Source: Medium
Didi, Lyft, Ola, Uber and other transport network companies (TNC) are transforming people’s daily lives — and in the process capturing the ire of incumbent taxi companies, the enthusiasm of investors, and the attention of regulators.
Recent investments signal these companies’ evolution from daring upstarts to a mainstay of transportation infrastructure. An investment in Lyft by General Motors in 2016 valued the former at $5.5 billion, while only a year later another investment made by the private equity firm KKR implied a valuation of $7.5 billion. A stake in Uber taken by one of Saudi Arabia’s investment vehicles reinforced a $62.5 billion valuation for the largest global ridesharing platform, and many other similar mobility companies have arrived across the globe.
The rapid growth of mobility services comes with a substantial impact. A recent study by number of well-respected economists estimated the social welfare generated by UberX alone in 2015 to be around $6.8 billion. A separate report by the Ruderman Family Foundation has drawn attention to the huge potential benefits of so-called “autonomous vehicles” (driverless cars) to disabled communities — benefits that possibly could be magnified when integrated in a ridesharing fleet.
It is at this point quite evident that these technologies, and the firms driving them, are prompting a major upheaval in the menu of mobility options and the relative price and service differentials among them. What is less certain, however, is what this upheaval means for the energy and carbon intensity of the transport sector. Indeed, this uncertainty is all the more pronounced as TNCs move to integrate vehicles with varying degrees of autonomous driving capabilities into their operations.
But what are the climate implications of such integration? Will Uber and other TNCs cut transport emissions, or lead to a disastrous increase? What are the conditions that will determine these outcomes?
What has thus far been lacking in public discussion is a cogent, comprehensive framework for the many variables that shape the ultimate climate impact of shared mobility services, and a nuanced understanding of which variables could be reasonably shaped by regulators, and which are largely outside of the control of governments. This paper aims to discuss these variables in a robust manner, and outline the importance of understanding them in developing climate and transportation policies.
It is imperative that policymakers who are responsible for increasing the energy and carbon efficiency of the transport sector fully consider different future scenarios for shared mobility services, but to do so, several factors pertaining to mobility services’ emissions impact must first be investigated with more vigor.
Why Mobility Services Matter for the Climate
The case for climate change as a cross-cutting imperative for government policy seemingly grows with every passing month; 2016 registered as the hottest year since instrumental records began in 1880 and the third such consecutive record-breaking year. Notably, the transport sector is arguably the greatest decarbonization challenge of the twenty-first century: oil-reliant transport recently surpassed the power sector as the single largest source of carbon emissions in the United States. Passenger vehicles account for around two-thirds of the sector’s emissions. Any move to shrink this footprint involves many more actors and uncertainties than such efforts in the power sector, considering the 260 million passenger vehicles in the United States today (versus 7,658 power plants) and the highly transformative technologies, such as vehicle automation or TNCs, poised to change the way they are utilized.
Ideally, regulators will see their role as not only to mitigate the most undesirable impacts of TNCs, but also to leverage the significant opportunities afforded by these new technologies and business models.
The U.S. Environmental Protection Agency (EPA) recently underscored the rationale for such an approach, noting that in order for the United States to meet its Paris climate targets, the existing policy framework of vehicle fuel efficiency standards will be insufficient. With respect to TNCs and autonomous vehicles, one official emphasized that “There must be a way for public policy to be as innovative as some of these things we’re seeing...I don’t know what that looks like yet, but I know for sure we need to think differently post-2025.”
Before a specific set of policies can be envisioned and implemented, there is a need to first understand the various dynamics that may lead shared mobility services to increase or reduce emissions, and to have a practical framework for thinking through the ways that policy might bend these dynamics in a more desirable direction.
What Levers do Policymakers Have?
Policies can reduce transportation emissions by addressing one of the following: increasing the fuel efficiency of vehicles, reducing the carbon intensity of fuel, or reducing vehicle miles traveled (VMT). Each of these approaches, termed “Technology, Fuels, and Activity” by the U.S. EPA, have played a role in transportation regulation. Traditionally, policies such as corporate fuel economy standards have tweaked “technology,” Zero Emission Vehicle (ZEV) mandates and emission intensity standards have pushed markets toward cleaner fuels (“fuels”), and policies such as HOV lanes and fuel taxes have sought to reduce emissions through changing travel patterns and choices (“activity”).
