Inherent in discussions about autonomous vehicles (AVs) are the costs associated with them. These costs are often presented in terms of cost of design/manufacture, purchase/lease, maintenance costs, insurance costs, and tax/licensing policy which promotes the development and deployment of AVs. The one aspect of government policy that has been overlooked in this discussion, so far, is the topic of the actual infrastructure associated with the operation of ALL vehicles, including those that are autonomous.
One of the costs associated with AVs is the anticipated growth in traffic to city streets that are often, in many urban areas, at capacity. In the ideal version of AVs people abandon car ownership in favor of using AVs. People of all stripes will be able to motor about in increasing numbers. This will increase traffic. The traditional answer to this is to build more lanes, even some dedicated to AVs. The challenge with this is the concept of induced demand. When a street is enlarged, congestion is eliminated temporarily. In time, the number of drivers increases to meet the increased capacity. Congestion returns. And, historically, we build more capacity to meet the demand and create more induced demand.
This building of more road capacity costs money. Where will this money come from? State gas taxes range from a high of $0.582/gal in Pennslyvania to $0.1225/gal in Alaska. The average is $0.3104/gal. The federal gas tax is $0.184/gal. These taxes along with tolls and motor vehicle license fees covered 41.4% of state and local road spending in 2013. The remaining funds come from general federal , state, and local revenues. Depending on where one lives the mix of user fees/taxes and general funds with more rural states tending to subsidize transportation spending from general revenues.
What does this mean for AVs and transportation policy? Pervasive AV use will create more traffic and congestion, even with expanding road systems. The footprint of 200 people in 177 cars is much larger than 200 people on three buses or on 1 light rail train. When the Brooklyn Bridge was converted from part rail to all cars in 1948 it went from carrying 400,000 people/day to carrying only 170,000 people/day. In congested places, AVs are not a solution to transit.
In addition there are other policy implications with AVs, particularly in the development of communities. AVs serve to heighten suburban-style development. Such development is a function of many factors, particularly highways. AVs will serve to use of cars and, serve to expand our cities horizontally. This creates the need for more infrastructure to build and maintain. The tax revenue from low-density places is not enough to repair/replace the roads and pipes once they fail. AVs serve to induce more suburban development much as car ownership does.
Another policy implication is associated with safety. By increasing support of AVs and cars in general, we implicitly decrease support from what makes our communities livable. More car lanes mean narrower sidewalks, thus decreasing use due to proximity to traffic and difficulty to navigate. Narrower sidewalks lead to fewer merchants along those streets as people will not be walking to them. Fewer walkers will serve to increase our nation’s obesity rate and lead people into more sedentary lifestyles.
With all this said, I do not want to write off AVs. The unfortunate thing is the focus is on how we can make cars safer and autonomous. In that way people with disabilities can get to where they want to go. I suggest we look at how we can provide cost-effective transportation options (including AVs—such as Ford’s AV bus initiative) to people so that we have strong communities for all.
Acknowledgment--Thanks to Jeff Speck for many of the ideas above