle occupants over those outside the vehicle.
Other area that need more openness and honesty is around the discussion of congestion,
electricity demands of AVs, and projections for transit; on these issues too, the AV ecosystem has
been quiet for all its claims to environmental and social consciousness. These would be
meaningful things to discuss in public if their messaging about equity and ethical matters were
genuine. They don’t because such discussions won’t help sell their product, which is
understandably their overarching objective which history shows has always been the case for
transportation technology. If equity and ethics were overriding principles that would draw out the
very significant potential impacts and enable the public to make informed choices. To be fair,
their role is to sell Stuff, it is not their responsibility to present the options to the public; but they
should stop using ethics and equity as a marketing strategy; these concepts are too important to
use in this way.
The power of personality
The history of transportation technology has shown how key personalities have been
important in driving the advancement of technologies. Henry Ford for the automobile, President
Eisenhower and Larry Page for the high speed national highway system, Travis Kalanik for TNCs,
Elon Musk for EVs.
While transportation will not be directly under Musk’s remit, autonomous driving lends
itself well to a re-framing as an efficiency issue – such as efficiency of the transportation system
and of commercial logistics, helping the national economy. Given the President Elect’s choice for
Secretary of State for Transportation, Sean Duffy, has very little actual transportation experience
248
we might assume his views are more flexible and open to persuasion than one who has had a
career of developing a strong viewpoint. Duffy will be responsible for overseeing projects that
involve companies run by Elon Musk, which have billions of dollars in contracts with the federal
government across a number of agencies, including the Department of Transportation. This
combination looks set for the AV industry to receive federal support for the technology that has
been so important in historic examples of transportation technology in achieving a tipping point
for its acceptance and rollout, irrespective of longer-term impacts.
The loosening of regulation, while it might help America develop AV technology faster,
comes with longer-term risks. As Professor Koopman told Congress in 2023, ‘A weak regulatory
system that does not encourage companies to emphasize safety and reliability will hurt us more
than it helps us in this race, by forcing our companies to respond to investor demands for more
deployment without providing regulatory pressure for them to spend sufficient resources on
safety’
473
. Without independent and verifiable data on safety, AV crashes and injuries could
attract enough public attention and concern to set the technology back a generation.
A collaborative way forward
Public policy works best when it is a collaboration - and a coming together on equal terms
- of urban planners, engineers, economic developers and politicians. This is an important step
towards creating the mobility future that many cities want, one that harnesses the power of
technology while shaping it towards the longer-term needs of society and the environment. More
needs to be done to examine the impacts of AVs using informed, honest dialogue with politicians,
473 US Congress (2023, July 26). Philip Koopman Testimony on Self Driving Cars
https://d1dth6e84htgma.cloudfront.net/IDC_Philip_Koopman_Ph_D_Testimony_Self_Driving_Cars_AV_Hearing_2
023_07_26_c2ebaa103f.pdf
249
governments and the public if we are to harness the potential benefits of autonomy while
avoiding the potential disbenefits.
An important feature of the past has been the weakness of the role of urban planners in
decision-making in transportation technologies, but they have a vital role and need to be part of a
collaborative effort to improve cities. Planners have a strategic overview of a city and its people
and are responsible for planning for the long-term and making informed estimates of changing
needs and the consequences of a variety of impacts, including technology, climate change,
population growth, and trends in preferences and public health.
Unlike politicians, planners are not tied to the electoral cycle and the pursuit of quick wins
- although they do of course (at least to some extent) work to the political directive of politicians.
Their wide remit means they are not tasked with solving other societal problems – such as
economic growth, jobs, skills and productivity – but they do play a role in achieving these
objectives.
Planners have a role in ensuring politicians know the long-term impacts of AVs for a city
and to help them make informed decisions on behalf of the people they represent. Planners can
help draw out the discussions that need to be had - such as between citizens and their elected
representatives and to handle more congestion - and encourage open discussion with decisionmakers at state level. This may mean difficult discussions about congestion charging and
infrastructure costs.
Planners at state level may be dominated by transportation planners with a more
engineering, rather than livability and placemaking, focus. They need to bear in mind that such a
perspective is outdated and they need to engage with their cities to understand impacts and
future infrastructure and societal needs to inform their decision-making now. That could shape AV
trials, for example, that focus on encouraged transit use.
250
If developed, autonomous vehicle technology has the potential to reduce road crashes and
save lives but unless we learn the lessons of the past and have an open and honest appraisal of
its likely impact in the future we will find ourselves driving in circles.
A power of urban planners
Given the strength of political power in matters of transportation technology - and that
planners have no formal role in the process - it seems hard to see how planners could exert the
necessary influence to change things. There may be ways however to strengthen their role, which
could be achieved in two ways.
First, as part of a necessarily informed democratic process, the public sector urban
planning profession should have a formal seat at the table in state government discussion and not
one watered down by a multitude of other interests. Practically speaking they would provide a
specific, prominent section early on within all agenda reports that are used to inform debates and
votes. Within the body of the report it would need to provide sufficient text to convey the views of
planning with accompanying detailed appendix, which would provide wider research and, where
appropriate, differing views within the profession.
This process would necessitate an engagement of planners across the state and could be
achieved through the extensive state-level network of the American Planning Association. The link
to the APA is important here because of the existing network which would help ensure the person
who holds that seat is representative of planners across the whole state. They or their family
members should not have been paid in the past or currently by the industry.
The second means of achieving this is through the ethical professionalism of the APA. The
APA could emphasize the ‘duty’ of planners to provide an honest view and steer. The appointment
of such people would inevitably involve a more specific network from within the APA at a national
251
level. This should be afforded the prominence it deserves – not as a minor element of their
transportation work. They could harness the insights from other national independent bodies,
especially the Transportation Research Board (TRB). Knitting together the various independent
bodies that have been working on AVs would help avoid ‘capture’ by special interests in favour or
against AVs (there is of course a history here, notably with the appointment of John Reynolds at
the CPUC who was Managing Counsel for Cruise from 2019-2022). Tying these matters to the
ethics of the APA would help keep it tied to the profession as a whole which is important because
reputationally to the profession.
The planner that represents the profession at state level would be appointed through the
APA (outside the state government appointment system). The increased influence of the role
would encourage a more charismatic and knowledgeable person to emerge. It should probably be
the President of the state chapter of the APA rather than a specific postholder, which would more
easily be subject to undue ecosystem influence. The benefit of the President is that they will owe
their position to the urban planners across the state, it also provides a prominence to the issue, it
brings more clout in discussions with politicians (because of seniority) and it also necessitates a
wider discussion within the profession rather than it being hived off to an ‘AV rep’ which could act
more as a ‘bolt-on’ interest group.
It would be expected that this would encourage discussion and debate within the states
and at the APA annual conference, again helping the issues to be subject to wider urban planning
thoughts. It would be highly beneficial for the APA to review their 2018 guidance and ensure that
those who drive it in the future are not purely academics (who have an interest in carving a
position that can provide the basis of citation scores) nor led by people who have held positions
for AV companies or received support (or expect to receive support) in some way, such as free AV
rides other than for themselves alone (not for an organized wider group). Close relationships to
those who also benefit should also be considered (such a close family member who benefits in
252
some way from an AV or auto company). There should be an expectation that if the authors of the
new publication benefit in the future in this way then the publication’s independence is
compromised. Declarations of interest should be made available within the publication that would
reflect the APA’s professional ethical integrity.