Emerging shared mobility services will ideally, though not necessarily, engender new policy options across all three approaches, especially those related to “activity.” From an asset-utilization perspective, automobile transportation appears to be begging for disruption: Census data shows that 86 percent of commuters go to work in an automobile, that 76 percent of them ride in their vehicle alone, and that those vehicles spend at least 93 percent of their existence parked, not being used at all. The impending shift toward flexible “mobility-on-demand” (MOD) service models will reshape our transportation to be much more efficient, and a corresponding evolution in transportation policies will also take hold.
This asset inefficiency has persisted due to individuals’ high implied demand for the “option value” afforded by private vehicles combined with insufficient alternatives, whether for lack of coverage (transit), high prices (taxis), or coordination frictions (traditional carpooling). New mobility services, however, are upending these dynamics by reducing the barriers to coordinating various vehicles with passengers. As service models change, regulatory models must adapt in kind.
In a world of increasingly shared vehicles, there may be new opportunities to regulate at the source. For example, previously impractical concepts like real-time congestion pricing could actually become a reality in the age of connectivity. Additionally, regulations operating at the vehicle level may be rendered ineffective if vehicle type loses importance relative to vehicle function. Shared-use vehicles, especially if they are to become automated, could generate significantly more mileage per vehicle than conventional automobiles, meaning that any fuel efficiency improvements in those shared vehicles is vastly more impactful from an energy and emissions standpoint than the car parked in a family’s garage.
Local and national governments have already turned to new mobility services as instruments for improved transportation service coverage and safety; some cities have begun subsidizing certain Lyft and Uber rides connecting to transit stops, and others have included private mobility services as part of “guaranteed ride home” programs. In addition to these uses, policymakers should also consider the potential environmental impacts, both positive and negative, that shared mobility services can unleash. Their growing importance has even prompted MOD services to be included in some regional long-range transportation plans.
Of course, a necessary prerequisite for effective policymaking around MOD travel is available data, which has been a central challenge for local and regional governments pondering how to engage with services like Uber and Lyft. While private mobility companies may be hesitant to share data, local authorities have many bargaining chips with which they could reach data sharing agreements, such as airport pickup lane access, streamlined payment with public transit services, the inclusion of qualifying services into HOV or other specialized lanes, or even subsidizing certain trips.
Utilizing shared mobility services for climate goals would not require data on specific trips or users, but only on measures such as aggregate ride distance, vehicle type, and empty miles between rides, which could put less of a strain on user privacy (a top concern for mobility providers like Uber). In a positive turn of events, Uber has launched a new Uber Movement website, which allows tracking of travel data in several cities – indicating a strong departure from its prior, controversial reluctance to relinquish data to city officials.
Emerging public-private partnership models may be able to accelerate this trend further. For example, the United Kingdom (UK) government is backing a consortium that includes Oxbotica, an Oxford University machine learning spin-out, and XL Catlin, an insurer, with £8.6m in funding for a project to run an ongoing convoy of six autonomous vehicles between London and Oxford with the goal of gathering data with diverse applications - from cybersecurity to insurance to road safety - that may ultimately expedite the move toward a future of autonomous, shared mobility services dominating the commute between the two cities. Governments funding such trials would likely have more direct access to and control over resulting data than in other arrangements, and with the added ability to encourage certain key actors to be included or excluded from a given consortium based on policymakers’ objectives.
Factors Driving Emissions Impacts
In order to accurately quantify the emissions impacts of emerging MOD services, policymakers and research professionals should make the best effort to consider each of the following factors. In some cases, quantitative data on these factors is currently proprietary or simply unattainable, but each is needed to fully investigate whether Uber, Car2Go, Via, or any other emerging service is a bane or boon to their local and global environments. As governments begin to consider these services as tools for environmental gain, further efforts to quantify each of these are necessary.
1) The grams of CO2e emitted per passenger mile of the service
Perhaps the greatest, or at least most salient, benefit offered by shared vehicle services is the ability to move multiple individuals around in a more efficient manner. It may be directionally accurate to say that as more passengers are integrated into a single trip, a greater volume of passenger miles will be distributed over the same (or marginally more, as in the case of a “pool” trip that requires circuitous routing to accommodate all passengers) fuel use and CO2e emissions. But even this depends upon the circuitousness of a given shared trip, as well as how many rider-less miles (deadhead miles) are incurred in the process of driving to, and waiting for, riders.
