As has been the case historically, we should expect that emerging transportation technology will transform cities.
Autonomous Vehicles
Autonomous vehicles, or self-driving cars or driverless cars, are able to drive from one point to another with little or no direct input from any passengers. As of 2025, limiting ridesharing services without human drivers are available in select cities throughout the world, and so the impacts are largely hypothetical.
In an editorial, Romem (2023) suggests four impacts that autonomous vehicles will have on city shape.
- They will lower travel costs per kilometre.
- Will Expand AV availability to improve worker well-being, travel lengths, and city size.
- Land rents will rise in the centre while falling in the periphery.
- Potential AV daytime parking locations will be investigated.
Travel Costs
Romem’s (2023) first point is very plausible. Litman (2019) finds that about 10% of the private cost of car ownership and operation is for insurance, and that could be greatly reduced with improved safety. Zhang et al. (2022) conduct field experiments comparing fuel consumption of manual and autonomous driving. Theory suggests that, by maintaining a more steady speed and reducing acceleration and deceleration, autonomous driving should reduce fuel consumption relative to manual driving. The field experiments confirm this, with manual driving requiring 5.6% more fuel per kilometer at an average speed of 20 kilometers per hour, and 14.7% more fuel at an average speed of 40 kilometers per hour. Modeling by Gueriau and Dusparic (2020) finds that high penetration rates of autonomous vehicles should reduce collisions by 50-80%.
Travel time may be the biggest cost reduction. The Oregon Department of Transportation’s (2025) latest estimate is of a value of travel time of $24.79 per hour per person for commuters in passenger vehicles, with national estimates ranging from $10 to $100 per person per hour. This means that a measure that saves one person one hour of driving time, all else being equal, should be valued at $24.79. The American Automobile Association (2023) reports that the average American spends 60.7 minutes each day driving 29.1 miles, for an average speed of about 29 miles per hour. In other words, each mile of driving has a time cost of 86¢, comparable to the direct monetary costs of Litman (2019).
There are two mechanisms by which autonomous vehicles might reduce travel time costs. The first is by freeing drivers from the necessity of manual driving and allowing the time to be used for more productive or enjoyable activities. Second, autonomous vehicles may reduce travel time, as modeling by Zhang and Gao (2020) incidates.
While the precise effect may be impossible to quantify while autonomous vehicles are still uncommon, it is reasonable to expect that their proliferation will lower travel costs on a per-kilometer basis, measured in direct monetary terms, time, and safety. This will likely have a significant impact on the shape of cities.
Rebound Effect
Since the spread of autonomous vehicles should reduce per-kilometer travel costs, as measured in direct monetary terms, time, and safety, it is reasonable to expect that they will increase the total number of kilometers driven by means of the rebound effect, or induced demand.
Medina-Tapia and Robusté (2019) model the impact of autonomous vehicles on a hypothetical circular city. First, they consider direct effects that should be expected, assuming no change in travel patterns. Less road space should be required, as autonomous vehicles can more efficiently avoid congestion and follow at shorter distances. The total cost of driving should be reduced by a third, including a 20% reduction in travel time. However, with decreased cost, the authors posit a rebound effect, or induced demand, of as much as 50%. Some of the increase is the result of automobility being made available to some groups, such as the young, the elderly, and the disabled, who are not served by conventioned driving. The authors expect an increase in the total amount of driving and city size and amiguous impacts on overall travel time.
Building a model based on survey results, Batur et al. (2015) identify several possible rebound mechanisms: an increase in the number of trips, mode shifts, increased trip length, and residential relocation. They find that rebounds are likely to be greater with privately owned automated vehicles rather than with a mobility-on-demand service. They conclude that policies to curb private automated vehicle ownership will limit induced demand.
Property Values
Under the (monocentric) standard urban model, the sum of property values and transportation costs to the center of a city should be constant. Cities thus pose a tradeoff between housing and transportation costs: areas near the center tend to be more expensive but with lower transportation costs, while areas near the fringe tend to be less expensive but with higher transportation costs. By reducing the per-kilometer travel costs, it is reasonable to expect that autonomous vehicles will flatten the housing price gradient; in other words, property values in the periphery of a city should rise relative to those in the center.
This is what Huang, Li, and Ross (2018) find in another context related to the cost of driving. In Singapore, a Certificate of Entitlement is required to legally drive a car, and the Singapore government restricts the availability of Certificates of Entitlement so as to manage congestion. The authors founds that as the cost of a Certificate of Entitlement rises, the housing price gradient steepens; in other words, property near the city center increases in value relative to property at the periphery. Note that this is the opposite of what Romem (2023) suggests.
