Socioeconomic Impacts of Automated and Connected Vehicles (2018)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

62 APPENDIX B Briefing Papers on Exploratory Topics Alexandra Millonig, AIT Austrian Institute of Technology, Vienna, Austria Susan Shaheen, University of California, Berkeley, USA Barbara Lenz, German Aerospace Center, Berlin, Germany Barry Einsig, Cisco Systems, Inc. (currently at CAVita, Washington, D.C.) Alex Karner, University of Texas, Austin, USA Marcin Stepniak, Polish Academy of Sciences, Warsaw, Poland Satu Innamaa, VTT Technical Research Centre of Finland Ltd., Helsinki, Finland Matthew W. Daus, Windels Marx Lane and Mittendorf, LLP, New York, USA customize their scenarios in varying degrees but still offer two opposing views based on the same scenario vectors of market penetration levels and public policy contexts, as described below. In the scenarios, we focus on different CAV market penetration levels (strong and weak forces) and public policy contexts (high and low government regulation). We present two contrasting scenarios to launch our dis- cussions for each subtopic and each of the cross-cutting issues: • Economics and the workforce, • Equity, • Data access and privacy, and • Safety and security. The first scenario is primarily market driven with little regulation. The second reflects a more highly regu- lated world in which the public is much less accepting of automation. Not surprisingly, it is also much less market driven. Scenarios and Context As noted, we have created two scenarios to assist us in assessing the impacts of CAVs in the future. It is important to note that our breakout discussions are not restricted to these worlds, which reflect high and full Overview Alexandra Millonig and Susan Shaheen Introduction This briefing paper overview sets the stage for our break- out group discussions as part of the EU-U.S. Sympo- sium on the Socioeconomic Impacts of Automated and Connected Vehicles (CAVs). In this overview, we describe two sample scenarios that could greatly impact four key subtopics related to the socioeconomic impacts of CAVs: • Travel behavior, • Freight, • Land use, and • Policy/governance (stakeholder response). Please note that the travel behavior and land-use breakouts use these same scenarios. Both the freight and policy/governance1 (stakeholder response) subtopics 1 The freight subtopic defines the two opposing scenarios as: (1) “Auto- mated Vehicles Taking Over Transport Business” and (2) “Automated Roadway Freight Restrained by Policy.” The policy/governance (stake- holder response) subtopic defines its two opposing scenarios as: (1) “CAVs on the Rise” and (2) “CAVs Tamed by Policy.”

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 6 3 levels of automation (Levels 4 and 5).2 These scenarios are intended to accelerate our discussions in each of the breakout discussions and to aid us in developing high priority research problem statements in the four subtopic areas: (1) travel behavior, (2) freight, (3) land use, and (4) policy/governance, which also address the four cross-cutting issues: (1) economics and workforce issues, (2) equity, (3) data access, and (4) safety and security. We define our two framing and contrasting scenarios (travel behavior and land-use breakouts) as: (1) “CAVs on the Rise” and (2) “CAVs Tamed by Policy and Peo- ple.” Please see Figure 1, which positions these scenarios in opposite quadrants based upon two axes: (1) high/ low government regulation and (2) market forces strong/ weak. While these scenarios focus on the highest levels of automation, we should also think about the transition to highly automated vehicles as part of our discussions and in our research problem statement formulation. Scenario 1: CAVs on the Rise CAV services are offered by different competing private companies or private people using their vehicles as an additional source of income. CAVs come in different sizes and models for different purposes (e.g., business vehicles for working, entertainment vehicles) and can be booked at lower rates for regular and/or shared trips or specifically hailed causing higher costs, as fleet manage- ment must reschedule trips to serve new requests. Sharing is common for saving costs, but some groups refrain from using shared services with strangers (especially at night), although vehicles are equipped with a surveillance system and users are identifiable—there is even a rating system for co-riders. Almost everyone is using CAVs also for short trips, even if the traffic is congested or slow, as it is very con- venient and time can be spent for different purposes. As the demand for other transport modes has been dramat- ically decreasing, public transport has been reduced and few people use bikes or walk. However, several areas in the city have been reserved for pedestrians and micro- vehicles (like hoverboards and electro scooters) for recreational and retail purposes, although most people rarely go shopping as most purchases are made online. 2 The Society of Automotive Engineers (SAE) has defined five levels of vehicle automation, with Level 1 signifying vehicles that automate only one primary control function (e.g., self-parking or adaptive cruise control) and Level 5 referring to vehicles capable of driving in all environments without human control (U.S. Department of Transportation 2016]. The majority of AV pilots thus far are target- ing Level 4 automation, where a human operator does not need to control the vehicle as long as it is operating in a suitable operational design domain (ODD) given its capabilities. (Stocker and Shaheen 2018, Forthcoming). Market Forces Strong Market Forces Weak Scenario 2: CAVs Tamed by Policy & People --More digital substitution of travel and lower per capita travel --Profound changes in travel demand also result from strong regulatory policy --Road pricing is sophisticated and effective --CAVs have not been widely accepted by the public Low Government Regulation High Government Regulation Scenario 1: CAVs on The Rise --Changing face of retail and healthcare --CAVs, SAVs, and UAM (private/freight) growing --Increased congestion and VMT result --Benefits for older adults (e.g., medical appointments) --Freight and retail make use of CAVs for long-haul freight and delivery to customers FIGURE 1 Travel behavior impacts of automated and connected vehicles (Millonig and Shaheen, see briefing paper on Exploratory Topic 3).

