Salt Lake City just approved a $180 million highway widening project that won't be completed until 2029. Austin is building a new downtown parking garage with 1,200 spaces, set to open in 2028. Phoenix is rezoning entire neighborhoods to require more parking per square foot of retail space. All of this is happening while the transportation industry races toward a future where most of these investments will be obsolete before the concrete finishes curing.
American cities are caught in a planning paradox: making massive infrastructure commitments based on assumptions about car ownership and usage patterns that autonomous vehicles will likely shatter within the decade. The disconnect isn't just expensive — it's about to reshape urban life in ways most city planners haven't begun to consider.
Consider what happens when car ownership drops from 91% of American households to, say, 40% over the next fifteen years. That's not speculation — it's the inevitable math of ride-sharing becoming cheaper than owning. When a robotaxi can pick you up in three minutes and costs less per mile than insurance payments alone, why would you keep a car in your driveway?
But our cities are doubling down on the opposite bet. Minneapolis is building light rail stations surrounded by massive park-and-ride lots. Denver is expanding its highway system under the assumption that traffic will keep growing. These projects represent billions in public investment designed for the current reality of individual car ownership — a reality that's already disappearing in major cities.
The economic implications are staggering. Parking requirements inflate housing costs, force businesses to devote prime real estate to empty spaces, and make dense development nearly impossible in many areas. A typical apartment building in Los Angeles spends $50,000 per unit just on required parking. Once those requirements become meaningless, the entire economics of urban development shift overnight.
Transit agencies face an even starker reckoning. Light rail systems cost billions and take decades to build, justified by ridership projections that assume people need fixed routes between specific points. But autonomous vehicles offer point-to-point service that's faster, more convenient, and increasingly price-competitive. Phoenix's light rail extension, currently under construction, may serve a transportation landscape that no longer exists by the time it opens.
Some cities are beginning to hedge their bets. Portland is piloting "flex zones" that can be converted from parking to housing or retail as demand changes. Columbus won a federal smart city grant partly by proposing infrastructure that can adapt to autonomous vehicles. These are exceptions, though. Most American cities are still planning like it's 2010.
The speed of change makes this particularly urgent. Autonomous vehicle technology isn't advancing on the timeline of highway construction. Tesla's Full Self-Driving is already handling complex urban scenarios. Waymo is expanding its robotaxi service. Chinese companies are deploying autonomous vehicles at scale. The transition won't be gradual — it will be a cascade, with adoption accelerating rapidly once the technology crosses reliability and cost thresholds.
- Reduce or eliminate parking requirements in new construction
- Design streets that can be easily reconfigured as traffic patterns change
- Plan transit investments that complement rather than compete with autonomous vehicles
- Prepare zoning codes for the conversion of parking lots to housing and retail
The opportunity cost is enormous. Every billion spent widening highways is a billion not invested in fiber-optic networks, renewable energy infrastructure, or housing. Cities that recognize this shift early will have competitive advantages in attracting residents and businesses. Cities that don't will be stuck with expensive infrastructure designed for problems they no longer have.
This isn't an argument against all transportation infrastructure — cities will still need roads, bridges, and transit systems. But the specific investments being made today often assume static demand patterns just as those patterns are about to be revolutionized. The city that figures out how to plan for 2035 instead of 1995 will be the one that thrives when the autonomous future arrives.
The irony is that many of these infrastructure decisions are being made by the same officials who acknowledge that transportation is changing rapidly. They understand the technology is coming — they just can't seem to incorporate that knowledge into their planning process. Perhaps because admitting that your current projects might be obsolete is not great for getting them funded.
But reality has a way of asserting itself regardless of our planning assumptions. American cities can either start preparing for the transportation revolution or get caught building the infrastructure equivalent of video rental stores. The choice is still ours, but not for much longer.