While this may appear a complex new layer, the issues of transportation technology and
the potential impacts on cities and their people are potentially too big to leave to the current
processes.
Learning from history
The history of transportation technologies and the progress so far on vehicle automation
point to the same consequences of previous technologies. Much is promised but the same unsafe
roads, congestion and environmental degradation have ensued. We already know now how to
make roads safer, reduce congestion, improve livability and lessen the impact of mobility on the
environment – that’s active travel, which is the very thing that AVs would displace and the very
thing that the planning profession works towards. The planning profession and academics have
been too quick to look at a new technology and find a way of connecting it to the future they want
to see, providing an optimistic view of the technology. A better way to understand and predict the
future though is to look to the past.
Limitations of this research
If this research were to be repeated it could add further case studies to the two included
here. Although the two chosen are very different it would be interesting to contrast with
253
experiences elsewhere, especially ones that may have worked more extensively with transit
providers.
The need for further research
There is an urgent need for further research to understand the costs of achieving fully
automated systems, especially the societal costs. This should then be compared to the perceived
benefits of fully automated systems.
Such research should include the costs of infrastructure improvements that would
maximize the benefits of AV technology. The cost of V2I (vehicle to infrastructure) has not been
widely examined but could have important implications to how well the technology could develop.
As in other areas of significant public cost there needs to be an open discussion on cost
implications. This research should include an examination of who should bear the cost of
increased congestion through congestion pricing for example. While technology may enable more
throughput of traffic there may need to be limitations on vehicles for livability reasons. Part of this
further research should also include possible exceptions to restrictions - such as for example - for
freight, people with disabilities or robotaxis. This could be a way of encouraging robotaxis and
discouraging AVs for personal use but would be likely to create conflict unless such rules are
discussed early on.
Further research might also examine the issue of flexibility in the physical shape of
vehicles if they are autonomous. So for example in the future an AV need not be a standard car
shape but could be more in the style of a small living room or even a whole modular room that
plugs into the rest of your home on arrival. The implications of this would be the basis for
interesting further research.
254
Autonomous vehicles provide an opportunity to increase the safety of roads but
transportation technologies of the past have boasted of being solutions to pressing problems too.
What’s needed now is honesty and an informed, realistic discussion about the future of this
technology - and urban planners have an important role to play.
255
Appendix 1
Common questions to all case study interviewees
Theme Question
Organizations
involved
The role of your organization [check understanding]
What influence has it had on AVs?
Where in your strategy are AVs (if mentioned)
Perception of the organizations leading/ involved – and what are their
interests in doing so?
Which organizations have exercised the most influence?
Communication What impact are the AVs having on SF? [Newspapers make it seem a
bigger deal than it is?]
How have AVs been sold to case study area? (Any documents?) [Largely
on safety?]
Public How positive is the public about AVs? [Apart from Street Rebel coning,
actually not much active opposition?]
Has this changed at all during pilot?
How do you know?
Expectations What’s your perception of how successful they’ve been?
What are the indicators of success?
Who thinks they are/ aren’t?
Who benefits? Who is benefitting from the pilots?
Who is not benefitting?
256
Going forward How do you see things developing?
(Inevitable?)
What do you think about impact on traffic or transit?
How do you see the future?
257
Bibliography
Air Quality Expert Group (2019) Non-Exhaust Emissions from Road Traffic, Department of the
Environment, Food and Rural Affairs, UK Government.
Altshuler, A. A., & Luberoff, D. E. (2004). Mega-Projects: The Changing Politics of Urban Public
Investment (1st ed) Brookings Institution Press.
Anderson, J. M., Nidhi, K., Stanley, K. D., Sorensen, P., & Samaras, C. (2014). Autonomous Vehicle
Technology: A Guide for Policymakers (1st ed.). RAND Corporation.
Aoyama, Y., & Alvarez Leon, L. F. (2021). Urban governance and autonomous vehicles. Cities, 119,
103410.
Arthur, B., (1994) Increasing Returns and Path Dependence in the Economy. University of
Michigan Press, Ann Arbor.
Babar, Y., & Burtch, G. (2020). Examining the Heterogeneous Impact of Ride-Hailing Services on
Public Transit Use. Information Systems Research, 31(3), 820–834.
Barrios, J. M., Hochberg, Y. V. and Yi, H. (2019, April 3) The Cost of Convenience: Ridehailing and
Traffic Fatalities.
Bass, F. M. (2004). A New Product Growth for Model Consumer Durables. Management Science, 50
(12_supplement), 1825–1832.
Bass, K (2023, November 1) Letter to the honourable Alice Busching Reynolds, President CPUC, Re:
Driverless autonomous vehicle passenger deployment programme in Los Angeles. From Karen
Bass, Mayor, City of Los Angeles https://mayor.lacity.gov/sites/g/files/wph2066/files/2023-11/2023-11-
01%20Mayor%20Karen%20Bass%20Letter%20to%20CPUC%20re%20AV%20Deployment%20%281%29.pdf
Behroozi, M. (2023) Understanding the Impact of Ridesharing Services on Traffic Congestion. In:
Heydari, B., Ergun O., Dyal-Chand R., Bart Y., eds. Reengineering the Sharing Economy: Design,
Policy, and Regulation. Cambridge University Press; 2023:119-145
Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs : toward a theory of sociotechnical change.
MIT Press.
Bijker. W. E. (2015). Social Construction of Technology. In International Encyclopaedia of the
Social & Behavioral Sciences (Second Edition, Vol. 24, pp. 136). Elsevier Ltd.
Bijker, W. E., Hughes, T. P., & Pinch, T. (2012). The Social Construction of Technological Systems:
New Directions in the Sociology and History of Technology (Anniversary edition.). The MIT Press.
Bissell, D. (2018) Automation interrupted: How autonomous vehicle accidents transform the
material politics of automation. Political Geography, 65, 57–66.
258
Bliss, L. (2019, May 30) Why aren't cities getting ready for autonomous vehicles? Citylab
Transportation https://www.bloomberg.com/news/articles/2019-05-30/study-u-s-cities-mustplan-for-self-driving-cars
Bloom, N. (2023). The Great American Transit Disaster: A Century of Austerity, Auto-Centric
Planning, and White Flight. Chicago: University of Chicago Press.
de Bortoli, A., & Christoforou, Z. (2020). Consequential LCA for territorial and multimodal
transportation policies: Method and application to the free-floating e-scooter disruption in Paris.
Journal of Cleaner Production, 273, 122898.
Brehm, S. S., & Brehm, J. W. (2013). Psychological Reactance: A Theory of Freedom and Control.
Academic Press.