6 4 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S Outside the cities, there is practically no other transport mode used besides CAVs. Scenario 2: CAVs Tamed by Policy and People Due to international commitments to dramatically improve climate conditions and reduce the consumption of land and resources, governments have implemented effective measures on transportation to significantly decrease traffic volumes (e.g., road pricing). This development is supported by considerable changes in the production sector (e.g., emerging technologies like automation and 3D-printing accelerate local productions and teleworking becomes common). Further, a paradigm shift occurs towards more sustainable, local lifestyles. This results in a landscape of smaller communities where people spend most of their time, but they are highly connected to other parts of the country through telecommunication. If possible, people follow their daily routines in the close vicinity, working in local community centers offer- ing teleworking spaces and covering the majority of the distances by walking or using shared bikes or micro- vehicles (e.g., hoverboards and electro scooters). Smaller CAVs are only rarely used for passenger transport (e.g., by people with disabilities or to reach larger hubs for mass transportation) and are strongly regulated as the use is only granted to specific groups, but are common for transporting goods, as people prefer to get their pur- chases delivered than carry them. For longer distances, automated public mass transport is used, but people usually avoid having to travel longer distances on a regu- lar basis. Only a small elite employ private AVs, mainly to display their status (as these are very expensive) and distinguish themselves from other citizens. explOratOry tOpic 1 SyntheSiS Of the SOciOecOnOmic effectS Of cOnnected and autOmated vehicleS and Shared mObility-freight Barbara Lenz and Barry Einsig Introduction This briefing paper focuses on freight—both long-haul goods transport and urban or regional delivery. The fun- damentally different situation for long-haul and urban delivery represents a particularly complex challenge when designing a “world of road automation.” While keeping this challenge in mind we oriented our thoughts along the public policy context (high and low government regulation) on the one hand, and the opportunities as perceived by actors on the markets on the other (fast and decelerated implementation). We defined two scenarios to launch our discussions for each of the cross-cutting issues: economics and the workforce, equity, data access and privacy, and safety and security. Scenarios and Context The purpose of the scenarios is to assist us in assess- ing the impacts of vehicle automation in the future. We defined our scenarios using a contrasting approach: (1) automated vehicles taking over transport business and (2) automated road freight restrained by policy. Figure 2 positions these scenarios in opposite quadrants based upon two axes: high/low government regulation and strong/weak. We provide a brief discussion of each scenario and outline research questions based on the four cross-cutting issues related to freight impacts. Scenario 1: Automated Vehicles Taking Over Transport Business Automated trucks have become available since the early 2020s, and were quickly adopted by logistics providers, because they dramatically reduced the need to hire driving staff. The implementation produced need for additional infrastructure and maintenance of existing infrastructure; motorways were reconfigured by allocating one track for automated trucks exclusively. National and state funding was provided to build new infrastructures to ease “mixed use” of motorways by fully automated and non-automated vehicles. Investments went in particular in infrastructures that allowed other motorway users to enter and exit the motorways safely. Meanwhile this problem has been solved by software-based solutions as all motor vehicles are automated. As automation provides high flexibility for transport times and modalities, most logistics providers are now relying on an automated truck fleet either as owners or charterers. Additionally they use all kinds of automated delivery vehicles (including drones). The ongoing imple- mentation of automation in production and commerce has massively pushed road freight automation to create fully automated supply chains. Road freight automation is not well accepted by soci- ety as vehicle automation has resulted in freight trans- port shifting predominantly to roads. Rail freight was no longer competitive, so that rail lines were transformed into roads for automated long-haul by truck. While freight operations became less costly, they increased the environmental load coming from goods transport. In the following, we explore possible impacts resulting from automation of freight across the four dimensions (1) economics and the workforce, (2) equity, (3) data access and privacy, and (4) safety and security. For each area, we pose key questions related to associated impacts of the first scenario.

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 6 5 Economics and Workforce In the freight sector, automation may require consider- able investment from the state. For logistics service pro- viders, it is connected with higher vehicle cost. Higher vehicle cost and reintegration of supply chains will impact small and independent trucking and distribu- tion companies who are no longer competitive, and the reduction in transport cost will stimulate the rise of oli- gopolies. Will the state be able and willing to invest in new infrastructures? Will society accept that the state pays for investments from which only a small group of companies will benefit? Or will new business models come up to provide automation-ready infrastructure requiring motorway tolls? At the same time road freight automation will cause tremendous job losses among drivers. What could be alternatives to these jobs in the short and in the medium run, but also to job losses that will result from automa- tion along the entire supply chain? What about alter- natives for jobs that are currently held by people with lower levels of education and training? Which are the jobs required by an automated freight system? Will the education system be able to provide the required educa- tion and training? And what about jobs and professions related to roles such as planners, traffic operations, parking authorities? Will Artificial Intelligence and Machine Learning from vehicles communicating all the time eliminate these roles currently held by humans? Will the lower cost of trans- portation cause manufacturers to move further away from populated areas, what will the impact be? Equity New road freight corridors will arise and existing ones will carry more shipments resulting in increasing volumes of noise and air pollution. Off-peak delivery into urban areas could disrupt the quality of life, either night life or sleep. Which are the options to avoid a negative environ- mental burden for those living close to road infrastruc- tures or areas of frequent deliveries? As automation of freight and deliveries might require specific infrastruc- tures for those who want or need home delivery, there will be residential areas where delivery is not or no longer Market Forces Strong Market Forces Weak Low Government Regulation High Government Regulation Scenario 2: Road Freight Restrained by Policy - Governments impose rail and waterways shares in modal split - Automation in rail and waterways extends capabilities of networks - Cities provide specific hubs and corridors - Significant job losses, governmental support for exploring alternatives - Logistics providers reorganize delivery fleets with automated trucks and additionally use all kinds of automated delivery vehicles (incl. drones) - Cities restrict automated delivery to areas with automated hubs and specific corridors - New infrastructures in housing areas allow for full automation of deliveries. Scenario 1: Automated vehicles taking over transport business - Automation provides high flexibility for transport : Logistics providers reorganize long-haul fleets with automated trucks - Freight transport shifts completely to road -> need for additional infrastructure - Automated freight strongly interacts with automation of production and commerce - Higher vehicle cost, but reduction in transport cost stimulates the rise of oligopolies - Tremendous job losses, little immediate alternatives - Automated freight strongly interacts with automation of production and wholesale trade FIGURE 2 Impacts of automated and connected vehicles on freight.

6 6 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S possible. Does this mean that delivery deserts will add to already existing retail and food deserts? Data Access and Privacy Smaller firms might lose access to supply-chain data thus no longer being able to collaborate with other companies in the chain. Since the value of the digital information is more valuable to the manufacturer or larger retailer the cost for the implementation may be forced on the smaller trucking company or distribution facility while the economic gains and customer intimacy go to the large manufacturing or retailer without any cost recov- ery available to the smaller players in the ecosystem. Access to data will probably be part of the freight busi- ness. Will small producers, dealers and logistics provid- ers still have the opportunity to be part of the business? Or will large e-commerce platforms become the control center of the entire chain thus applying full domination on production, commerce and consumption? Can block- chain help secure the route of trust in the supply chain and allow all the stakeholders to participate equitably? Will this force consolidation of the supply chain? Could this ultimately drive up transport cost for freight due to the limited options available? May this eliminate par- ticular elements of the market (e.g., wholesale)? Safety and Security If automation takes over too soon, it may cause safety issues and lead to resistance against freight automa- tion. How can it be ascertained that software that is used for freight automation provides safe operations for both the driving and the loading and unloading? How can current regulations (e.g., for control of load by police during transport) be “translated” to the case of automation? For instance, is it possible for the police to stop a truck to control the load during its cruise on the motorway? Will we be able to monitor and control smuggling and illegal imports and exports with no humans in the supply chain? Who has access and who can get access to the data that go with freight during transport? How can control about the security of goods be achieved? Scenario 2: Automated Road Freight Restrained by Policy Acknowledging that automation of road freight will considerably lower the cost for road freight, thus leading to a complete shift of freight transport to road, govern- ments started early to control and modulate the process. To keep or extend the modal split towards environment- friendlier modes like rail and waterways capabilities of their networks were extended, automation on rail and waterways was heavily subsidized, and larger cities and agglomerations provided specific intermodal hubs and corridors that included facilities for automated oper- ation. Shippers have to follow specific rules to keep an environment-friendly modal split. States and cities have severe regulation for freight automation to make it a sustainable mode of transport i.e., environment-friendly, safe and efficient. As auto- mated logistics are operating with the highest respect possible towards other road users, including pedestrians and cyclists, children and seniors, it is broadly accepted also in cities and city centers. By creating connectiv- ity systems cities can allow check-in and check-out of delivery vehicles and schedule drop-off times and loca- tions reducing double parking and congestion in the city center. This has helped to further extend home deliver- ies across cities and regions. New business models have emerged for dual use of automated vehicles: commuters during peak commute times and freight delivery during off-peak times. Another important factor to generate acceptance within the population for the automation of freight was the better use of resources such as alternative power by automated vehicles, and the obligation to run automated vehicles zero emission either electrically, CNG, or Fuel Cell thus contributing to cleaner air and less CO2. In the following, we explore—similar to the exercise for Scenario 1—possible impacts resulting from automa- tion of freight across the four dimensions (1) econom- ics and the workforce, (2) equity, (3) data access and privacy, and (4) safety and security. For each area, we pose key questions related to associated impacts of the second scenario. Economics and Workforce Full automation will need regulation to give enough time to make an impact that labor can be reskilled to better quality of life positions. How much will actors in the field of automation—firms, states, cities and citizens— accept any deceleration of automation and its opportu- nities for new businesses, but also for a new transport system? Which are short and medium term alternatives for those who are already working in the sector? What needs should be formulated towards education and training? What will be the effect of automation on regional cir- cular economy wherever it exists? Could the dependency on automation empower local restraint of trade and force urban areas’ preference for locally made, grown, or developed products and delivery system?

