Robotaxis: Uber’s Autopilot Anxiety

Recently, there have been a few notable announcements in mobility and ride-sharing. 

Firstly, Waymo has obtained permission to broaden its paid robotaxi offerings to include freeways and select highways in the San Francisco Bay Area. Until now, Waymo has only conducted tests of its autonomous robotaxis on San Francisco’s freeways with a safety driver present.

This development has been in the works for some time, and it’s becoming increasingly evident that Waymo is progressing toward a scalable self-driving solution. 

Despite Waymo’s announcement, we’re still far from fully autonomous cars becoming mainstream. Interestingly, Tesla is also considering launching their own robotaxis, starting in China. 

A company that should be particularly concerned by this is Uber. Uber, who shut down its autonomous car unit a few years ago, decided, around the same time, to partner with firms working on autonomous mobility solutions.

In a bold move, Uber’s CEO noted that Tesla will allow them to have access to customers for its robotaxis. 

Bold… since if there’s one thing we know about Elon Musk, it’s that he doesn’t easily admit to depending on others. 

The central question of this discussion is how the world will look when self-driving cars become widespread. What types of ownership structures will emerge, and who will benefit the most—customers, platforms, or vehicle owners? 

Who will Benefit From RoboTaxis?

Many articles on this topic assume a model in which ride-hailing platforms own the assets or partner with the OEMs. However, since Uber currently doesn’t own any of the cars used by its drivers, it’s unclear what the main ownership structure will be in the future. 

The paper “Ride-Hailing Platforms: Competition and Autonomous Vehicles” by Auyon Siddiq and Terry Taylor, investigates the impact of autonomous vehicles (AVs) on the profitability and competitive dynamics of ride-hailing platforms like Uber and Lyft. The authors explore whether AVs are beneficial or detrimental to these platforms, to human drivers, and to society as a whole.

The central question is: How do AVs’ accessibility, and ownership structure influence the profitability of ride-hailing platforms, the welfare of drivers and passengers, and the social welfare overall?

To analyze the decision-making processes of ride-hailing platforms regarding pricing, wages, and fleet size, the authors use a game-theoretic model (i.e., this isn’t an empirical study, but one that tries to anticipate the type of behavior we should expect). 

The ownership structures that are considered are the following:

1. Platform-owned AVs: Platforms own and manage their AV fleets. This can happen either through full ownership or a partnership. For example, according to a joint announcement, Uber is re-entering the robotaxi market through a decade-long, multi-market agreement with Motional —a collaboration between Hyundai and Aptiv— to introduce autonomous vehicles across its ride-hailing and delivery services.

2. Individually owned AVs: The AVs are owned by private individuals or third parties but are operated by the platforms. Imagine you own a car, but while you’re at work for example, and you’re not using it, you allow Uber to operate it. Elon Musk has mentioned in the past that he envisions Tesla vehicle owners using a smartphone app to activate a feature that allows their cars to enter commercial service and pick up passengers on the company’s network. 

The trade-offs in the model include:

Cost of AVs (high vs. low): This determines how platforms strategize around AV deployment.

Price and Wage Competition: This has to do with how platforms adjust prices for passengers and wages for drivers to maximize their profits while managing the shift to AVs.

AV Ownership Structure: This affects how costs are distributed, and impacts competition dynamics between platforms.

The following graphic summarizes the paper’s main results: 

Parsing these results, the main findings are:

Platform Profits: When AVs are owned by the platform and their costs are high, this can lead to reduced profits. When AVs are individually owned, low costs lead to reduced profits due to intensified price competition.

Human Drivers Welfare: This factor generally decreases with access to AVs due to job displacement of drivers unless AV costs are low and the benefits for consumers outweigh drivers’ losses.

Social Welfare: Access to AVs can increase or decrease social welfare depending on ownership structure and AV costs. Individually owned AVs tend to increase social welfare across various cost scenarios.

But you may be wondering about the implications.

At this point, the cost associated with AVs is high, so the implications for companies like Uber and Tesla can be quite significant:

For Uber (as a ride-hailing platform):

1. Increased Cost Pressure: The high cost of AVs could strain Uber’s financial resources, especially if they choose to own and operate their own AV fleet. Managing these costs while maintaining competitive pricing for customers could be challenging.

2. Strategic Pricing and Wage Decisions: The model suggests that Uber’s profitability could be negatively impacted if the firm can’t effectively manage the balance between lowering consumer prices and reducing driver wages. This balance becomes even more critical as Uber moves toward a mixed fleet of human-driven and autonomous vehicles.