Brown, A., & Rodriguez, G. (2018, December) Federal, state, and local governance of automated
vehicles. UC Davis Institute of Transportation Studies & Policy Institute for Energy, Environment,
and the Economy, Issue Paper, https://epm.ucdavis.edu/sites/g/files/dgvnsk296/files/inlinefiles/AV%20Gov_IssuePaper_FINAL_TH_2018_14_12%20%281%29.pdf
Brown, J. (2005). A tale of two visions: Harland Bartholomew, Robert Moses, and the development
of the American freeway. Journal of Planning History, 4 (1), 3–32.
Brown, J. R., Morris, E. A., & Taylor, B. D. (n.d.). Planning for Cars in Cities: Planners, Engineers,
and Freeways in the 20th Century. Journal of the American Planning Association, 75(2).
Burrows, M., Burd, C. & McKenzie, B. (2021, April) Commuting by Public Transportation in the
United States: 2019 American Community Survey Reports.
https://www.census.gov/content/dam/Census/library/publications/2021/acs/acs-48.pdf
Butler, S., & Topham, G. (2017, September 22) Uber stripped of London license due to lack of
corporate responsibility. The Guardian
https://www.theguardian.com/technology/2017/sep/22/uber-licence-transport-for-london-tfl
California Public Utilities Commission (2023, 10 August) CPUC approves permits for Cruise and
Waymo to charge fares for passenger service in San Francisco. Press release, CPUC.
https://www.cpuc.ca.gov/news-and-updates/all-news/cpuc-approves-permits-for-cruise-andwaymo-to-charge-fares-for-passenger-service-in-sf-2023
Calthorpe, P., & Walters, J. (2017, March 1) Autonomous vehicles: hype and potential. Urbanland.
https://urbanland.uli.org/sustainability/autonomous-vehicles-hype-potential
Caltrans (2021, February) California Transportation Plan 2050, California Department of
Transportation (Caltrans). https://dot.ca.gov/-/media/dot-media/programs/transportationplanning/documents/ctp-2050-v3-a11y.pdf
Campbell, H. (2018). Who Will Own and Have Propriety over Our Automated Future? Considering
Governance of Ownership to Maximize Access, Efficiency, and Equity in Cities. Transportation
Research Record, 2672(7), 14–23.
259
Castell, S. & Stilgoe, J. (2019) ‘Driverless Futures?’, Ipsos Mori, London
https://www.ipsos.com/sites/default/files/ct/publication/documents/2019-04/driverless-futuresipsos-2019.pdf
Carmichael, M. (2018. March 29) Will a driverless future be heaven or hell? Ipsos
https://www.ipsos.com/en-us/knowledge/new-services/will-driverless-future-be-heaven-or-hell
Cavoli, C. et al., (2017.) Social and behavioural questions associated with Automated Vehicles. A
Literature Review, London: Department for Transport.
Chatman, D. G., & Moran, M. (2019). Autonomous Vehicles in the United States: Understanding
Why and How Cities and Regions Are Responding.
City and County of San Francisco (2022, December) City Policy - Autonomous Vehicle Declaration
of Policy. File No. 221212. Resolution No. 529-22. Adopted 12 December, 2022.
https://sfbos.org/sites/default/files/r0529-22.pdf
Clewlow, R. R. (2017, October 11) New Research on How Ride-Hailing Impacts Travel Behavior.
Planetizen https://www.planetizen.com/features/95227-new-research-how-ride-hailingimpacts-travel-behavior
Clewlow, R. R., & Mishra, G. S. (2017). Disruptive Transportation: The Adoption, Utilization, and
Impacts of Ride-Hailing in the United States.
Cohen, T., & Cavoli, C. (2019). Automated vehicles: exploring possible consequences of
government (non)intervention for congestion and accessibility. Transport Reviews, 39(1), 129–
151.
Cohen, T., & Jones, P. (2020). Technological advances relevant to transport – understanding what
drives them. Transportation Research. Part A, Policy and Practice, 135, 80–95.
Cohen, A. & Shaheen, S. (2016) Planning for Shared Mobility American Planning Association PAS
Report 583.
Cohen, T., Stilgoe, J. & Cavoli, C. (2018) Reframing the Governance of Automotive Automation:
Insights from UK Stakeholder Workshops. Journal of Responsible Innovation 5: 3 257–279.
Cohen, T., Stilgoe, J., Stares, S., Akyelken, N., Cavoli, C., Day, J., Dickinson, J., Fors, V., Hopkins, D.,
Lyons, G., Marres, N., Newman, J., Reardon, L., Sipe, N., Tennant, C., Wadud, Z., & Wigley, E.
(2020). A constructive role for social science in the development of automated
vehicles. Transportation Research Interdisciplinary Perspectives, 6, 100133.
Collingridge, D. (1980) The Social Control of Technology. St. Martin’s Press, New York.
Crute, J., Riggs, W., Chapin, T. and Stevens, L. (2018) Planning for autonomous mobility, American
Planning Association PAS Report 592.
Currie, G. (2018) Lies, Damned Lies, AVs, Shared Mobility and Urban Transit Futures, Journal of
Public Transportation Vol. 21 No1, 19-30.
260
Dafoe. (2015). On Technological Determinism: A Typology, Scope Conditions, and a Mechanism.
Science, Technology, & Human Values, 40(6), 1047–1076.
Deakin, E., Frick, K., & Shively, K. (2012) ‘Dynamic Ridesharing’. University of California
Transportation Center, Working Papers, University of California Transportation Center.
Di Lillo, L., Gode, T., Zhou, X., Atzei, M., Chen, R., & Victor, T. (2024). Comparative safety
performance of autonomous- and human drivers: A real-world case study of the Waymo
Driver. Heliyon, 10(14), e34379-.
Diao, M., Kong, H. & Zhao, J. (2021) Impacts of transportation network companies on urban
mobility. Nat Sustain 4, 494–500.
Duarte, F., & Ratti, C. (2018). The impact of autonomous vehicles on cities: a review. The Journal
of Urban Technology, 25(4).
Dubal, V. B. (2017). The Drive to Precarity: A Political History of Work, Regulation, & Labor
Advocacy in San Francisco’s Taxi & Uber Economies. Berkeley Journal of Employment and Labor
Law, 38(1), 73–135. http://www.jstor.org/stable/26356922
Dunn, J. A. (1998). Driving forces : the automobile, its enemies, and the politics of mobility.
Brookings Institution Press.
Ehsani, J. P., Michael, J. P., Igusa, T., Mueller, J., Chang, C. H., & Yenokyan, G. (2024). Advancing
Transportation Equity and Safety Through Autonomous Vehicles. Health Equity, 8(1), 143–146.
Fagan, M., Comeaux, D., & Gillies, B. (2021) Autonomous vehicles are coming: five policy actions
cities can take now to be ready. Taubman Center for State and Local Government at the Harvard
Kennedy School.
https://www.hks.harvard.edu/sites/default/files/Taubman/Research/Autonomous%20Vehicles%20
Are%20Coming_Five%20Policy%20Actions%20Cities%20Can%20Take%20Now%20to%20Be%20Ready.p
df
Fagnant, D.J., & K. Kockelman (2015) Preparing a Nation for Autonomous Vehicles: Opportunities,
Barriers and Policy Recommendations. Transportation Research Part A: Policy and Practice 77:
167–81.