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 6 7 Equity The reduction of transport cost for manufactured goods by automated freight can increase the buying power of all including poor and middle class. What would this mean for household consumption as well as the pro- duction and trade of consumer products? Can automa- tion be made a driver of overall economic development including all segments of the population? Equity issues could arise if all of the preference is shown for delivery vehicles over other vehicle types or part of the urban environment that is impoverished or outside the local norms or social and political views. Will strong government regulations and significant value creation in the private sector create issues with corrup- tion in the way access is granted by governments to pri- vate sector? Can we reimagine a localized logistics and supply chain so as to bring smaller more local focused distribution and help eliminate food and services des- erts? Can we use this to free up land in the urban areas that can lead to better utilization for equitable housing, healthcare, or education? How can regulation ensure that private issues are estimated equivalent to business issues? Can we create an equitable system for curb access for freight delivery, with other shared mobility services? Data Access and Privacy Governments of all sizes from cities, regional states, provinces, and federal agencies understand the need to be able to exchange data between the many actors in the supply chain and still be compliant with General Data Protection Regulation (GDPR) or other privacy provi- sions as well as proprietary data from the private sector, and so they create a data broker and exchange to enable the connected and automated vehicles. Much of the data that the private sector has, regarding product mix, ship- ping times, and even end customers, is very proprietary financial information to the manufacturer or retailer. Whose data should be included in this broker/exchange? Who should operate this system and why? Can the gov- ernment collect enough of its own valuable data to share and make it an equitable exchange that the private sector will collaborate? How can we secure this data? How will it be exchanged? Will real-time safety and mobility data be enough for the private sector to see the clear advan- tage? Can enabling these systems create a competitive advantage for regional governments and cities that will differentiate them in the areas of throughput, freight, and velocity from ports, rail, and trucking to retail or indi- vidual consumers? Can we build an exchange/brokerage that is beneficial enough for the private companies to want to participate while secure enough to protect the critical supply-chain data of the private companies so that their information cannot get out and have a negative impact on their financial performance (i.e., stock share price or leak out to a competitor to give an advantage to them by exposing shipping dates, quantities, end customers or other key financial metrics)? Safety and Security Automated freight vehicles would have to obey laws and enable digital inspections allowing regulators and pub- lic safety officials to focus on more positive impacting behaviors. Which actors and aspects should be consid- ered then? Which kind of inspections will still be needed given the fact that the entire supply chain is “digitized,” including information about origin and destination of parts and products? Will the public trust these systems for mixed traffic? Automation of urban delivery systems could create a safer environment. Shift to automated freight and pla- tooning might reduce congestion and fatalities on major intercity relations. How does the operation of vehicles and infrastructures need to be designed to meet this goal? How must—for the “urban case”—delivery vehi- cles be designed so that pedestrians, the disabled, and other “vulnerable road users” are accounted for? explOratOry tOpic 2 placeS where peOple live, wOrk, and play Alex Karner and Marcin Stepniak Introduction This paper addresses potential land-use changes that are likely to arise under two very different transportation automation futures. Land use generally refers to the type of structure or allowed activities located in a particular geographic area and their intensity (e.g., residential den- sity, employment density, square feet of retail space). We concentrate on the possible evolution of the places where people live (residential locations), work (employment zones), and play (recreation and entertain- ment locations). Historically, changes in transportation technology have preceded substantial changes in the urban form. Faster modes have facilitated the coverage of ever further distances within fixed travel time bud- gets, extending the scale of human settlement from city to region to megaregion. Following the common structure of all briefing papers for the symposium, we discuss possible socioeconomic impacts using two contrasting scenarios of the deploy- ment of automation related to the four cross-cutting themes. We propose that in Scenario 1, the widespread implementation of CAVs will drive travel costs to near- zero, reducing the importance of specific residential or

6 8 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S workplace locations. On the other hand, the reduction of vehicle use envisioned in Scenario 2 will increase the importance of place. In Scenario 1, access to CAVs is a key factor, whereas in Scenario 2, access to places is of primary importance. Scenario 1: CAVs on the Rise The first scenario assumes spectacular, unregulated development of CAVs driven by strong market forces. It leads to the development of a rich portfolio of CAVs of many sizes, with multiple available models serving many purposes. The widespread public acceptance of CAVs means that they are widely used and completely replace traditional vehicles in short order. This trans- port revolution is likely to deeply transform land use in urban, suburban, and rural areas. The complete domination of CAVs as the main trans- port mode means that they will be used even for short trips where previously non-motorized modes may have been attractive. Increased travel comfort, safety, and the ability to use travel time productively will engender a significant increase in trip lengths as the effective cost of travel tends towards zero. The confluence of these factors means that the footprints of regions are likely to expand, increasing demand for land on the periphery of regions for residential purposes. At the same time, city centers are likely to be substantially affected by reduced parking demand, enabling a higher density of economic activities. The space saved from parking in central busi- ness districts is likely to be transformed into other uses. However, even shared CAVs require areas for charging, cleaning, maintenance, and parking during low demand periods (e.g., off-peak and overnight). This means that questions remain about the future location and scale of parking in urban areas. Moreover, CAVs are likely to increase demand for curbside drop-off and pickup areas which might partly counterbalance reductions in park- ing requirements. Economics and Workforce The widespread uptake of CAVs envisioned under Scenario 1 will deeply affect the location choices of businesses. Specifically, an increased concentration of economic activity in city centers might be observed, thanks to repurposing land previously used for parking. Rather than needing to park close to a destination, CAVs could easily travel to find parking further away or in centralized locations relatively far from downtown. In the absence of major increases in roadway capacity, the result of this shift is likely to be crippling congestion as more people demand travel to downtown destinations and as CAVs make more empty trips. Although the cost of travel will be reduced, travel time budgets are likely to have a practical limit. As congestion increases, public transit operating on dedicated rights of way becomes a more viable alternative. These possible outcomes raise multiple questions related to the evolution of the future city. Under Scenario 1, cities might evolve in more or less transit- and non-motorized-friendly directions. How will public transit availability play into firms’ location choices under Scenario 1? On the other hand, the reallocation and concentra- tion of retail and service centers may occur. The expected increase in the comfort of travel will favor longer trips to “service hubs,” which offer more complex services rather than shorter trips to several small services with more limited offerings. As a result, small- and medium- sized retail and service centers might disappear. Because working while traveling will become the norm, a fur- ther dissolution between personal and work time will be observed. Would these changes provoke the devel- opment of more polycentric cities (e.g., due to increas- ingly congested city centers) or, on the contrary, boost a monocentric urban structure? Equity Widespread CAV adoption and implementation has a great potential to influence social, economic and spatial equity. On the one hand, CAVs should enhance acces- sibility for people without access to cars and those who are unable or unwilling to drive (e.g., the young, older adults or people with disabilities, etc.) as they will be able to use CAVs (shared or owned). On the other hand, this enhanced mobility will still have associated costs and will likely require access to mobile technology, rais- ing questions about the digital divide. If public transit services are scaled back and newly available CAV ser- vices are higher priced, vulnerable social groups could be placed at a disadvantage. Issues related to shared ver- sus private ownership of CAVs must also be addressed. Thus, ensuring an adequate level of access to CAVs will be crucially important from an equity perspective in Sce- nario 1. Further, the widespread adoption of CAVs will have an impact not only on business location choices, but also on residential location choices. One of the possi- ble scenarios leads towards demographically homoge- neous residential areas, as particular social groups can more easily concentrate separately from others as they accept longer (but comfortable) travel (to jobs, services or recreation areas). Finally, full dependency on CAVs might also limit access to recreation and entertainment areas for vulnerable groups. They might be affected by limited access to CAVs (due to limited public transport options) or they may face discrimination due to a hard- wired inability to reach desired destinations (e.g., lack of