While many TNC advocates believe TNCs represent a significant net efficiency improvement, a serious approach to determining the marginal emissions cost or benefit of shared vehicle services will require more information to be collected and shared.
For services that run homogenous fleets of identical vehicles, such as Bridj, calculating this metric is relatively simple, and can be determined with data on the total operational miles of travel (including miles between and before passenger pickup), the type of vehicle utilized, the fuel powering that vehicle, and the total passenger miles traveled.
However, if a fleet contains a variety of different vehicles, as is true in business models such as Uber that harness a heterogeneous mix of private cars, then additional detail is needed to determine the average emissions per mile traveled. In such cases, accurately determining the grams per passenger mile of the entire service requires pairing which vehicles drove which miles. For example, “full time” Uber or Lyft drivers may make up a minority of their respective fleets, but may contribute to the majority of miles driven. These drivers may choose more fuel efficient cars than passive drivers who base their vehicle purchases primarily on other activities. In such a case, calculating the grams per passenger mile of a typical Uber ride based on the average vehicle emission rate would overestimate the emissions from Uber travel. Many services may already collect the vehicle-level data necessary to calculate this metric, but if not, this issue could also be circumvented by tracking fuel usage by all service vehicles and applying the appropriate emissions factors.
One interesting implication of having TNCs collecting fuel consumption information across their fleet, either voluntarily or via mandate, is that it would make it easier to facilitate downstream (at the pump) discretion between different fuels based on their total lifecycle emissions, from initial extraction of the feedstock to the final product. A large entity such as Uber or Lyft would be more likely to have the resources and motivation necessary to make such decisions, and thus there would be greater momentum behind a system of labeling or other such transparency measures that would allow biofuels and electricity to compete against different gasoline or diesel derived from various different crude-to-product pathways. As TNCs become responsible for an increasingly larger share of total transport fuel purchases, innovative policies such as emissions intensity standards may also proliferate and become more effective.
2) The degree of substitutability between a given service and other transportation modes.
A heated debate continues about whether new mobility providers are substituting or complementing public transportation. Surveys have shown that mobility services are used in place of a variety of other options, including taxis, trains, walking, and personal driving, and that usage patterns differ regionally and across various types of private services. Some research has suggested that TNCs were complements with public transit rather than competitors in that they helped facilitate a carless lifestyle, and that individuals foregoing car ownership or selling their car may also start taking public transit and walking much more often.
Many cities have even sought to harness TNCs as a means to increase transit ridership. For example, a Florida transit agency is experimenting with suspending two regular bus lines and using the funds to instead subsidize select Uber and taxi rides, while Lyft has worked out a similar arrangement with a Colorado town. Most recently, a California Bay Area transit agency is experimenting with subsidizing both Lyft and Uber rides—up to an aggregate total of $200,000—that take place between two test areas. In each case, public authorities are attempting to expand total usership of public services by expanding the range of available transportation modes.
If the user base of mobility services expands and if prices fall, they may also start heavily competing with public transit and having adverse impacts on transit ridership. In a growing number of cases, city officials are tackling this issue head-on by providing incentives for TNCs to support public transit systems. Ridesharing platforms have expanded their New York City ridership base at an exponential pace, and increased from five million passengers in 2015 to around 16 million passengers in October 2016 alone. However, 2016 was also the first year since 2009 that saw a decline in New York City’s subway ridership, driven primarily by a 3 percent decrease in weekend trips. A new analysis of New York City travel data corroborated the suspicion that TNCs were behind the transit ridership decline and indicated that TNCs have induced a net gain in vehicle travel in the city.
Meanwhile, in an example of cross-subsidization cutting in the opposite direction, Massachusetts is preparing to assess a 20 cent per trip fee on all shared mobility services platforms, with ten cents being allocated to municipal budgets, five cents allocated to a state transportation fund, and five cents allocated to the traditional taxi industry.