3. Competitive Dynamics: If Uber’s competitors also gain access to AV technology (which is a real possibility), the intensified competition could further pressure prices and wages. This scenario could lead to a “race to the bottom” in pricing, potentially eroding profit margins. 

For Tesla (as an AV manufacturer):

1. Market Opportunity: As a manufacturer, Tesla stands to benefit from selling AV technology to ride-hailing platforms. Apart from selling cars, Tesla’s strategy also includes the possibility of licensing software or offering fleet service directly to companies like Uber, or selling directly to individuals who make them available on ride-hailing platforms for some extra income.

2. Technology Leverage: Tesla’s quick progress in AV technology can become a significant competitive advantage. By providing platforms like Uber with AV technology, Tesla can secure a separate, steady revenue stream. 

3. Partnership Potential and Risks: Engaging in partnerships with ride-hailing services could expose Tesla to the volatile dynamics of the ride-hailing market. However, these partnerships could also provide valuable data and real-world testing environments for Tesla’s AV technology, speeding up innovation and improvements.

In essence, the transition to AVs in the ride-hailing industry represents both an opportunity and a challenge, which will require strategic foresight and careful management of financial, technological, and regulatory risks. 

What will the Transition Look Like?

If the analysis above sounds a bit futuristic, given how good (or bad) AVs are, you’re not wrong. 

So considering that we’ll still have human-driven and self-driving cars, the question is, “How will the world transition?”

The paper “Getting Out of Your Own Way: Introducing Autonomous Vehicles on a Ride-Hailing Platform” by Francisco Castro and Andrew Frazelle explores the strategic challenges and implications for a ride-hailing platform when integrating autonomous vehicles (AVs) alongside human drivers. It delves into the dynamics of AV fleet management, human driver recruitment, and the competitive interplay between human-driven vehicles and AVs, especially in a transitioning phase where both coexist.

The central questions addressed in the paper are:

1. How should a ride-hailing platform set wages for human drivers and manage its AV fleet size to optimize operations while recruiting enough human drivers to maintain service reliability?

2. What is the impact of human drivers’ strategic decisions on joining the platform when they anticipate a future increase in the platform’s AV fleet?

The authors develop a game-theoretic model, similar to the previous paper we discussed, where the ride-hailing platform operates a fleet of AVs and recruits human drivers who make strategic decisions about joining based on their expected earnings. 

One of the most interesting dynamics the paper demonstrates is the “race to the top.”

The study illustrates how a ride-hailing platform might unexpectedly harm its own financial performance when it chooses to increase its AV fleet after observing the levels of human driver participation. 

Why?

The authors explain that when anticipating that the platform will add more AVs, human drivers expect reduced earnings. Consequently, the platform is compelled to offer higher wages to attract the desired number of drivers. This necessity pushes the platform to a “race to the top,” where it must continue to raise wages whilst expanding its AV fleet. 

This cycle not only diminishes the platform’s profitability by limiting its ability to attract a substantial number of human drivers, but also increases the costs associated with recruiting drivers. The findings emphasize the critical role of the platform’s credibility. 

Success in avoiding this “race to the top” depends on the platform’s ability to assure human drivers that it won’t expand its AV fleet on impulse, as failure to provide such assurance could lead to significant profit losses.

The implication is that the platform (e.g., Uber) must manage the integration of AVs strategically to avoid wage spirals and uncontrolled fleet expansion. In other words, maintaining a balance between human drivers and AVs is crucial to ensure profitability. 

Bottom Line

As we edge closer to a future dominated by autonomous vehicles, the landscape of urban mobility and ride-sharing is set to undergo significant transformations. The implications of this shift extend beyond the technical feats of engineering; they encompass a profound reimagining of how services are delivered and who will reap the economic benefits. With companies like Waymo leading the charge with regulatory approvals for their robotaxis, and Tesla potentially joining the fray, we are glimpsing the contours of a new competitive arena.

As we delve into these questions, stakeholders must engage in thoughtful dialogue and strategic planning. The transition to autonomous ride-sharing is not merely a technological upgrade, but a pivotal moment that could redefine the rules of engagement across multiple sectors. Whether we are heading toward a future of shared prosperity or increased centralization of wealth and control remains a critical question of our times. 

Combining the findings from the two research papers, it’s not easy to see how specifically Uber can ever be widely profitable. I know this is a long-term vision, so looking at the transition period, and examining the “race to the top” in wages and “race to the bottom” in pricing, may give us a hint regarding the firm’s long-term success.