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 6 9 a “permission to drive” CAVs to particular destinations or through particular areas). To what extent are these scenarios possible? Which social groups might be (nega- tively) affected by the deployment of these scenarios? What policies can be designed to avoid discrimination and segregation of vulnerable groups? Safety and Security A full deployment of CAVs will significantly increase safety due to theoretically full protection against human- error-based traffic accidents. Nevertheless, the increase in traffic safety might not be spread evenly among all areas. Places with low CAV access, where walking and bicycling might prevail, could be at greater risk than others for injury or death. Are these differences fair? Further, the induced travel demand resulting from CAV implementation might significantly limit the walkability of residential and office areas. Moreover, some transport users might prefer to avoid shared modes when travel- ing from particular locations (or time periods). Thus the “security of a place” (e.g., residential or employment areas) might be related to the modes connecting particu- lar zones. Will travel security (or personal security while traveling) become a factor that stratifies future space and societies? Data Access and Privacy The massive implementation of CAVs would facili- tate the integration of personally identifiable informa- tion with information about residential, workplace, and transportation-related data (e.g., departure times, modes used, traveling companions). Based on this information, the precise characteristics of any indi- vidual could be known in principle, including histori- cal activity locations and real-time whereabouts. This data situation presents a great risk as it might provoke discriminatory policies from the state (e.g., in case of individuals/behaviors not supported by a current gov- ernment), as well as the private sector (e.g., in case of more/less prospective clients). As a result, selected resi- dential areas might be excluded from CAV services and their inhabitants may face serious limitations in their mobility. However, it might also provide an opportu- nity to provide specialized transport services directly to the areas where and when they are needed most. Again, the key policy question is what are the regulatory and policy frameworks needed to ensure that new oppor- tunities are seized, while avoiding the risks associated with potential discriminatory policies? Another potential data access and privacy impact relates to land values. The detailed knowledge of who is attracted to a given area (and when) is a powerful tool for both land management and land speculation. Thus, key questions are which CAV data might be used for land-use management, who should have access to them, and how should they be protected? Scenario 2: CAVs Tamed by Policy and People Scenario 2 assumes that overall traffic volumes would be reduced because of a renewed effort to mitigate environ- mental impacts and create dense communities where most needs can be met nearby using non-motorized modes or public transit. Rather than individual CAVs, smaller, slower, and shared multimodal AVs would become the norm (e.g., smaller CAVs, electric scooters). In general, CAVs would be used for goods movement rather than passenger travel. The latter would be limited to connections between main hubs and would have limited applicability to commuting trips. This scenario would entail a major change in prevail- ing land-use regimes in the United States where an ethos of “local control” currently dominates. It envisions sub- stantial government intervention in land-use decisions. Transit oriented development and the domination of “small communities” would change the character of places across the land-use spectrum. Thus, the overall density of residential areas would increase, together with their supply of basic services, while teleworking supple- mented by local employment centers would dominate the future economy. But without additional intervention to curtail prevailing patterns of segregation, Scenario 2 would also entail substantial equity issues, as discussed further below. Economics and Workforce The telework-dominated future envisioned under Sce- nario 2 raises multiple questions related to economics and the workforce. For most workers, commute times and distances will be dramatically reduced. Additionally, the reduced emphasis on vehicle travel will likely result in improved levels of well-being and health across the population as non-motorized modes increase in popu- larity and utility. Because not all jobs are well-suited to telecommuting, there will also be some negative economic and work- force effects under Scenario 2. The societal acceptance of telework as a viable option is likely to vary substantially across and within economic sectors. Substantial low- wage work is concentrated in the service and manual labor sectors. Jobs like landscaping, construction, food preparation, and others will still need to occur in spe- cific locations. How would the reduction of travel sup- ply affect economic sectors that require employment to occur in specific locations? How would commuting trips for workers in these sectors be organized?