Fagnant, D. J., Kockelman, K. M., & Bansal, P. (2015). Operations of Shared Autonomous Vehicle
Fleet for Austin, Texas, Market. Transportation Research Record, 2563(1), 98–106.
Faisal, A., Kamruzzaman, M., Yigitcanlar, T., & Currie, G. (2019). Understanding autonomous
vehicles: A systematic literature review on capability, impact, planning and policy. Journal of
Transport and Land Use, 12(1), 45–72.
Fallon, M. (2018). Self-driving cars : The new way forward. Lerner Publishing Group.
Ferlis, R. A. (2007, July/ August) The Dream of an Automated Highway Public Roads.
Flink, J. J. (1970). America adopts the automobile, 1895-1910. MIT Press.
261
Flink, J. J. (1990) The Automobile Age. MIT Press
Flyvbjerg, B. (2012). Why mass media matter and how to work with them: phronesis and
megaprojects. In Real Social Science (pp. 95–121). Cambridge University Press.
Flyvbjerg, B. (Ed.). (2014). Megaproject planning and management: Essential readings.
Cheltenham, UK: Edward Elgar.
Flyvbjerg, B., Landman, T, & Schram, S., eds., Real social science: Applied phronesis (pp 95–121).
Cambridge, England
Forlano, L. (2019) International Journal of Communication 13, 2811–2838.
Ford, L. R. (1994) Cities and Buildings Baltimore: Johns Hopkins University Press.
Foster M (1981) From Streetcar to Superhighway Temple University Press, Philadelphia.
Freemark, Y., Hudson, A., & Zhao, J. (2019). Are Cities Prepared for Autonomous Vehicles?:
Planning for Technological Change by U.S. Local Governments. Journal of the American Planning
Association, 85(2), 133–151.
Funk, J. (2019, Fall) What’s behind technological hype? Issues in Science and Technology 36-42.
Gehrke, S. R., & Huff, M. P. (2024). Spatial equity implications and neighborhood indicators of
ridehailing trip frequency and vehicle miles traveled in the Phoenix metro region. Transportation
(Dordrecht), 51(1), 271–295.
Gehrke, S.R., Huff, M.P., & Reardon, T.G. (2021) Social and trip-level predictors of pooled ridehailing service adoption in the Greater Boston region. Case Stud. Transport Policy.
Gkartzonikas, C., & Gkritza, K. (2019). What have we learned? A review of stated preference and
choice studies on autonomous vehicles. Transportation Research. Part C, Emerging
Technologies, 98, 323–337.
Gkartzonikas, C., Ke, Y., & Gkritza, K. (2022). A tale of two modes: Who will use single user and
shared autonomous vehicles. Case Studies on Transport Policy, 10(3), 1566–1580.
Goodman, E. P., & Powles, J. (2019). Urbanism Under Google: Lessons From Sidewalk
Toronto. Fordham Law Review, 88(2), 457–498.
Grush, B., & Niles, J. (2018). The End of Driving: Transportation Systems and Public Policy Planning
for Autonomous Vehicles (1st ed.). Elsevier.
Guerra, E. (2016). Planning for Cars That Drive Themselves: Metropolitan Planning Organizations,
Regional Transportation Plans, and Autonomous Vehicles. Journal of Planning Education and
Research, 36(2), 210–224.
262
Hall, J. D., Palsson, C., & Price, J. (2018). Is Uber a substitute or complement for public
transit? Journal of Urban Economics, 108, 36–50.
Harvard Business Review (2018, February 15) How Arizona is leading the autonomous vehicle
revolution. Sponsor content from Arizona Commerce Authority.
https://hbr.org/sponsored/2018/02/how-arizona-is-leading-the-autonomous-vehicle-revolution
Heermann, P. D., & Caskey, D. L. (1995). Intelligent Vehicle Highway System: Advanced Public
Transportation Systems. Mathematical and Computer Modelling, 22(4), 445–453.
Henderson, J. (2013). Street Fight: The Politics of Mobility in San Francisco (1st ed.). University of
Massachusetts Press
Hewitt, C., Politis, I., Amanatidis, T., & Sarkar, A. (2019). Assessing public perception of selfdriving cars: the autonomous vehicle acceptance model. Proceedings of the 24th International
Conference on Intelligent User Interfaces.
Hughes, T. P. (1994) Technological Momentum in Smith, & Marx, L. (1994). Does technology drive
history? : the dilemma of technological determinism. MIT Press.
Ikuta, Y. (1988) The American Automobile: Advertising from the Antique and Classic Eras Chronicle
Books, San Francisco, CA.
Interrante J. (1980) The Road To Autopia: The Automobile And The Spatial Transformation Of
American Culture. Michigan Quarterly Review. 1980;19(4):503.
Isaac, L. (n.d.). How Local Governments Can Plan for Autonomous Vehicles. In Road Vehicle
Automation 3 (pp. 59–70). Springer International Publishing.
J.D. Power (2023, October 19) Most consumers not embracing robotaxis in their community. Press
release, J.D. Power. https://www.jdpower.com/business/press-releases/2023-us-robotaxiexperience-study
J.D. Power and MIT Advanced Vehicle Technology Consortium (2023, October 4) Stakes are high
and consumer confidence is fragile for automated vehicles, J.D. Power finds. Press release
https://www.jdpower.com/sites/default/files/file/2023-
09/2023130%20U.S.%20Mobility%20Confidence%20Index_0.pdf
Jackson, K. T. (1987). Crabgrass frontier : the suburbanization of the United States. Oxford
University Press.
Jasanoff, S. (2021). The dangerous appeal of tech: Technology doesn’t rule us. We direct it, but
often by inaction. In Technology review (Vol. 124, Number 4, pp. 16-). Technology Review, Inc.
Jiang, J. (2019) More Americans are using ride-hailing apps, Pew Research Center.
https://www.pewresearch.org/short-reads/2019/01/04/more-americans-are-using-ride-hailingapps/
263
Jiang, Q., Yueshuai He, B., & Ma, J. (2022). Connected automated vehicle impacts in Southern
California part-II: VMT, emissions, and equity. Transportation Research. Part D, Transport and
Environment, 109, 103381-.
Kang, C. (2017). Where self-driving cars go to learn. 11 November 2017 New York: New York
Times Company. https://www.nytimes.com/2017/11/11/technology/arizona-tech-industryfavorite-self-driving-hub.html
Kaplan, S., Gordon, B., El Zarwi, F., Walker, J. L., & Zilberman, D. (2019). The Future of Autonomous
Vehicles: Lessons from the Literature on Technology Adoption. Applied Economic Perspectives and
Policy, 41(4), 583–597.
Kassens-Noor, E., Dake, D., Decaminada, T., Kotval-K, Z., Qu, T., Wilson, M., & Pentland, B. (2020).
Sociomobility of the 21st century: Autonomous vehicles, planning, and the future city. Transport
Policy, 99, 329–335.
Kline, R., & Pinch, T. (1996). Users as Agents of Technological Change: The Social Construction of
the Automobile in the Rural United States. Technology and Culture, 37(4), 763–795.