3) The emissions per passenger mile of other transportation modes
As a jumble of different services continue to enter the transportation space, the task of equally weighing them against one another becomes increasingly difficult, but important nonetheless. For example, the emissions impact of trips being drawn from taxis to Uber is as much dependent on the cleanliness of existing taxi agencies as on the disruptor. Since emissions intensities of taxis and public transit systems differ across cities, this is an important consideration on a local level. Carbon intensity of transit services also vary regionally, with some metro rail and bus systems producing more than five-fold the emissions per passenger mile of other similar systems.
4) The number of induced trips
Early research suggests that these effects are modest; few trips done by ridesourcing services such as Uber and Lyft would have been avoided, rather than substituted, if the services were unavailable. Given the nascence of the ride-sourcing industry, it is unsurprising that few folks have built their lives around the existence of these services. However, as new mobility services become more established, entire communities will build their lifestyles around expanded travel possibilities, and overall trips may increase more substantially.
If vehicles were to become automated and built specifically for ultra-high utilization, it has been postulated that the costs to provide on-demand passenger transportation could sink as low as $0.18 per mile. Yet others believe that the introduction of advertising and other revenue streams into automated TNCs could even result in a zero- or even negative-cost per mile, under certain circumstances. For example, one could imagine riders with certain behavioral patterns being offered free rides at certain times in exchange for captive exposure to advertising.
As many companies (e.g. Uber, Lyft, Via, Chariot) experiment with ride-matching services that group riders together going in similar directions—also known as pooling, ride-splitting or concurrent sharing—prices are falling further. These pooled rides could have a favorable emissions impact by virtue of consolidating trips, though that could be overshadowed by induced automobile travel. Via, a ride-sourcing company specializing in ride-splitting services, offers a flat rate of as low as $3 per ride—making it nearly competitive with local transit services. While low transportation costs and increased mobility are laudable aspirations, they could—if managed poorly—induce more mileage, thus clogging streets, polluting the air, and causing traffic incidents.
With new mobility technologies and services still in their incipient phase, it is also important to recognize that pricing models are still very much in flux, and in many cases prices do not reflect cost. Uber, for example, reportedly lost more than $2 billion in heavily subsidized rides in China while battling Didi for market share before capitulating in mid-2016.
5) Impact on car ownership
Services such as Uber, Car2Go, Capital Bikeshare, and Bridj all work together to make a carless, or fewer car, lifestyle more appealing by providing a variety of on-demand services, which can in turn spur more use of transit, walking, and biking. In this way, a given mobility service may cause substitution between two other modes of transportation. Past research of individual carsharing services has indicated that those services reduced car ownership and emissions among users, but groups of services could be even more effective at reducing car ownership rates. Indeed, other surveys challenge the premise that shared mobility services will quickly lead to avoided car ownership, pointing out that household car ownership has remained relatively constant even in metropolitan areas with the greatest “density and intensity of carsharing fleets.” Since personal car ownership’s appeal is largely rooted in its versatility, the advent of any singular niche service, however convenient it may be, is unlikely to challenge private vehicle ownership. However, a suite of services capable of collectively tackling mobility needs will be much more competitive, especially if it can do so at a lower cost.
The shift from personal car ownership to mobility as a service (MaaS) has been a tantalizing prospect for transportation planners for years, but the recent advent of communications technology and innovative service models has bolstered the feasibility of such a concept. MaaS systems work by allowing users to subscribe to a transportation package provider for a flat monthly fee, and then use their credits for a variety of services available to them. The first MaaS provider, Whim, is currently operating in Finland now, and individual transportation providers like Uber and Lyft are moving in a similar direction by offering flat rate monthly packages. Even traditional automotive manufacturing companies are following suit, with many having announced plans for launching their own mobility services.
Long Term Implications and Other Considerations
Some impacts of MOD services may be immediately realized, such as variance in usage of local transit services (which could be either positive or negative), but others may be more structural in nature and could take years, even decades, to fully develop. With continued growth, mobility services may soon start to shape where people choose to live, what kinds of vehicles are produced, and how cities are planned.
1. Parking and Urban Form
Efforts to model ubiquitous shared vehicle fleets have shown vast benefits, such as each shared vehicle replacing around ten private vehicles. Since shared (and potentially autonomous) vehicles could be perpetually moving, parking requirements could be slashed. Recent research by the International Transportation Forum finds that, with full market penetration, parking requirements could be cut by 95 percent. In the long term, shared fleets could allow for more densely packed and efficient cities without restricting mobility. This is insufficient to understand the full emissions implications of such a change, however. More will need to be done with respect to characterizing the relationship between city density and emissions intensity, or more specifically, understanding the emissions penalty imposed by parking lots.