7 0 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S Equity It is unlikely that the relocalization envisioned under Sce- nario 2 will result in the reversal or erasure of racial- and income-based segregation or employment discrimination that prevailed throughout the 20th century in locations around the world. This means that without explicit poli- cies aimed at integration and inclusion, segregation is likely to continue. Specifically, some areas will be rela- tively prosperous, with high shares of telecommuters, strong concentrations of local services, and increasingly livable environments. Less prosperous areas will be pop- ulated by those unable to telecommute who will likely need to commute into the more prosperous areas for work. Any limitation on transport options or vehicles raises the possibility of discrimination against people with disabilities. What public policies will most effectively combat these residential and employment location patterns? Under what conditions will the public sector be willing to intervene to ensure progress towards integration? The key equity-related question that runs through each of the cross-cutting themes for Scenario 2 is that non-CAV policies will be required to ensure that the socioeconomic benefits of automation are not felt by a select few but are more broadly shared across the economy. Safety and Security The research on neighborhood effects demonstrates the profound impacts of an individual’s residential setting on public health, well-being, and related outcomes. In high-opportunity areas, schools are high quality, services are located nearby, crime rates are low, and strong social ties prevail. Those located in Scenario 2’s newly compact and convenient areas, generated through increased regu- lation and public skepticism will likely enjoy improved traffic safety as the “safety in numbers” effect would increasingly dominate pedestrian and cyclist safety out- comes. What about traffic safety in areas populated by non-telecommuting workers (with the continued domi- nance of vehicle travel)? Should we expect spatial dis- parities in traffic safety outcomes between particular locations? What kind of policy should be designed to limit the potential negative outcomes of these differences? The land-use patterns that would prevail under Sce- nario 2 also present issues from a disaster preparedness perspective. If CAVs are in short supply, large-scale evac- uations would become difficult. Of course, in a strong land-use-regulatory environment it’s likely that govern- ments would have the power to ensure that development only emerged in locations in which evacuations would be feasible. Which areas would need this kind of treat- ment or consideration and what kind of actions should government undertake? Data Access and Privacy In Scenario 2, limited CAV use would result in the col- lection of much less personal and location data than in Scenario 1. Nevertheless, a data-related threat to privacy will still exist. Limited options for long- and medium- distance trips will lead to data collection on the positions of particular individuals, even though they would not be as precise and complex as in Scenario 1. What kinds of threats are related to the collection of data through places? To what extent will they differ from those data collected via CAV tracking? The strong land-use regulations inherent in Scenario 2 also likely entail significant government intervention in the private housing market. One result could be the centralization of personally identifiable information in large databases held by a single agency. These data would potentially be vulnerable to compromise. Given the strong environmental ethic that underlies Scenario 2, how can individual freedom and data about where people live, work, and play be protected given the environmental and quality of life imperatives embodied in Scenario 2? explOratOry tOpic 3 cOnnected and autOmated vehicleS and travel behaviOr impactS Alexandra Millonig and Susan Shaheen Introduction This briefing paper focuses on travel behavior as it relates to the automation of the transport of goods and people. In this paper, we describe two scenarios that could greatly affect travel behavior given different connected and auto- mated vehicle (CAV) market penetration levels and policy contexts. We present two contrasting scenarios to drive our examination of travel behavior impacts across four cross-cutting issues: (1) economics and the workforce, (2) equity, (3) data access and privacy, and (4) safety and security. The first scenario is primarily market driven with little regulation. The second reflects a more highly regu- lated world in which the public is much less accepting of automation; not surprisingly, it is also much less market driven. Below we describe our scenario approach and outline key questions across the cross-cutting themes. Scenarios As noted, we have created two scenarios to assist us in assessing the travel behavior impacts of automated and connected vehicles in the future. We define these scenarios as: (1) “CAVs on the Rise” and (2) “CAVs Tamed by Policy and People.” Please see Figure 1,

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 7 1 which positions these scenarios in opposite quadrants based upon two axes: (1) high/low government regula- tion and (2) strong/weak market forces. We provide a brief discussion of each scenario and outline numerous questions based on the four cross-cutting issues related to travel behavior impacts. Scenario 1: CAVs on the Rise CAV services are offered by different competing private companies or private people using their vehicles as an additional source of income. CAVs come in different sizes and models for different purposes (e.g., business vehicles for working, entertainment vehicles) and can be booked at lower rates for regular and/or shared trips or specifically hailed causing higher costs, as fleet manage- ment must reschedule trips to serve new requests. Shar- ing is common for saving costs, but some groups refrain from using shared services with strangers (especially at night), although vehicles are equipped with a surveil- lance system and users are identifiable—there is even a rating system for co-riders. Almost everyone is using CAVs also for short trips, even if the traffic is congested or slow, as it is very con- venient and time can be spent for different purposes. As the demand for other transport modes has been dra- matically decreasing, public transport has been reduced and only a few people use bikes or walk. However, sev- eral areas in the city have been reserved for pedestrians and micro-vehicles (like hoverboards and electro scoot- ers) for recreational and retail purposes, although most people rarely go shopping as most purchases are made online. Outside the cities, there is practically no other transport mode used besides CAVs. In this scenario, “CAVs on the Rise,” we explore pos- sible travel behavior impacts across four dimensions: (1) economics and the workforce, (2) equity, (3) data access and privacy, and (4) safety and security. In each area, we pose key questions in the context of associ- ated impacts, which could result from the “CAVs on the Rise” scenario. Economics and Workforce In this scenario, many new business models related to transport services will be developing, especially in the private sector. How will the increased demand for pri- vate CAV services and related businesses (e.g., main- tenance, traffic management) affect different types of professions in the transport sector? Personalized requests and increasing individualization will open new chances for start-ups and smaller busi- nesses, but at the same time, few globally acting compa- nies will dominate the market. How will this affect the chances and barriers for workers? Which social groups might benefit and which could be disadvantaged (e.g., competences/education levels or socio-demographic characteristics)? What would be the consequences of a highly com- petitive transport market for the definition of work (e.g., hourly employee versus salaried employee, quality of work, work-life-balance)? Could this lead to a growing amount of uncertain jobs? Who might be affected? Equity How would such a transportation system impact socio- economic disparity? Is there a risk of increasing social gaps as individuals with financial resources or in power- ful positions receive a higher priority on their trips, while others may have to accept longer commutes due to lim- ited resources and longer travel times? How might this impact health due to reduced physical exercise and social contacts due to long commutes? Would competing services facilitate accessibility for all groups of people, including previous non-drivers like older adults, children, or people with disabilities? Or could access to transport become more restricted with no more alternatives being available, as private operators might discriminate or ban specific groups from using their services? With CAVs being the dominant transportation mode, could this result in discrimination of other road users? For instance, could pedestrians and cyclists be restricted to limited areas to avoid “disturbances” with connected fleets)? Data Access and Privacy Social groups with limited financial backgrounds or long travel distances may be forced to give up privacy and sell their data (e.g., when and where they travel, with whom, reveal their activities during travel) to be able to get around. This could also impact young people who may not take this seriously. What could be the consequences of such a development? With customers becoming more and more transpar- ent, service providers can use this information to either restrict or ban certain groups (discriminatory practices) or manipulate their behavior by putting them on specific routes or exposing them to specific information. What could be the effects? People who are especially concerned about their pri- vacy will find it challenging to protect their data and might greatly reduce their transport needs as a result. How will the users’ trust in how the different service providers are handling their data impact their freedom