Koehn, J. (2023, August 24) Here's how much robotaxi companies spent to influence California
regulators. The SF Standard https://sfstandard.com/2023/08/23/cruise-waymo-spent-millionsto-lobby-for-free-rein-on-san-francisco-streets/
Koopman, P (2021, October 22) Debunking AV Industry Positions on Standards and Regulations
Safe Autonomy https://safeautonomy.blogspot.com/2021/10/debunking-av-industry-positionson.html
Lawson, C. (2007). Technology, technological determinism and the transformational model of
social activity. In Contributions to Social Ontology (1st ed., pp. 32–49). Routledge.
Leavitt, H. (1970) Superhighway, Superhoax Doubleday & Co (New York).
Linja, A., Mamun, T. I., & Mueller, S. T. (2022). When Self-Driving Fails: Evaluating Social Media
Posts Regarding Problems and Misconceptions about Tesla’s FSD Mode. Multimodal Technologies
and Interaction, 6(10), 86-.
Liu, C. Y., & Painter, G. (2012). Travel Behavior among Latino Immigrants: The Role of Ethnic
Concentration and Ethnic Employment. Journal of Planning Education and Research, 32(1), 62–80.
Lovelace, E. (1993). Harland Bartholomew: His contributions to American urban planning. Urbana,
IL: Department of Urban and Regional Planning, University of Illinois, Urbana-Champaign.
Luger, S. (1999). Corporate Power, American Democracy, and the Automobile Industry (1st ed).
Cambridge University Press.
Magassy, T. B., Batur, I., Mondal, A., Asmussen, K. E., Bhat, C. R., & Pendyala, R. M. (2024).
Multidimensional Analysis of Willingness to Share Rides in a Future of Autonomous
Vehicles. Transportation Research Record, 2678(6), 865–880.
264
Maharawal M (2017) San Francisco’s tech-led gentrification: Public space protest and the urban
commons. In J Hou and K Knierbein (eds) City Unsilenced: Urban Resistance and Public Space in
the Age of Shrinking Democracy (pp 30–43). New York: Routledge.
Maharawal MM. (2023) Infrastructural Activism: Google Bus Blockades, Affective Politics, and
Environmental Gentrification in San Francisco. Antipode. 2023;55(5):1454-1478.
Maia, S. C., & Meyboom, A. (n.d.). Understanding the Effects of Autonomous Vehicles on Urban
Form. In Road Vehicle Automation 4 (pp. 201–221). Springer International Publishing.
Manville, M., Taylor, B. D., & Blumenberg, E. (2018). Falling Transit Ridership: California and
Southern California.
McAslan, D., Gabriele, M., & Miller, T. R. (2021) Planning and Policy Directions for Autonomous
Vehicles in Metropolitan Planning Organizations (MPOs) in the United States, Journal of Urban
Technology, 28:3-4, 175-201.
McAslan, D., Najar Arevalo, F., King, D. A., & Miller, T. R. (2021). Pilot project purgatory? Assessing
automated vehicle pilot projects in U.S. cities. Humanities & Social Sciences
Communications, 8(1), 1–16.
McCarroll C & Cugurullo F (2022) No city on the horizon: Autonomous cars, artificial intelligence,
and the absence of urbanism. Front. Sustain. Cities 4:937933.
McLaren, Duncan, & Julian Agyeman. Sharing Cities : A Case for Truly Smart and Sustainable
Cities, MIT Press, 2015.
McShane, C. (1997). The Automobile: A Chronology of Its Antecedents, Development, and
Impact (1st edition., Vol. 6). Routledge.
Meredith-Karam, P., Kong, H., Wang, S., Zhao, J. (2021) The relationship between ridehailing and
public transit in Chicago: A comparison before and after COVID-19. J. Transp. Geogr. 97, 103219.
Metz, C. (2024). When Self-Driving Cars Don’t Actually Drive Themselves: Times Insider. New York
Times (Online).
Metz, D. (2018). Developing Policy for Urban Autonomous Vehicles: Impact on Congestion. Urban
Science, 2(2), 33.
Michalek, J., Armanios, D., Nock, D (2022) Ridehailing Service Equity in Normal and Rare
Conditions Mobility 21, USDOT.
Milakis, D. (2019). Long-term implications of automated vehicles: an introduction. Transport
Reviews, 39(1), 1–8.
Milakis, D., van Arem, B., & van Wee, B. (2017). Policy and society related implications of
automated driving: A review of literature and directions for future research. Journal of Intelligent
Transportation Systems, 21(4), 324–348.
265
Misa, T. J. (1988). How Machines Make History, and How Historians (And Others) Help Them to
Do So. Science, Technology, & Human Values, 13(3/4), 308–331.
Mohl, R. A. (1993). Race and space in the modern city: Interstate-95 and the black community in
Miami. In A. R. Hirsch & R. A. Mohl (Eds.), Urban policy in twentieth-century America (pp. 100–
158). New Brunswick, NJ: Rutgers University Press.
Mohl, R. A. (2004). Stop the Road: Freeway Revolts in American Cities. Journal of Urban
History, 30(5), 674–706.
Moretti, E. (2012). The new geography of jobs. Houghton Mifflin Harcourt
Mullin, K. and Pelosi, N. (2023, September 19). Letter from Members of Congress to Honorable
Ann Carlson, NHTSA Acting Administrator. https://kevinmullin.house.gov/sites/evosubsites/kevinmullin.house.gov/files/evo-media-document/3b080d17-b33e-4de0-a762-
e1dd005633ee.pdf
Nader, R. (1965) Unsafe at Any Speed: The Designed-In Dangers of the American Automobile,
Grossman Publishers.
Naumov, S., Keith, D. R., & Fine, C. H. (2020). Unintended Consequences of Automated Vehicles
and Pooling for Urban Transportation Systems. Production and Operations Management, 29(5),
1354–1371.
Nordhoff, S. (2024). A conceptual framework for automation disengagements. Scientific
Reports, 14(1), 8654–8654.
Norton, P. D. (2011). Fighting Traffic: The Dawn of the Motor Age in the American City (1st ed).
The MIT Press.
Norton, P. (2021). Autonorama: The Illusory Promise of High-Tech Driving (1st ed.). Island Press.
Nunes, A., Huh, L., Kagan, N., & Freeman, R. B. (2021). Estimating the energy impact of electric,
autonomous taxis: evidence from a select market. Environmental Research Letters, 16(9), 94036-.
O'Connell, J. & Myers, A. (1966) Safety Last: An Indictment of the Auto Industry. New York:
Random House.
Osman, T. & Maury-Holmes, S. (2022, February) An Analysis of App-Based Drivers in California.
UC Riverside School of Business, Center for Economic Forecasting and Development.
Othman, K. (2021) Public acceptance and perception of autonomous vehicles: a comprehensive
review. AI and Ethics, 1: 355-387.