2. Vehicle Build and Electrification
Vehicles for perpetual passenger transportation will be built vastly differently than those sitting in your garage. Rather than being carefully calculated generalists, they will be dedicated specialists for specific trip types, likely with a focus on dense seating, higher fuel efficiency due to increased operational expenditure, and an emphasis on passenger comfort.
As many have speculated, these design changes could have major impacts on vehicle emissions, especially if low-cost electricity spurs service providers to utilize electric vehicles. With high utilization rates, the comparatively larger up-front costs of electric vehicles are rendered less important, as these costs can be quickly amortized over many vehicle miles traveled, shortening pay-back periods. According to figures from Lex AutoLease, a UK fleet leasing adviser, the monthly savings on a fully electric hatchback (the Nissan LEAF Acenta) versus a similar standard hatchback (the Ford Focus 1.5 EcoBoost Zetec S) equate to an average of £77 per month, or £3,700 over a four-year operating cycle.
A recent study by MIT researchers using detailed GPS data has shown that the energy requirements of 87 percent of vehicle-days (an awkward but useful measurement for vehicle use) can be met by an affordable, extant electric vehicle. It is the remaining 13 percent of mobility-related energy demands that cannot be met by current electric vehicle options that remains a barrier, both in physical reality and even more so in inflated psychological terms, to electric vehicle adoption. By deploying electric vehicles for trip-by-trip services, it circumvents the issue of electric vehicles not being able to serve the full suite of buyer needs. This also points to a potential role for coordinated government policies to support a complementary portfolio of lower-carbon mobility (shared mobility services enabling electric vehicle deployment) through subsidies and other mechanisms that address market externalities or irrational behavior.
Current market actions support the idea that high-utilization, shared vehicles are likely to be electric, and that theory does indeed mirror reality. Pilot projects for full autonomous shared vehicles are in various stages of development by companies such as Google, Uber, nuTonomy, GM+Lyft, Faraday Future, and Tesla. In each case, the deployed vehicles are fully electric vehicles or plug-in hybrids. To date, we are not aware of a single shared, autonomous vehicle pilot project that has stuck entirely to traditional petroleum-fueled vehicles (either plug-in hybrids or full-electric vehicles have been used in all cases).
Uber, specifically, has recently introduced programs in first the UK (London), and now the United States (Portland), that sees it managing a fleet of ridesharing EVs and a dedicated network of charging stations. BMW’s ReachNow service, more analogous to ZipCar than a true TNC, has also deployed 30 i3 electric vehicles into its Portland fleet. Perhaps most notably, is GM and Lyft’s stated intention to bring forth a fleet of all-electric Chevy Bolts into the Lyft fleet.
TNCs, in this sense, may represent an important set of balance sheets that can be brought to bear in financing and funding charging stations across key urban areas, particularly when a critical mass of charging stations can be reached that help improve the efficacy of subsidies, standards, or other policy mechanisms. Daimler’s Car2Go program had previously introduced electric vehicles into its Portland fleet, for example, only to remove them shortly thereafter due to lack of charging infrastructure. While these charging networks may begin their existence as proprietary, one could envision policy mechanisms that would create incentives for TNCs to open their charging networks up (such as the opportunity to generate credits in emissions intensity standards systems) where such actions would help advance climate and/or environmental goals.
Geospatial Variance
Evidence from the early years of MOD services indicates that its usage varies across cities, and even within them, and these variances could be critical factors in determining whether new services are a bane or boon to the climate. Multiple components of local geography—population density, transit availability, grouping of business and residential areas, weather, etc.—can play a role in determining the usage patterns of mobility services. This variability is apparent in the range of approaches undertaken by policymakers toward regulating and/or partnering with mobility providers. Government can engage these services in a multitude of ways, such as taxing them, restricting usage in areas such as airports or, alternatively, subsidizing certain services in order to achieve specific regional goals.
As an example of variability, TNCs may be having a less ecofriendly impact in regions with harsh winters or scorching summers because they may be more likely to replace walking to and from metro stops, or metro (transit) trips altogether. Conversely, as evidence suggests, they may be primarily replacing taxi trips in cities with high density and available rail systems, and because TNCs can offer pooled rides and presumably a lower share of deadhead miles, this could yield environmental benefits. In a city such as New York, which is subject to both example conditions, these effects and many others need to be weighed against one another to reveal the overall environmental impact.