7 2 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S of choice, especially in areas where there are just one or two providers? Safety and Security Will personal safety become a matter of financial resources, when CAVs get to “decide” who to sacrifice in the case of an unavoidable accident (e.g., higher worth buys you safety/security)? How can the personal percep- tion of safety and security affect the behavior of spe- cific groups? For example, women might avoid booking cheaper shared trips at night time (causing limited access), or active mobility could perceived be as much less safe, resulting in longer-term health issues. Some groups are more reliant on transportation access than others, (e.g., commuters or emergencies). Which groups are especially at risk, if the connected transport system fails due to malfunction or a cyber- security breech? Scenario 2: CAVs Tamed by Policy and People Due to international commitments to dramatically improve climate conditions and reduce the consumption of land and resources, governments have implemented effective measures on transportation to significantly decrease traffic volumes (e.g., road pricing). This devel- opment is supported by considerable changes in the production sector (e.g., emerging technologies like auto- mation and 3D-printing accelerate local productions and teleworking becomes common). Further, a para- digm shift occurs towards more sustainable, local life- styles. This results in a landscape of smaller communities where people spend most of their time, but they are highly connected to other parts of the country through telecommunication. If possible, people follow their daily routines in the close vicinity, working in local community centers offer- ing teleworking spaces and covering the majority of the distances by walking or using shared bikes or micro- vehicles (e.g., hoverboards and electro scooters). Smaller CAVs are only rarely used for passenger transport (e.g., people with disabilities or to reach larger hubs for mass transportation) and are strongly regulated as the use is only granted to specific groups, but are common for transporting goods, as people prefer to get their pur- chases delivered than carry them. For longer distances, automated public mass transport is used, but people usually avoid having to travel longer distances on a regu- lar basis. Only a small elite employ private AVs, mainly to display their status (as these are very expensive) and distinguish themselves from other citizens. We also explore possible travel behavior impacts across four dimensions in this second scenario: “CAVs Tamed by Policy and People”: (1) economics and the workforce, (2) equity, (3) data access and privacy, and (4) safety and security. In each area, we pose key ques- tions related to associated impacts, which could result from the second scenario. Economics and Workforce The automobile sector has shifted its focus from vehicles for private transport to community vehicles for special services and smaller vehicles for the transport of goods, many of them are produced in the region. In parallel, the construction of walking and cycling infrastructure and a variety of walking aids (e.g., small transport robots) create new businesses. How would such developments affect professions in the transportation sector? With jobs becoming available in the vicinity due to teleworking, specific groups (e.g., people taking care of family members, disabled people) have improved job access. At the same time, the local community can only offer a limited amount of jobs and professions, which may force people to move to find a job. Which jobs and competences may benefit more or less in such a scenario? Teleworking at local community centers and at home enables people to have more flexible working times, but it also can interfere with their private life, affecting their work-life-balance and productivity. What could be the consequences for work, families, and communities? Equity Reduced travel distances also imply that access to sev- eral facilities (work, education, leisure) cannot be com- pensated by “virtual trips,” which are not accessible in the community. Some groups may therefore be forced to travel. Which groups will most likely be affected by this? Is there a risk for certain groups that are disadvantaged? Some groups are more privileged than others in this scenario (e.g., by regulations or private resources). Could this contribute to increased social disparities in local communities, leaving individuals not eligible for public support and without the financial resources behind? Experience shows that frequently encountering unfa- miliar people and groups increases tolerance. Is there a risk that fairly closed communities may lead to increas- ing tensions between different local communities, result- ing in psychologically perceived in- and out-groups, even causing instability in a region or nation? Data Access and Privacy With daily activities being concentrated within a small spatial range, the boundaries between different aspects

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 7 3 of life (e.g., work and private life, family, and commu- nity) become blurred. As people also put a lot of trust in communities, the handling of sensitive personal or professional data may become negligent. What are the risks and which groups may be more prone to experi- ence disadvantages? Strong local communities can increase the sense of security, but they are also known for a high level of social control. As people know each other, each movement and delivery can be registered by people in the surroundings. How can close communities impact the freedom of indi- viduals to get around and keep their privacy? For efficiency reasons, regional companies strive to combine the production of goods and delivery of ser- vices in the region. These businesses gain access to a very rich and comprehensive database about their local customers. What could be positive or negative conse- quences for consumers? Safety and Security As CAVs have not succeeded in being widely accepted and local transport (walking and cycling supported by partially automated aids) has strongly increased, the positive effect of CAVs on road safety is not achieved. What are the consequences to no improvement in road safety or more but less severe accidents? Longer distances are much less traveled; hence, the capacity of transporting large amounts of people or goods is considerably reduced. What could be the impacts in the case of a natural disaster or terrorist attack when fast evacuation or quick supply of goods is vital? explOratOry tOpic 4 cOnnected and autOmated vehicle impactS Of StakehOlder reSpOnSeS Satu Innamaa and Matthew W. Daus Introduction This briefing paper focuses on the interaction between various regulatory approaches to CAVs and various socioeconomic impacts and issues due to different stake- holder actions. The briefing identifies: (1) stakeholders in the CAV implementation, technology and policy framework evolution process; (2) CAV governance scenarios across a spectrum of potential market forces versus government regulatory controls, and various outcomes of CAV implementation under such scenarios; and (3) key socioeconomic impacts related to the cross- cutting themes: (1) economics and welfare, (2) social equity, (3) data access and privacy, and (4) safety and security. Stakeholders The introduction of CAVs will influence many differ- ent types of stakeholder groups, and stakeholders may undoubtedly conversely initiate, react and/or mold policy and CAV frameworks—on both the public side (as constituents and interest groups) as well as on the private side (as consumers of private company services and products). Stakeholders can be grouped into catego- ries that include: (1) public and quasi-public entities; (2) users or impacted people or entities; (3) automakers; (4) private mobility companies; and (5) technology com- panies. In terms of governance and the scenarios identi- fied below and in this briefing paper, the influence of the public versus the private sector will be highlighted as the two major categories. Other subcategories that will be identified include organizations that represent groups of stakeholder interest groups (for example, technology industry think tanks, and non-profits that represent or further the interests of groups of consumers or issues). Such groups, to name a few, could include: representation of privacy protections or transport modes; private trade organizations representing industries or owners; con- stituent interest groups representing causes that promote sustainability or disability rights; and even organizations that represent numerous local or regional governments. The identification of as many stakeholders as possible is critical, as outreach to these groups to ascertain the socioeconomic impacts of CAVs will help not only to identify the research needs, but also to conduct further research by working with such groups to collect and ana- lyze data from as many diverse perspectives as possible. In terms of public and quasi-public stakeholders, groups would include, of course, government agencies, as well as quasi-public entities, like economic development, tourism bureaus and airports. Users or impacted people or entities would include not just passengers of CAVs, but also those who would interact with CAVs, like bicyclists, pedestri- ans, drivers of conventional cars and interest groups that represent their viewpoints. In terms of automakers, the nuances and subdivisions of this vast industry will be fur- ther identified as there are competing stakeholders not just in terms of manufacturers, but component parts, systems, dealers and advertisers, making up the vehicle industry ecosystem. Private mobility companies affected by CAVs include competing modes and sub-modes, such as transportation network companies, taxicabs, bike- share and microtransit, as well as future entities and joint ventures that are or may be developing (i.e., subdivisions or wholly owned spin-offs from tech companies or auto- makers that will be developing multimodal mobility ser- vices to supplement and/or promote the introduction and use of CAVs). Finally, technology companies, which are evolving in terms of their function and reach, include not just internet providers, smartphone application and big