Hiramatsu (2022) conduct a simulation of the effect of the adoption of autonomous vehicles and find,
The results show that increased prevalence of AVs steers people toward suburbs with poor public transportation. Thus, high-income workers react more to technological progress, while low-income workers react more to lowered ownership costs.
Modeling in the 20 largest metropolitan areas in the United States with the exception of New York City, Wang (2023) also finds that autonomous vehicles should decrease commuting cost and time, and thus it should flatten the price gradient from the city center. Additionally, she finds that most cities, such as those in the San Francisco Bay Area, will have an overabundance of parking, which can be converted to housing and relieve the supply shortage.
City Size
Again under the standard urban model, the boundary of a city occurs where transportation costs are so high, and thus property values are so low, that the land is equally valuable as either built-up urban land or as farm land. Beyond this boundary, it is uneconomical to convert land into urban land, and this mechanism defines the size of the city. If the spread of autonomous vehicles lowers transportation costs and flattens the housing price gradient as one moves from the center of a city, then under the standard model, it should also result in geographically larger cities. Romem (2023) asserts that this will happen.
This is the conclusion of Zacharenko (2016). He builds a simple model in which economic activity is centered around a port, a point through which all imports and exports must pass. The central business district is a small semicircle around the port, while residences comprise a larger semicircle. Zacharenko’s (2016) model also finds the emergence of a parking belt around the central business district, to which cars would drive and park while their owners were at work. Average commuting distances would increase. However, in contrast to the above results, Zacharenko (2016) finds that property values in the central business district would rise relative to those in the periphery. The reason for this is that the reduced parking needs in the central business district increase the value of land.
Economic Impact
Based on our analysis of agglomeration economies, it may be expected that policies that increase the size of cities as functional labor markets will create wealth through agglomeration economies. Autonomous vehicles may have the potential to do this by increasing the throughput of traffic on a road network.
Winston and Karpilow (2017) build a model of the economic impact of congestion and estimate the economic gain from 50% adoption of autonomous vehicles based on congestion relief. Their model exploits California’s self-help county tax, which the authors argue is an exogeneous to economic performance measures. They then extrapolate the California results to the rest of the United States. They find that 50% autonomous vehicle adoption could increase gross domestic product by $214 billion, or about 1%, create 2.4 million jobs, and generate $90 billion in income.
References
Romem, I. “How will Driverless Cars Affect Our Cities?”. Energy Theory. November 2023.
Litman, T. “Transportation Costs & Benefits”. Victoria Transport Policy Institute. Updated September 2019.
Zhang, L., Zhang, T., Peng, K., Zhao, X., Xu, Z. “Can Autonomous Vehicles Save Fuel? Findings from Field Experiments”. Journal of Advanced Transportation 2022(1): 2631692. July 2022.
Oregon Department of Transportation. “Value of Travel Time Estimates”. February 2025.
American Automobile Association. “American Driving Survey: 2023”.
Zhang, T., Gao, K. “Will Autonomous Vehicles Improve Traffic Efficiency and Safety in Urban Road Bottlenecks? The Penetration Rate Matters”. 2020 IEEE 5th International Conference on Intelligent Transportation Engineering (ICITE). September 2020.
Gueriau, M., Dusparic, I. “Quantifying the impact of connected and autonomous vehicles on traffic efficiency and safety in mixed traffic”. 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC). December 2020.
Hiramatsu, T. “Impact of autonomous vehicles on the choice of residential locality”. Transportation Planning and Technology 45(3), pp. 268-288. April 2022.
Wang, S. “The Impact of Autonomous Vehicles on the Real Estate Housing Market in the United States”. Thesis, Master of Science in Real Estate Development at the Massachusetts Institute of Technology. February 2023.
Zacharenko, R. “Self-driving cars will change cities”. Regional Science and Urban Economics 61, pp. 26-37. November 2016.
Medina-Tapia, M., Robusté, F. “Implementation of Connected and Autonomous Vehicles in Cities Could Have Neutral Effects on the Total Travel Time Costs: Modeling and Analysis for a Circular City”. Sustainability 11(2), 482. January 2019.
Batur, I., Mondal, A., Alhassan, V.O., Asmussen, K.E., Bhat, C.R., Pendyala, R.M. “The induced demand implications of alternative adoption modalities of automated vehicles”. Transport Policy 175: 103879. October 2025.
Winston, C., Karpilow, Q. “A New Route to Increasing Economic Growth: Reducing Highway Congestion with Autonomous Vehicles”. Mercatus Working Paper, available at SSRN. January 2017.