The energy source that fuels a trip—whether by taxi or ridesharing service or otherwise—also matters greatly. Just as the emissions of petroleum-fueled vehicles depend significantly on the particular lifecycle emissions of the fuel’s production pathway, if services begin adopting higher percentages of electric vehicles, then local grid emissions factors will also be important. Early steps are being made in London, where Uber deployed 50 electric vehicles to its fleet in 2016, with hundreds more to come throughout 2017, and many other services have stated goals of utilizing electric vehicles.
Private services such as TNCs can also make small tweaks to their algorithms that maximize certain variables, which could be done at the local level. For example, services providing pooled rides can navigate between the extremes of providing quick pick up times for riders, versus streamlined routing; by relaxing the requirements for one vehicle to be able to pick up riders, pickup times will drop, but riders may be taken further off course to pick up additional passengers. The desired calibrations here may also vary temporally and regionally as, for example, one could envision shorter wait times and longer drive times being appropriate late at night when an individual is deemed safer inside a vehicle than wandering the street.
From an environmental standpoint, services that optimize the efficiency of pairing drivers with passengers (presumably at the expense of quick wait times) will, all else being equal, produce the least emissions from fuel use. This will be especially important as TNCs shift from being used for occasional social trips to being a consistent source of commuting, such as is the case with Via. These services, which include Uberpool and Lyft Line, have begun to be included in employee benefits in several cities including New York City, Chicago, and Boston. And one company in Nashville is offering employees fully paid ridesharing transportation to work via Uber or Lyft.
Why Innovation Needs Governance
As mobility services have proliferated rapidly, information on their local and global impacts has lagged far behind, much to the dismay of governments and citizens alike. Since the impacts of mobility services vary locally, most policy decisions are best left to city and regional governments, but national governments also have a role in streamlining and harnessing this transition.
Specifically, federal governments should consider directly regulating mobility services as a means of achieving climate goals. There is a balance to be struck between allowing sub-national entities, such as states, to be laboratories of innovation without allowing a burdensome patchwork of incongruent regulations to proliferate. The move toward new mobility holds great promise for the environment and for other policy objectives, but also the potential for material risks and disruptions in anything from labor markets to safety concerns to traffic flows.
In such a dynamic, the best response is arguably not ad-hoc interventions and blanket restrictions on the part of local and regional governments that feel caught off-guard, but instead balancing the goal of unified standards with the reality that these services have different impacts depending on location. Ideally, governments will work toward promulgation of a thoughtful, stable, and balanced national policy framework in which the interrelated mobility trends of electrification, automation, and service-ification (such as via TNCs) can evolve. A fully formed framework might not be realistic at the outset, of course, but could come into place gradually out of an initial series of incentive mechanisms and government guideposts.
For example, EPA and/or the U.S. Department of Transportation could enact measures to incentivize TNCs to efficiently pool rides or to pair with conventional transit. Federal governments could also provide incentives for cities to enact forward-thinking policies, such as designating HOV lanes that include certain qualifying services. As shared rides consume more of total VMT, such options may be logical complements to regulating on a vehicle-by-vehicle basis.
At a more abstract level, though no less important, is the fundamental importance of governments understanding the opportunities and risks associated with new technologies as they enter their early stages of development, rather than reacting to these upheavals ex post—often with ineffective or draconian measures. A prime example of an institution designed to accomplish just this is the now defunct Office of Technology Assessment (OTA), which existed as an official component of the U.S. Congress from 1972 until its defunding in 1995 amid the implementation of the Republican majority’s “Contract with America” program. The OTA could, and should, be revitalized with a modernized design and focus that reflects the major technological trends, including those in the mobility sector, facing society today.
International bodies also have an important role to play. With automated driving and electrification rushing to further accelerate this transition, the global community may benefit from a coordinated effort to share data around key test cities that lead the curve. Much like DOT’s Smart City Challenge and Mobility Sandbox programs have spurred innovation within the United States, an international fund for testing shared, electric, and autonomous vehicles could benefit cities that are interested in the potential of transformative ideas, but lack the funding or confidence to invest without a precedent. All cities that are currently puzzling over if and how to embrace these technologies would benefit from gleaning insight from test cities.