7 4 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S data software or platform related service providers, but also telematics or safety equipment industry entities, or niche industry providers (e.g., taximeter manufacturers). Scenarios Two scenarios can assist in identifying key research needs with respect to stakeholder impacts on CAVs in the future. These scenarios are (1) “CAVs on the Rise”; and (2) “CAVs Tamed by Policy.” The figure below posi- tions these scenarios in opposite quadrants based upon two axes: government regulation high/low versus market forces strong/weak. In Scenario 1, the locus of control is on the private sector and, in Scenario 2, on the public sector. Both scenarios include assumptions on high pen- etration of CAVs in mixed traffic, and a high level of automation (SAE 4–5). For the identification of future research needs, the aspects of these two extremes should be analyzed. Figure 3 describes the two scenarios and addresses potential impacts, challenges and concerns relating to four cross-cutting themes: (1) economics and welfare, (2) social equity, (3) data access and privacy, and (4) safety and security. Scenario 1: CAVs on the Rise In Scenario 1, regulators choose to offer minimal inter- vention for the provision of mobility services. This approach may result in a wider array of innovative ser- vices being developed that may not necessarily serve the best interests of all segments of society, and could present more safety risks. The basis of the transport system relies on private competing mobility services and privately owned AVs; hence, the public transport service offering would be reduced. Different mobility services include carsharing and ridesourcing/transportation network companies. The eco systems behind these services also include meta- operators for mobility service concepts, payments, etc. Traveler services include infotainment, in-car-work enabling services (e.g., texting, email and teleconfer- ence) and parking services for private AVs. Economics and Welfare In Scenario 1, there could be high job elimination or re-classification for many for-hire drivers and other Market Forces Strong Market Forces Weak Government Regulation Low Government Regulation High Scenario 2: “CAVs Tamed by Policy” • Regulators: Prescriptive over the introduction of CAV based shared mobility services • Transport system: Public transport, incl. demand-responsive public transportation and mass-transit utilizing CAVs; Public/private collaboration; Subsidized by the public sector; Travel chains well functioning and intermodal, including walking and biking; Private AVs expensive and quite heavily taxed • Services: Multimodal mobility services (MaaS / MOD smartphone apps & platforms); First-mile, last-mile services; Intelligent journey planners; Infotainment Scenario 1: “CAVs on the Rise” • Regulators: Minimal intervention • Transport system: Private competing mobility services: market operated fleets for carsharing and ridesourcing/TNCs with service level-based pricing, Door-to-door services, public transport service offering reduced • Services: Carsharing, Ridesourcing/ TNCs; Meta-operators for mobility service concepts, payments, etc.; Infotainment; In-car-work enabling services; Automated parking FIGURE 3 Impacts of automated and connected vehicles on stakeholders.

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 7 5 mobility operators (buses, etc.). In addition, there could be changes in the employment responsibilities and work descriptions for the on-the-job drivers, as driving is replaced by other activities while the car is driving to its destination. If there is an increase in unemployment, it affects transport demand indirectly because reduced income causes loss of ability to buy/pay for transport. Is employment support needed for the most affected labor groups due to job loss from vehicle automation? How can we effectively support affected job classes to adapt such employment opportunities to the advent of automation? The impacts on the make-up of the workforce needed could directly affect the income of the transport service providers. In addition to the savings related to driver costs, by increases in vehicle utilization of CAV-based services, operating costs may be reduced and profitability may therefore increase. However, it is unclear whether passengers will benefit from reduced fares as a result of these decreased costs. In Scenario 1, innovative mobility businesses emerge in a variety of different service models and providers in the private sector. Ecosystems emerge to facilitate end- user services. Consequently, there will be an increase in workforce demand for service provision and service con- cept development, fleet management, etc. It is likely that there will be a large number of small service operators in addition to large, even global operators. In this scenario, there is a possibility for additional income for citizens via the offering of private vehicles for carsharing or ride- sourcing services. How can we support innovations (e.g., policy, training) that create jobs that benefit from CAV introduction? In Scenario 1, outside-city-center malls are popular for commercial services. There are also cost-efficient delivery services to support shopping in them. Can all citizens access these services? Social Equity In Scenario 1, private mobility services may not be afford- able for all, or may be subsidized in the short term and then could become unaffordable for many. The use of ser- vices will include also indirect costs (e.g., it may include a requirement for access to smartphones or similar plat- forms) and the use of mobile payment services. This sce- nario may cause social inequities. There may also be a variety of enhanced service-levels resulting from higher prices, which could further increase inequities. There is a risk for the selection of customers if the service is banned for some. The challenge is how we can guarantee afford- able mobility for all citizens (all income classes and user groups). What role will or should private equity subsidies play in both the short-term and long-term provisions of services? As public services have largely been replaced by com- peting private businesses in this scenario, transport options for people with special needs (e.g., wheelchair users) have become quite expensive, as the vehicles require specific equipment and the target group is comparatively small. This development might be compensated by, for example, advocacy groups collecting donations to support people in need, or insurance companies offering premium reduc- tions or other incentives to facilitate such services. This approach would require a paradigm shift as such services are now viewed as involving increased risks and premi- ums. What could be the potential solutions provided, who would offer these solutions, and what would the conse- quences be for different social and other groups? There is a risk of regional inequity in service offerings if market forces focus the service provisions only on areas where the profit margin is greater and there are no ser- vices for areas with low demand where service provision would be non-profitable. There is also a risk for unbal- anced routing via different neighborhoods in navigation if routing scenarios can be influenced by payments to the navigation service provider. On the other hand, in Sce- nario 1 with a larger number of smaller vehicles and more flexible routes, there may also be more opportunities for network level optimization of traffic than in Scenario 2, with more fixed routes. How can we manage the equi- table provision of mobility across regions? Other key questions include: What are the objectives for different private stakeholders in the mobility sector? Are there conflicting interests? What are the key areas we need to understand in CAV shared mobility services to further benefit citizens and businesses? How can we balance the benefits of the private sector and of cities in the regulatory environment in which they are deployed? Data Access and Privacy In Scenario 1, the private service providers collect big mobility data from their customers, and may engage in consumer profiling. There is little anonymity/privacy for travelers as they are identified when utilizing such services, and their mobility habits may be monitored. Mobility service providers may have additional business interests and revenue streams that result by vending the data or information retrieved from mobility data. On the other hand, these datasets are unlikely to be openly avail- able anytime soon absent intervention. There is a high demand for cybersecurity services. Some of the questions asked will include whether private service providers col- laborate to develop industry standards or protocols (best practices) for the ownership, licensure and dissemination/ use of data? Will the tug-of-war over data ownership, use