Perkins L., Dupuis N., and Rainwater B. (2018) Autonomous Vehicle Pilots Across America –
Municipal Action Guide, National League of Cities
Pernicka, S. (2019). The disruption of taxi and limousine markets by digital platform corporations
in Western Europe and the United States: Responses of business associations, labor unions, and
266
other interest groups. Working paper. Retrieved from Institute for Research on Labor and
Employment, University of California, Los Angeles website: https://irle.ucla.edu/wpcontent/uploads/2019/06/Disruption-of-Taxi-and-Limousine-Markets.pdf
Peters, J. D. (2017). You Mean My Whole Fallacy Is Wrong. Representations (Berkeley,
Calif.), 140(1), 10–26.
Pinch, T. J., & Bijker, W. E. (1984). The Social Construction of Facts and Artefacts: Or How the
Sociology of Science and the Sociology of Technology Might Benefit Each Other. Social Studies of
Science, 14(3), 399–441.
Poel, I. van de. (2020). Three philosophical perspectives on the relation between technology and
society, and how they affect the current debate about artificial intelligence. Human Affairs
(Bratislava, Slovakia), 30(4), 499–511.
Porter, L. (2018). The autonomous vehicle revolution: Implications for planning. Planning Theory &
Practice, 19(5), 753–778.
Pourrahmani, Jaller, M., Maizlish, N., & Rodier, C. (2020). Health Impact Assessment of Connected
and Autonomous Vehicles in San Francisco, Bay Area. Transportation Research Record, 2674(10),
898–916.
Rainie, L., Funk, C., Anderson, M., & Tyson, A. (2022, March 17) Americans cautious about the
deployment of driverless cars. Pew Research.
https://www.pewresearch.org/internet/2022/03/17/americans-cautious-about-the-deploymentof-driverless-cars/
Randazzo, R. (2018, 11 December) A slashed tire, a pointed gun, bullies on the road: Why do
Waymo self-driving vans get so much hate? The Republic, AZCentral.com
https://www.azcentral.com/story/money/business/tech/2018/12/11/waymo-self-drivingvehicles-face-harassment-road-rage-phoenix-area/2198220002/
Randazzo, R. (2018, 18 December) Self-driving vehicles often harassed on valley roads. Arizona
Republic USA Today network.
Rayle, L., Dai, D., Chan, N., Cervero, R., & Shaheen, S. (2016). Just a better taxi? A survey-based
comparison of taxis, transit, and ridesourcing services in San Francisco. Transport Policy, 45, 168–
178.
Regional Development Plan, ‘Momentum 2050’ (2021) Maricopa Association of Governments
https://azmag.gov/Portals/0/Transportation/RTP/2022/RTP-Momentum-2050-v2.pdf
Rethinking Transportation 2020-2030: A Rethink X Sector Disruption Report (May 2017) Arbib,
James and Seba, Tony https://www.rncan.gc.ca/sites/www.nrcan.gc.ca/files/energy/energyresources/Rethinking_Transportation_2020-2030.pdf
Rip, A. & Kemp, R. (1998). Technological change. In S. Rayner & E.L. Malone (Eds.), Human choice
and climate change (pp. 327–399). Columbus, Ohio: Battelle Press.
267
Robillard, K. (2014, August 26) Plouffe at Uber A Tough Start Politico.
https://www.politico.com/story/2014/08/david-plouffe-uber-110366
Roeh, I., & Feldman, S. (1984). The rhetoric of numbers in front-page Journalism: How numbers
contribute to the melodramatic in the popular press. Text & Talk, 4(4), 347.
Romero, S. (2018, December 31) Wielding Rocks and Knives, Arizonans Attack Self-Driving Cars.
New York Times. https://www.nytimes.com/2018/12/31/us/waymo-self-driving-cars-arizonaattacks.html
Rose, M. H., & Mohl, R.A. (2012) Interstate: Highway politics and policy since 1939. University of
Tennessee Press.
Rouse, D. J. Henaghan, K. Coyner, L. Nisenson, & J. Jordan (2018) Preparing Communities for
Autonomous Vehicles. American Planning Association, Washington D.C., USA.
Rushkoff, D (2016) Throwing Rocks at the Google Bus. Portfolio/ Penguin (New York).
St. Clair, D. J. (1986). The motorization of American cities. New York: Praeger.
Saleem, F., & Head, L. (2017). Safety and mobility advances through: Maricopa County, Arizona's
SMARTDrive Program: A look at the anthem connected vehicle test bed. ITE Journal (Institute of
Transportation Engineers), 87(10), 27-32.
Seely, B. E. (1987). Building the American highway system: engineers as policy makers. Temple
University Press.
SF City Attorney (2023, August 17) San Francisco seeks reprieve from CPUC decision allowing
unfettered autonomous vehicle expansion. Press release.
https://www.sfcityattorney.org/2023/08/17/san-francisco-seeks-reprieve-from-cpuc-decisionallowing-unfettered-autonomous-vehicle-expansion/
SF Police Officers Association (2023, June 22) Letter to CPUC. Re: Provisional authorizations for
Waymo and Cruise for autonomous vehicle passenger service https://catimes.brightspotcdn.com/67/15/d1d49d394702af4879f766f35b74/police-letter.pdf
SFMTA and SFCTA (2020, January) Before the Public Utilities Commission of the state of
California. Initial opening comments of the SFMTA and SFCTA on ALJ ruling ordering parties to
comment on questions regarding CPUC regulation of autonomous vehicles. R.12-12-011
SFMTA and SFCTA (2022, September 21) Letter to NHTSA from Jeffrey Tumlin (Director of
Transportation, SFMTA) and Tilly Chang (Executive Director, SFCTA)
https://www.sfmta.com/media/37591/download?inline
SFMTA, SFCTA and City and County of San Francisco (2023, January 23) Letter to California Public
Utilities Commission Re.: Protest of Waymo LLC Tier 3 Advice Letter from Jeffrey Tumlin and
Nicole Bohn (Mayor’s Office on Disability, City and County of San Francisco).
https://www.sfmta.com/sites/default/files/reports-anddocuments/2023/01/2023.01.23_protest_of_waymo_llc_tier_3_advice_letter_0001.pdf
268
SFMTA, SFCTA and City and County of San Francisco (2023, January 23) Letter to California Public
Utilities Commission Re.: Protest of Cruise LLC Tier 2 Advice Letter from Jeffrey Tumlin and Nicole
Bohn (Mayor’s Office on Disability, City and County of San Francisco).
https://www.sfmta.com/sites/default/files/reports-anddocuments/2023/01/2023.01.25_ccsf_23.0125_cpuc_cruise_tier_2_advice_letter_protest_002.pdf
https://www.sfmta.com/sites/default/files/reports-anddocuments/2020/01/49._initial_opening_comments_of_the_sfmta_and_sfcta_on_alj_ruling_ordering_parties_to_comme
nt_on_questions_regarding_cpuc_regulation_of_autonomous_vehicles.pdf
Shapiro, R. and Yoda, I. (2023, July) Innovation highway: unlocking the social and economic
benefits of autonomous vehicles. US Chamber of Commerce Technology Engagement Centre.
https://www.uschamber.com/assets/documents/CTEC_InnovationHighwayReport_July23.pdf
Silver, D. (2019, January 3) Don't Throw Rocks At The Google Self-Driving Cars. Forbes
https://www.forbes.com/sites/davidsilver/2019/01/03/dont-throw-rocks-at-the-google-selfdriving-cars/
Simons, R. A., & Arancibia, A. (2020). Will driverless vehicle adoption cause more urban sprawl or
prevent it? The Journal of Sustainable Real Estate, 12(1), 51–68.