The Way Forward
The new mobility revolution stands at a unique intersection in time, particularly when set against the backdrop of a new administration in the United States that comes to power unencumbered by a long history of commitments, ideologies, or experiences in the space. TNCs, particularly when combined with autonomy and electrification, provide an avenue for emissions reduction while also achieving a litany of other policy goals, and also enjoy the benefit of being amazingly—if fleetingly—apolitical in the current moment.
Secretary Chao of the U.S. Department of Transportation recently announced that she is in the process of reviewing guidance that was issued last year by the Obama administration, via the National Highway Traffic Safety Administration (NHTSA), that had promoted flexibility on the part of the Federal government so as to not discourage innovation in the still-inchoate autonomous vehicle industry. This has the potential to be a positive step forward, in particular if it takes a broad view of the spectrum of risks and opportunities posed by autonomous, shared mobility services, rather than focusing narrowly on just a single technology (e.g. automation) or a single policy equity (e.g. safety).
Specifically, it would be encouraging to see other agencies, including the EPA and U.S. Department of Energy, involved in the DOT review so that multiple policy equities, including climate policy, are represented. These early cross-agency coordination efforts could in turn lay the groundwork for more structured research institutions and testing initiatives to quickly answer some key questions. The UK has pursued an approach akin to this, and has quickly advanced to a leadership position in the global supply chain for new mobility services.
If the Trump administration indeed plans to remain in the Paris Climate Agreement while simultaneously revising much of the policy architecture put in place by the Obama administration to meet the U.S. commitment therein, then a vacuum will be created for new tools and policies.
The Trump administration has sent early signals that in the power sector, this may take the shape of additional support for carbon capture and storage in lieu of the Clean Power Plan. As the Trump administration similarly steps back from more aggressive vehicle fuel efficiency targets put in place by the Obama administration, it should consider the role of new technologies and service models—such as those offered by TNCs—to help fill some, if perhaps not all, of the resulting void. One would hope that there is room for pragmatism in this space.
Innovation, whether in physical technologies, institutional arrangements, or business models, has long been a driving force for economic progress. Regardless of what priorities the current U.S. administration identifies with regard to climate and environmental policies, it would be costly to fail to pursue the full promise of innovations such as shared mobility services, in particular as these services move toward employing autonomous, electric vehicles into their fleets.
Innovations throughout the emerging “new mobility” sector, from ridesharing services to mapping companies and beyond, have thus far had a strong American provenance. If the United States chooses to arrest that momentum, or even neglects to capitalize upon it fully, the fruits of innovation will not disappear; they will simply migrate elsewhere.
As Raj Rajkumar, a professor of electrical and computer engineering and co-director of the General Motors-Carnegie Mellon Vehicular Information Technology Collaborative Research Lab, so aptly puts it, “Whether we do this or not in the U.S., the technology will continue to be developed in Europe, tested [and] deployed in Singapore, Qatar, etc. Since this technology can revolutionize transportation, it is imperative that the U.S. maintain its edge.”
The economic, safety, and technological edge arguments are those most often marshalled when futurists or policy planners paint a picture of a shared, autonomous, electric mobility future. Undoubtedly, directing policy resources to this opportunity is of paramount importance for countries seeking to find new productivity gains and drivers of growth. However, there is also a need to recognize that the proliferation of mobility services and TNCs will have potentially profound consequences for the emissions of the transport sector. Harnessed correctly, they may help catalyze a virtuous interplay between the optimization, electrification, and decarbonization of transport. If left unattended, there is a risk that they instead only serve to further ingrain a high-carbon, oil-dependent, over-congested transport system.
Those who purport to know exactly where the mobility transformation will lead—to utopia or dystopia—are shirking intellectual honesty. It is too early to tell, and too few policymakers are even beginning to grapple with the most important questions and trade-offs.
A useful starting point, however, is to take stock of the factors and uncertainties laid out in this paper and consider how best they can be addressed given the opportunities and challenges of unique contexts, from California to Kyoto to Kigali. No one policy approach will likely be correct in all circumstances, but any policy approach should consider the emissions determinants, and the levers to shape them, that we have laid out here. There has never been a better moment for policy to climb into the driver’s seat.
