As ride-hailing firms drive into the future, who is being left behind?

This article is brought to you thanks to the collaboration of The European Sting with the World Economic Forum. Author: Joanna Moody, Ph.D. Research Program Manager, Mobility Systems Center, MIT Energy Initiative, Massachusetts Institute of Technology & Scott Middleton, Planning and Research Analyst, Massachusetts Institute of Technology & Amitai Bin-Nun, Director of the Autonomous Vehicle Initiative, Securing America’s Future Energy (SAFE) & Maya Ben Dror, Lead, Autonomous and Urban Mobility, Centre for the Fourth Industrial Revolution of the World Economic Forum

---

The on-demand mobility services offered by providers like Uber and Lyft have experienced significant growth since their introduction in the US in 2009. While the rapid adoption of these mobility service providers (MSPs) and the diversification of services they offer have expanded people’s mobility options, the benefits of using these services have accrued to some more than others.

As MSP usage grows, it becomes increasingly important to identify and address inequities in on-demand car services. Visionaries look to a future of pooled, autonomous mobility services as a way to expand the benefits of MSPs to more people. While both pooling and automation do hold promise as gateways towards an inclusive MSP model, they bring challenges of their own which will require deliberate attention if we are to unlock their possible future benefits. Research in the US has increased our understanding of the distributed impacts of MSP services, and has illuminated the need for proactive and collective governance to ensure inclusivity of MSPs – both now and in the future.

The benefits of MSP services accrue almost exclusively to those who use them. MSPs have expanded mobility options for carless households and communities with limited or poor quality taxi and transit services. However, usage is higheramong MSP users in urban rather than suburban areas.

It’s also worth asking who does not have access to MSPs. Even with fares cheaper than traditional taxis, affordability remains a barrier to low-income individuals. And despite programs such as concierge services, the app-based platform is an additional barrier for the unbanked and digitally impoverished.

While outperforming taxis and paratransit services on some accessibility metrics, evidence shows that MSPs fail to provide an equivalent service for people with disabilities, particularly those who need access to wheelchair accessible vehicles.

Finally, concerns regarding safety and personal security may discourage women from using these services.

There is also evidence that, like the traditional taxi industry, MSPs provide a platform for discrimination. This can manifest in several ways:

· Drivers towards riders: electing not to drive in or provide coverage for certain types (minority or low-income) of neighbourhoods; declining to accept reservations from certain types of passengers or canceling pickups once the passenger’s identity is revealed; driving routes that increase costs and/or travel time based on race, gender, or perceived socioeconomic status. Multiple studies have found differences in wait times and cancellations between African American and Caucasian riders.

· Riders towards drivers: lower tips or ratings for drivers based on race, gender, or perceived socioeconomic status, which could impact earnings or influence termination practices; or

· Both ways: leaving low ratings for passengers or drivers based on race, gender, or perceived socioeconomic status.

While the benefits of MSPs accrue narrowly, negative externalities are felt more broadly. This is especially true for congestion on urban streets. Fortunately, pooling MSP rides has the potential to reduce traffic congestion by combining multiple trips into a single vehicle, improving efficiency, reducing traffic and expanding mobility to more people. From the user perspective, sharing could mean cheaper fares, since the cost of the trip is distributed over multiple riders. However, this comes at the ‘cost’ of a more circuitous and uncertain route and the possibility of sharing a ride with a stranger. Different users vary in how they value this tradeoff.

Although cost and time may be primary factors, there is also growing concern that some users may avoid pooled services because they prefer not to share space with strangers, with the potential for discriminatory outcomes. Building on the studies of rider-driver discrimination discussed above, other recent studies have investigated rider-rider discriminatory outcomes in pooled rides. One studyfinds that riders are less likely to pool in racially or ethnically diverse neighbourhoods. Other research finds that some MSP users harbour discriminatory attitudes towards fellow passengers of different social class and race, that fear of negative social interactions may reduce users’ willingness to request pooled rides, and that these riders prefer to have early information about potential fellow passengers. Another recent paper found that discriminatory attitudes are correlated with a lower level of satisfaction with pooling, a lower proportion of MSP rides that are pooled, and a lower willingness to consider pooling in the future. So rider-to-rider discriminatory attitudes may discourage sustained and frequent use of pooling among MSP users – a potentially problematic outcome for an increasingly shared (or pooled) future.

Autonomous (or driverless) vehicles, once technologically mature and broadly deployed, may accelerate a shift to MSP services. However, automation by itself may not solve issues of equity or discrimination. While the elimination of a driver removes one potential source of bias, there is potential for bias in AI algorithms, and the absence of a driver may change the social dynamics between passengers in the same vehicle, exacerbating existing rider-rider discrimination.

In addition, there is the matter of cost. Widespread adoption of autonomous vehicles and its ability to reach low-income households will likely only occur if consumers find them cheaper and/or more convenient than the alternative of privately owned vehicles. By removing the driver (and possibly shifting to a more efficient remote driver mode), autonomous MSP services could pass on savings in operational costs to the consumer. While some models that have assumed complementary innovation would allow pooled autonomous vehicles to reach a price point below personal ownership, others have argued that autonomous MSP services will remain more expensive than personal car ownership in the US unless automation is accompanied by increased utilization, significant pooling, or changes to other aspects of the taxi business model.

While the advent of MSP services has increased access and offered benefits relative to the status quo, evidence is mounting that their benefits are not afforded equally to all individuals. Unequal access resulting from self-selection of users, barriers to access for certain groups, and discrimination should be acknowledged, studied carefully and addressed with informed and proactive policy.

Sharing these services is seen as a key way to improve their efficiency, but it comes with its own challenges: primarily travel time detours and sharing trips with strangers. The potential deployment of driverless vehicles in the future could heighten the importance of the MSP model. By eliminating the driver and potentially reducing costs, automation could alleviate some equity concerns. However, it’s clear that sharing and automation (either independently or together) will not alone fix the many issues surrounding inclusivity of MSP use. Collective governance from both the public and private sectors is needed to address existing issues of inclusivity and to shape how MSPs will be used in the future. And these solutions will need to be sensitive to local context and incumbent mobility systems in order to address the remaining challenges in the mobility service provider model.

Contributing authors:

– Adam Cohen – Researcher, Institute of Transportation Studies at the University of California, Berkeley

– Susan Shaheen – Adjunct Professor of Transportation Engineering at the University of California, Berkeley