Smart Growth America (2022), Ranking the Most Dangerous Metro Areas. Online.
https://smartgrowthamerica.org/dangerous-by-design/ Smart highways an assessment of their
potential to improve travel : report to the Chairman,
Subcommittee on Transportation, Committee on Appropriations, U.S. Senate. (1991). The Office.
Soriano, B. (2018, August 14) California’s Autonomous Vehicle Testing Program, California DMV
https://catc.ca.gov/-/media/ctc-media/documents/tab-4-1-a11y.pdf
Soriano, B. C., Dougherty, S. L., Soublet, B. G., & Triepke, K. J. (n.d.). Autonomous Vehicles: A
Perspective from the California Department of Motor Vehicles. In Road Vehicle Automation (pp.
15–24). Springer International Publishing.
Soteropoulos, A., Berger, M., & Ciari, F. (2019). Impacts of automated vehicles on travel behaviour
and land use: an international review of modelling studies. Transport Reviews, 39(1), 29–49.
Sperling, D. (2018) Three Revolutions. Island Press.
Spurlock, C. A., Sears, J., Wong-Parodi, G., Walker, V., Jin, L., Taylor, M., Duvall, A., Gopal, A., &
Todd, A. (2019). Describing the users: Understanding adoption of and interest in shared,
electrified, and automated transportation in the San Francisco Bay Area. Transportation Research.
Part D, Transport and Environment, 71, 283–301.
Silver, C. (1984). Twentieth-century Richmond: Planning, politics, and race. Knoxville: University of
Tennessee Press.
269
State of California (2022, August) Driving the future: autonomous vehicles strategic framework
vision and guiding principles. https://calsta.ca.gov/-/media/calstamedia/documents/final_avsf_visionguidingprinciples-a11y.pdf
Stead, D., & Vaddadi, B. (2019). Automated vehicles and how they may affect urban form: A
review of recent scenario studies. Cities, 92, 125–133.
Stilgoe, J. (2023, February 24) Driverless cars: what we've learned from experiments in San
Francisco and Phoenix. The Conversation. https://www.ucl.ac.uk/news/2023/feb/opiniondriverless-cars-what-weve-learned-experiments-san-francisco-and-phoenix
Stilgoe, J. (2018). Seeing Like a Tesla: How Can We Anticipate Self-Driving
Worlds? Glocalism, 2017(3).
Stilgoe, J. (2020) Who’s Driving Innovation? Palgrave Macmillan, Switzerland.
Stilgoe, J., & Cohen, T. (2021). Rejecting acceptance: learning from public dialogue on self-driving
vehicles. Science & Public Policy, 48(6), 849–859.
Stilgoe, J., & Mladenović, M. (2022). The politics of autonomous vehicles. Humanities & Social
Sciences Communications, 9(1), 1–6.
Stopher, P.R., Magassy, T.B., Pendyala, R., McAslan, D., Arevalo, F.N., & Miller, T. (2021) An
evaluation of the Valley Metro-Waymo automated vehicle RideChoice mobility on demand
demonstration. (FTA Report No. 0198)
Strömberg, H., Ramos, É. M. S., Karlsson, M., Johansson, M., Ekman, F., Bligård, L.-O., & Bergstad,
C. J. (2021). A future without drivers? Comparing users’, urban planners’ and developers’
assumptions, hopes, and concerns about autonomous vehicles. European Transport Research
Review, 13(1), 1–12.
Sudhakar S, Sze V, Karaman S. (2023) Data Centers on Wheels: Emissions From Computing
Onboard Autonomous Vehicles. IEEE MICRO. 2023;43(1):29-39.
Talebian, A., & Mishra, S. (2018). Predicting the adoption of connected autonomous vehicles: A
new approach based on the theory of diffusion of innovations. Transportation Research. Part C,
Emerging Technologies, 95, 363–380.
Taylor, B. D. (2000). When finance leads planning: Urban planning, highway planning, and
metropolitan freeways in California. Journal of Planning, Education and Research, 20 (2), 196–214.
TRID (1994) What is IVHS IVHS Architecture Bulletin, Issue Number 1, IVHS America
https://trid.trb.org/view.aspx?id=389133
Tumlin, J. P. (2020, February 11) statement of Jeffrey P Tumlin Director of Transportation San
Francisco Municipal Transportation Agency on autonomous vehicles: promises and challenges of
evolving automotive technologies. Submitted to the US House of Representatives Committee on
Energy and Commerce Subcommittee on Consumer Protection and Commerce. February 11, 2020
https://docs.house.gov/meetings/IF/IF17/20200211/110513/HHRG-116-IF17-Wstate-TumlinJ20200211.pdf
270
Urry, J. (2008). Governance, flows, and the end of the car system? Global Environmental
Change, 18(3), 344.
USDoT (2016, September) Federal Automated Vehicles Policy: Accelerating the next revolution in
roadway safety. U.S. Department of Transportation.
https://www.transportation.gov/sites/dot.gov/files/docs/AV%20policy%20guidance%20PDF.pdf
USDoT (2021) An Evaluation of the Valley Metro–Waymo Automated Vehicle RideChoice Mobility
on Demand Demonstration, Final Report.
https://www.transit.dot.gov/sites/fta.dot.gov/files/2021-09/FTA0198-Research-ReportSummary-REVISED.pdf
Wachs, Martin., & Crawford, M. (1992). The Car and the city : the automobile, the built
environment, and daily urban life. University of Michigan Press.
Wang, K., & Akar, G. (2019). Factors Affecting the Adoption of Autonomous Vehicles for Commute
Trips: An Analysis with the 2015 and 2017 Puget Sound Travel Surveys. Transportation Research
Record, 2673(2), 13–25.
Wang, S., Jiang, Z., Noland, R. B., & Mondschein, A. S. (2020). Attitudes towards privately-owned
and shared autonomous vehicles. Transportation Research. Part F, Traffic Psychology and
Behaviour, 72, 297–306.
Ward, J. W., Michalek, J. J., Samaras, C., Azevedo, I. L., Henao, A., Rames, C., & Wenzel, T. (2021).
The impact of Uber and Lyft on vehicle ownership, fuel economy, and transit across U.S.
cities. iScience, 24(1), 101933–101933
Wasserman JL, Taylor BD. State of the BART: Analyzing the Determinants of Bay Area Rapid
Transit Use in the 2010s. Transportation research Part A, Policy and practice. 2023;172:103663-.
Webb, J., Wilson, C., & Kularatne, T. (2019). Will people accept shared autonomous electric
vehicles? A survey before and after receipt of the costs and benefits. Economic Analysis and
Policy, 61, 118–135.
Weber, B. S. (2019) Uber and urban crime, Transportation Research Part A: Policy and Practice,
Volume 130.