7 6 S O C I O E C O N O M I C I M P A C T S O F A U T O M A T E D A N D C O N N E C T E D V E H I C L E S and privacy be resolved by private industry litigation or public and private contractual agreements that evolve without any government regulatory interference or inter- vention? It is also quite possible that changing business relationships (mergers/acquisitions), joint ventures and mutual investment partnerships between technology/ software companies, auto manufacturers and mobility providers could shape the data/privacy framework inde- pendent of government involvement. Safety and Security In Scenario 1, the users of CAVs may be better protected than other road users in mixed traffic. The crash risk per distance traveled is decreased, but increase in traffic volumes and mileage may increase exposure to crashes, reducing the overall safety impact. Driverless ridesourcing may compromise the security of the travelers or at least their feeling of safety. Market forces, in terms of private concerns for liability exposure and safety risk, could dic- tate a manufacturing paradigm or scheduling for the mix of CAV fleets. To this end, will private companies decide that it is in their best interests economically—as well as for the safety and security of passengers, pedestrians and other motorists—to have all vehicles on the road be of the same SAE level (e.g., no mix of lower and higher automa- tion levels)? In Scenario 1, the economic interests of the private sector may lean towards what some view as more safety risks by CAV entry into urban markets, as opposed to less profitable implementation in areas where there are reduced safety risks—such as closed communities (e.g., amusement parks, retirement communities, college cam- puses). Is there a real possibility of the development of safety and cybersecurity standards by the private sector, internationally or domestically in the United States or the European Union? Scenario 2: CAVs Tamed by Policy In Scenario 2, regulators are more prescriptive than in Scenario 1 over the introduction of CAV shared mobility services, perhaps resulting in constraints on innovation in transport provision, but with a more defined vision over how such services should be used to improve safety and mobility for city residents and businesses. The basis of many transport systems rely on pub- lic transport, which includes demand-responsive public transportation and mass-transit utilizing CAVs. Travel chains are well functioning and may be intermodal, including walking and biking. There is public/private collaboration, and transport is subsidized by the public sector to ensure a minimum level of mobility services to all people. Private AVs are expensive and quite heav- ily taxed. Traveler services include multi modal mobil- ity services—Mobility-as-a-Service (MaaS) or Mobility on Demand (MOD) smartphone apps and platforms— first-mile, last-mile services; intelligent journey planners; and infotainment. Economics and Welfare In Scenario 2, it is likely that there are slower and more deliberate job protections (than in Scenario 1) due to labor organization input and legislation to protect work- ers, or that unionization is allowed and workers’ rights are enhanced. Still, there will be changes in workforce needs. Which work positions will be reclassified or pre- served, and are substantially different skills needed for those reclassified positions (e.g., elder care or paratransit specialists to assist in vehicle ingress/egress, school bus matrons, or customer service assistants to guard against passenger misconduct)? The workforce impacts will affect the income of the transport service providers. In addition to the savings related to driver costs, by increasing vehicle utilization of CAV-based services, operating costs can reduce and profitability may therefore increase. However, the changes in the transport system have also indirect impacts on the public bodies to provide subsidized transport services as fewer private vehicles lead to lower tax income, which has an effect on how to fund subsidies. There is an increased workforce demand for operat- ing and developing demand-based and other CAV-based public transport services. A workforce is also needed for development of different easy-to-access and easy-to-use information services related to them. In addition to public transport services, there may be public–private collabora- tions in provision of the mobility services. In Scenario 2, there may be less work in the private mobility service sec- tor, than in Scenario 1, and also fewer but larger service operators. In Scenario 2, transport hubs serve as important ser- vice hubs and are also used as commercial market places. Social Equity In Scenario 2, basic transport services are easier to guar- antee for all regions and social classes (low-cost mobility options with mass-transit), including rural areas, than in Scenario 1. These mobility services require the use of smartphone applications and smart cards, which may be challenging for older adults. The basic service fee for public transport is the same for all. However, there may be additional mobility options, such as “first-mile, last-mile” services, with additional fees that may cause social inequity. In this scenario, there is a large differ- ence in mobility for the affluent (private AVs) and the rest of society (public transport based mobility services). What role will or should public subsidies play in both the short-term and long-term provision of services?

A P P E N D I X B : B R I E F I N G P A P E R S O N E X P L O R A T O R Y T O P I C S 7 7 In Scenario 2, with public sector transport operators, it may be easier to achieve equity among neighborhoods in terms of traffic management, as political decisions can be better made by analyzing routings of traffic flows, than in Scenario 1. There may also be better compliance with network level optimization of road network use. A distribution of tariffs may allow low-income neigh- borhoods that are poorly served by public transporta- tion to be served by new innovative CAV-based mobility services at lower costs, whereas areas with good quality public transportation links and infrastructure that sup- port active travel, may have higher tariffs on new mobil- ity services to limit their use. In this scenario, there may be subsidized delivery services to support healthy living (e.g., food) for all. One of the key questions is how policies and regula- tion can support introduction of private CAV services and solutions that supplement the public transport and support the mobility strategies of municipalities (e.g., use of active modes of transport). How can CAV-based shared mobility services be supported by regulation and planning? In addition, what role should the gov- ernment play in the formation of laws and regulations that impact or distribute economic costs or obligations on industries, modes and/or sub-modes of transport ser- vices to provide wheelchair-accessible service to people with disabilities—in the private CAV transport market? For example, much debate has occurred in the United States following the passage and implementation of the Americans with Disabilities Act, and the U.S. Depart- ment of Transportation’s implementing “equivalent ser- vice” standards. These laws and regulations exempted taxicabs from providing wheelchair-accessible vehicle (WAV) services; however, local efforts to use taxicabs and for-hire vehicles to provide WAV services, when not mandated by federal law, are impeded by the “equiva- lent service” standards, which treat less-than-perfect service as discrimination. Such exacting or demanding service standards may not be fully achievable without significant subsidies to offset WAV ramp retrofitting and increased insurance costs. In the European Union, regulation imposes the rights of passengers, specifically mentioning equal rights also for those with a disability or reduced mobility. Data Access and Privacy In Scenario 2, public authorities are the owners or licensed users of mobility data. Anonymity/privacy of travelers can remain for some public transport services, but identification of the user is required for some services. There is more potential for the provision of open data with respect to mobility than in Scenario 1. In Scenario 2, there is a similarly high demand for cyber security services as in Scenario 1. In Scenario 2, one might question what the legal data paradigm would be in light of freedom of information laws, “right to be forgotten” laws (e.g., the United Kingdom), and the impact on public-private MaaS/MOD partnerships. The role that legislation will or could play to resolve and manage competing interests and concerns about data ownership and use could have a stabilizing effect on private business ground rules and expectations. However, one could argue to the contrary that government regulation of data could undermine the economic underpinnings or motivations for companies to enter and develop innovative products and services, if the right to access private data or ownership is under- mined or minimized. Safety and Security In Scenario 2, the crash risk per distance traveled may be reduced for CAVs and lower traffic volumes in terms of the number of vehicles possibly decreasing exposure to crashes (i.e., a decrease in relative crash risk) for all road users (including vulnerable road users). Driverless pub- lic transport in Scenario 2 may compromise the secu- rity of the traveler or at least their perception or feeling of safety. In this scenario, government regulators could control or dictate the CAV mix on the streets, and may determine whether the safety and security of passengers, pedestrians and other motorists will be enhanced by having all vehicles on the road be of the same SAE level (e.g., no mix of lower and higher automation levels), or by restricting the use of various SAE levels to certain high, low or mixed density areas. Can or should the gov- ernment control CAV rollout as a function of land use or urban planning, with safety as a priority? If so, what are the risks or safety concerns associated with implementa- tion of CAVs in urban, versus suburban, rural and closed or quasi-closed communities (e.g., amusement parks, retirement communities, college campuses)? Finally, is there an ideal governance structure for the development of uniform laws, rules and regulations addressing safety and cybersecurity standards by the public sector, inter- nationally or domestically in the United States or the European Union? referenceS Rohr, Charlene, Lissa Ecola, Johanna Zmud, Fay Dunkerely, James Black, and Eleanor Baker (2016). Travel in Britain in 2035: Future Scenarios and Their Implications for Tech- nology. RR-1377. Rand: Cambridge, UK. Stocker, Adam and Susan Shaheen (2018). Shared Automated Vehicle (SAV) Pilots and Automated Vehicle Policy in the U.S.: Current and Future Developments, Road Vehicle Automation 5. Springer: Cham, Switzerland. U.S. Department of Transportation (2016). Federal Automated Vehicles Policy, https://www.transportation.gov/sites/ dot.gov/files/docs/AV%20policy%20guidance%20PDF.pdf (accessed 08 November 2016).