Are consumers willing to pay to let cars drive for them? Analyzing response to autonomous vehicles

JavaScript is disabled on your browser. Please enable JavaScript to use all the features on this page. This page uses JavaScript to progressively load the article content as a user scrolls. Click the View full text link to bypass dynamically loaded article content. View full text

Highlights • This paper focuses on early response to selfdriving vehicles ownership. • Microdata from a nation-wide online panel of 1260 individuals is used. • Semiparametric estimates of the willingness to pay for automation are derived. • The mean WTP for full automation is $4500 in a fixed parameter model. • Substantial preference heterogeneity is found in random parameter models.

Abstract Autonomous vehicles use sensing and communication technologies to navigate safely and efficiently with little or no input from the driver. These driverless technologies will create an unprecedented revolution in how people move, and policymakers will need appropriate tools to plan for and analyze the large impacts of novel navigation systems. In this paper we derive semiparametric estimates of the willingness to pay for automation. We use data from a nationwide online panel of 1260 individuals who answered a vehicle-purchase discrete choice experiment focused on energy efficiency and autonomous features. Several models were estimated with the choice microdata, including a conditional logit with deterministic consumer heterogeneity, a parametric random parameter logit, and a semiparametric random parameter logit. We draw three key results from our analysis. First, we find that the average household is willing to pay a significant amount for automation: about $3500 for partial automation and $4900 for full automation. Second, we estimate substantial heterogeneity in preferences for automation, where a significant share of the sample is willing to pay above $10,000 for full automation technology while many are not willing to pay any positive amount for the technology. Third, our semiparametric random parameter logit estimates suggest that the demand for automation is split approximately evenly between high, modest and no demand, highlighting the importance of modeling flexible preferences for emerging vehicle technology.

Personal mobility is about to experience an unprecedented revolution motivated by technological change in the automotive industry (National Highway Traffic Safety Administration, 2013 ; Fagnant and Kockelman, 2014). The introduction of automated vehicles –in which at least some (and potentially all) control functions occur without direct input from the driver– will completely change how people move. The adoption of automated navigation systems has the potential to dramatically reduce traffic congestion and accidents, while creating substantial improvements in the overall trip experience as well as providing enhanced accessibility opportunities to people with reduced mobility (Fagnant and Kockelman, 2015).

Automated vehicles use sensing and communication technologies to navigate safely and efficiently with little or no human input. Automated navigation technology comprises any combination of (1) self-driving navigation systems informed by on-board sensors (autonomous vehicles) vehicle-to-vehicle (V2V) and (2) vehicle-to-infrastructure (V2I) communication systems that inform navigation and collision avoidance applications (connected vehicles). The National Highway Traffic Safety Administration (NHTSA) has suggested five levels of automated navigation: level 0 (no automation), where the driver is in complete control of safety-critical functions; level 1 (function-specific automation), where the driver cedes limited control of certain functions to the vehicle especially in crash-imminent situations (adaptive cruise control, electronic stability control ESC, automatic braking); level 2 (combined-function automation), which enables hands-off-wheel and foot-off-pedal operations, but the driver is expected to be available at all times to resume control of the vehicle (adaptive cruise control and lane centering); level 3 (limited self-driving or conditional automation), where the vehicle potentially controls all safety functions under certain traffic and environmental conditions, but some conditions require transition to driver control; and level 4 (driverless or full self-driving automation), where the vehicle controls all safety functions and monitors conditions for the whole trip.1

Imminent commercialization of automated cars is best exemplified by the recent announcement (October 2016) that all new Tesla vehicles will have full self-driving hardware.2 Several semi-autonomous features are already available in the automotive market, mostly in the form of in-vehicle crash avoidance upgrades with preventive warnings or limited automated control of safety functions, such as braking when danger is detected. Self-parking assist systems are another example of a more advanced upgrade that is currently available in select makes and models. These entry-level automation packages are possible as a result of vehicles being equipped with radar, cameras, and other sensors. Even though technology is still evolving, full automation is possible with the current stage of development. The Google car and its more than 2 million miles of driverless driving is the most publicized effort.3

The literature on vehicle-to-vehicle, vehicle-to-infrastructure, and control systems for safe navigation is extensive. Regulation, insurance, and liability are other areas where there is strong debate. However, little attention has been devoted to the analysis of automated vehicles as marketable products. Consumer acceptance is critical to forecast adoption rates, especially if one considers that there may be strong barriers to entry (potential high costs, concerns that technology may fail).

Our work contributes to two strands of literature on the demand for new technology. The first area is the recent development in understanding the demand, penetration, and policy implications of autonomous vehicle technology. Several recent studies attempt to understand how consumer preferences for attributes such as safety, travel time, and performance shape the demand for driverless cars. Kyriakidis et al. (2015) conducted an international public opinion questionnaire of 5000 respondents from 109 countries. Responses were diverse: 22 percent of the respondents did not want to pay any additional price for a fully automated navigation system, whereas 5 percent indicated they would be willing to pay more than $30,000. Payre et al. (2014) conducted a similar survey of 421 French drivers with questions eliciting the acceptance of fully automated driving. Among those surveyed, 68.1 percent accepted fully automated driving unconditionally, with higher acceptance conditional on the type of driving, including usage of highway driving, in the presence of traffic congestion, and for automated parking. Similar results were obtained in a survey of Berkeley, California, residents conducted by Howard and Dai (2013). Individuals in this survey were most attracted to the potential safety, parking, and multi-tasking benefits. Schoettle and Sivak (2014) conducted a much larger and more internationally based survey of residents from China, India, Japan, the United States, the United Kingdom, and Australia. The authors found that respondents expressed high levels of concern about riding in self-driving vehicles, with the most pressing issues involving those related to equipment or system failure. While most expressed a desire to own an autonomous vehicle, many respondents stated that they were unwilling to pay extra for the technology.

A paper related to our own is that by Bansal et al. (2016), which estimates willingness to pay for different levels of automation. They find that for their sample of 347 residents of Austin, Texas, willingness to pay (WTP) for full automation is $7253, which is substantially higher than our own estimate. The authors also estimate WTP for partial automation of $3300, which is similar to our estimate.

Our demand estimates contribute to the assessment of the social costs and benefits of autonomous vehicles. Fagnant and Kockelman (2015) estimate the external net benefits from autonomous vehicle penetration. They find that the social net benefits including crash savings, travel time reduction from less congestion, fuel efficiency savings, and parking benefits total between $2000 and $4000 per vehicle. These estimates, however, greatly depend on how the presence of autonomous vehicles will impact both vehicle ownership and utilization. For example if autonomous vehicles make owning a vehicle more desirable, then the stock and use of vehicles may increase, reducing the external net benefits.

We designed a web-based survey with a discrete choice experiment to determine early-market empirical estimates of the structural parameters that characterize current preferences for autonomous and semi-autonomous electric vehicles. The discrete choice experiment contained as experimental attributes three levels of automation: no automation, some or partial automation (“automated crash avoidance”), and full automation (“Google car”). Automation was allowed for alternative powertrains (hybrid electric, plug-in hybrid and full battery electric). Based on the results from this experiment, we estimate WTP for automation. Our estimates of WTP for privately owned autonomous vehicles take a first step to understanding the demand for this technology, which is critical for understanding how aggregate demand for vehicles and vehicle miles traveled will respond to the technology over time.4

In addition to the discrete choice experiment of vehicle purchase, the survey also contained an experiment to elucidate the subjective discount rate of potential vehicle buyers. Expanding on the work of Newell and and Si̇i̇kamäki (2013), we used the individual-level experimental discount rate to determine the present value of fuel costs for each alternative.