Weingroff, R. F. (1996, Summer) Federal Aid Road Act of 1916: Building The Foundation, Public
Roads https://highways.dot.gov/public-roads/summer-1996/federal-aid-road-act-1916-
building-foundation Accessed September 17, 2024
Weingroff, R. F. (nd) On the Road with Woodrow Wilson
https://www.fhwa.dot.gov/highwayhistory/wilson.pdf
Widen, W. H. & Koopman, P. (2023, September 7) Federal Action Is Needed to Maintain Trust in
Automated Vehicle Technology. JURIST – Academic Commentary,
https://www.jurist.org/commentary/2023/09/widen-koopman-federal-action-automatedvehicles/
271
Wolf, A. (2022, February 2) Testimony of Ariel Wolf, General Counsel, Autonomous Vehicle Industry
Association, to the US House of Representatives Committee on Transportation and Infrastructure
Subcommittee on Highways and Transit.
https://docs.house.gov/meetings/PW/PW12/20220202/114362/HHRG-117-PW12-Wstate-WolfA20220202.pdf
Wyatt, S. (2007) in Hackett, Amsterdamska, O., Lynch, M. E., Wajcman, J., Sismondo, S., Bijker, W.
E., Turner, S., Thorpe, C., Latour, B., & Clarke, A. E. (2007). Technological Determinism Is Dead;
Long Live Technological Determinism. In The Handbook of Science and Technology Studies. MIT
Press.
Yigitcanlar, T., Wilson, M., & Kamruzzaman, M. (2019). Disruptive impacts of automated driving
systems on the built environment and land use: An urban planner’s perspective. Journal of Open
Innovation, 5(2), 1–17.
Yin RK. (1994) Case Study Research : Design and Methods. 2nd ed. Sage Publications.
Yuen, K. F., Cai, L., Qi, G., & Wang, X. (2021). Factors influencing autonomous vehicle adoption: an
application of the technology acceptance model and innovation diffusion theory. Technology
Analysis & Strategic Management, 33(5), 505–519.
Yuen, K. F., Wong, Y. D., Ma, F., & Wang, X. (2020). The determinants of public acceptance of
autonomous vehicles: An innovation diffusion perspective. Journal of Cleaner Production, 270,
121904-.
Zmud, J., Sener, I. N., Wagner, J., (2016) Consumer Acceptance and Travel Behavior Impacts of
Automated Vehicles: Final Report. PRC 15-49 F . Texas A&M Transportation Institute
https://policy.tti.tamu.edu/technology/revolutionizing-our-roadways/prc-report-consumeracceptance-and-travel-behavior-impacts-of-automated-vehicles/
Abstract (if available)
Abstract
Transportation technologies of the past have been presented as solutions to one or more of the continuous challenges facing cities, especially improving road safety, reducing congestion or the environmental impact of vehicles. Such technologies though have failed to meet the hopes and expectations placed upon them. We have seen continued rising numbers of road deaths, rising congestion and continued environmental costs. This dissertation explores why transportation technologies have largely failed to solve these problems and how lessons of the past may explain and predict outcomes for the current “new” technology, automated vehicles.
Past technologies have failed in part because those who developed and promoted them prioritized goals other than mobility – their profits, their careers, or to find a fix for other social or economic problems – all to the end of selling the technology. Transportation technology has been promised to make life easier and more comfortable, generate wealth, create jobs, strengthen economies, along with making mobility safer, more efficient, and more environmentally friendly. Despite a century of failure, the public continues to believe that a world of safe, congestion-free, easy, and clean automobility is continually just around the corner. An ecosystem of people and organizations that benefit from the sale of cars continue to provide an illusion that it is.
This dissertation describes the history of the development of autonomous vehicles (AVs) and conducts two case studies, Phoenix, AZ and San Francisco CA. Using the lenses of theories of technology adoption and the history of prior transportation technologies, I find that the same dynamics and motivations of the past are playing out today, and they are coming from a similar coalition of interests. Claims for autonomous vehicles echo those used previously and are equally misleading. Negative impacts are overlooked, and losers are likely to be those who have lost out in the past. I conclude with some suggested ways forward. If developed, autonomous vehicle technology has the potential to reduce road crashes and save lives but unless we learn the lessons of the past and have an open and honest appraisal of AVs’ likely impact in the future, we will find ourselves driving in circles.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The long-term impact of COVID-19 on commute, employment, housing, and environment in the post-pandemic era
PDF
Spatial and temporal expenditure-pricing equity of rail transit fare policies
PDF
The impact of demographic shifts on automobile travel in the United States: three empirical essays
PDF
Unraveling decentralization of warehousing and distribution centers: three essays
PDF
Healthy mobility: untangling the relationships between the built environment, travel behavior, and environmental health
PDF
The development implications of China’s Belt and Road Initiative for Russia, Kazakhstan and Belarus
PDF
Location of warehouses and environmental justice: Three essays
PDF
Essays on congestion, agglomeration, and urban spatial structure
PDF
An application of aerial drones in high definition mapping for autonomous vehicles
PDF
Urban air pollution and environmental justice: three essays
PDF
The demand for reliable travel: evidence from Los Angeles, and implications for public transit policy
PDF
Environmental justice in real estate, public services, and policy
PDF
The use of mobile technology and mobile applications as the next paradigm in development: can it be a game-changer in development for women in rural Afghanistan?
PDF
Active travel, outdoor leisure, and neighborhood environment: path analysis, Los Angeles County
PDF
An examination of the law enforcement response to active shooter incidents in U.S. cities: findings and implications for improving future response efforts
PDF
Productive frictions and urbanism in transition: planning lessons from traffic flows and urban street life in Ho Chi Minh City, Vietnam
PDF
Avoiding middle-class planning 2.0: media arts and the future of urban planning
PDF
Community development agreements: addressing inequality through urban development projects
PDF
Urban spatial transformation and job accessibility: spatial mismatch hypothesis revisited
PDF
Health impact assessment, the concept, science, and application in China
Asset Metadata
Creator
Coombes, Alanna Clare Louise
(author)
Core Title
The history of autonomous vehicle development and its likely futures and consequences
School
School of Policy, Planning and Development
Degree
Doctor of Policy, Planning & Development
Degree Program
Urban Planning and Development
Degree Conferral Date
2024-12
Publication Date
01/14/2025
Defense Date
10/18/2024
Publisher
Los Angeles, California
(original),
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
autonomous vehicles,driverless cars,history of technology,new mobility,OAI-PMH Harvest,science and technology in society,transportation technology
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Giuliano, Genevieve (
committee chair
), Norton, Peter (
committee member
), Sloane, David (
committee member
)
Creator Email
aclcoombes@gmail.com,coombes@usc.edu
Unique identifier
UC11399FAXM
Identifier
etd-CoombesAla-13755.pdf (filename)
Legacy Identifier
etd-CoombesAla-13755
Document Type
Dissertation
Format
theses (aat)
Rights
Coombes, Alanna Clare Louise
Internet Media Type
application/pdf
Type
texts
Source
20250115-usctheses-batch-1235
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
autonomous vehicles
driverless cars
history of technology
new mobility
science and technology in society
transportation technology