The U.S., it is not controversial to point out, is a car-oriented country. A new report from the Brookings Institution explains this reality in stark numbers. On average, Americans travel about 21 miles per day. Every trip taken to work, to the store, to school, or to any other regular destination in the U.S. requires an average of about 7 miles of travel. The majority of this travel happens with a car.
These numbers were found by analyzing a relatively new source of travel behavior data that can be derived from mobile phones. Using de-identified geolocation data collected from mobile phones and tracking the distance and travel time of trips between Census tracts in six metropolitan areas before the pandemic, Brookings researchers learned not just how much people are driving in different parts of the country, but where they’re driving to and from. Combining this data with information on land use and urban development, they can draw a clear line between how a place is built and how much driving results.
But this precise geolocation data can do more than tell car-dependent Americans what they already know. It can also show urban planners and developers where new neighborhood-oriented services and amenities can be placed to reduce peoples’ reliance on their car for nearly every daily need.
According to Brookings’ Adie Tomer, coauthor of the report, this relatively new source of data can create a precise picture of travel behavior that stands to greatly improve the way urban areas are understood and developed. “Inside government, other research shops, academia—folks are starting to understand the power of this data,” he says.
The Brookings analysis found that work commutes tended to be the longest trips people made, with an average distance of about 11 miles, while trips to school tended to be shortest, at an average distance of about 3 miles. But travel distance varies depending on the place and its urban form. The type and age of neighborhoods influences driving distances, with urban cores and mature suburbs showing similar average trip distances, and longer trips in newer suburbs and exurbs. Driving levels also varied depending on the metropolitan area analyzed, with the Kansas City metro area having an average distance of 8.2 miles for all kinds of trips, while the average in Portland, Oregon, was just 6.2 miles.
“Even just comparing the average neighborhood in Portland to Kansas City, the difference can be 1,600 miles [of driving] per year. That’s a significant amount of gas, it’s significant wear and tear on one’s vehicle,” Tomer says. “There are real financial consequences for those households.”
Part of the explanation is as simple as the concept of density—places that are more dense require less driving, while less-dense places require more driving. But Tomer says there’s another level of causation here, and it lies in how traffic demands are measured.
Typically, transportation engineers rely on a standard known as level of service, or LOS: a letter grade that assesses whether a given road is meeting or exceeding its expected flow of traffic. Traffic flows are measured by small black tubes laid out on streets that count every car that drives over them. If traffic isn’t moving at the expected speeds, that has historically been interpreted as a problem. One way to solve it has been to divert traffic elsewhere. Another has been to simply build more car lanes.
“The second answer has been our primary response, and it’s happened in every single state, in every single metropolitan area across the country,” Tomer says. “It has not just been a preferred method of analysis, it has become dogmatic. And because it has lasted for decades, we are seeing the full results of that approach, which is a dramatic increase in lane miles.”
New-road building has outpaced population growth, according to a Brookings analysis. More lanes means more driving.
But this road-building spree has been based on limited information. The traffic counts used to determine LOS are a blunt tool that offers information about only a few streets, and only about the raw number of cars driving them. The precise origin-to-destination travel information offered by anonymized and encrypted mobile phone data, Tomer says, can be a much more comprehensive way to understand how traffic is flowing and where people are going. This data about the starting point, distance, duration, and ending point of trips can be used to better understand what type of travel is happening, and when long trips may be reduced by bringing services or new development closer to where people live.
Human-scaled development, even in the suburbs, can help reduce driving distances by enabling more of a person’s daily needs to be met in closer proximity to where they live. Policies that encourage this kind of development, Tomer says, can help reduce the amount people have to drive to meet their daily needs. He points to policies that states can adopt including requiring impact fees for new developments that shift some of the costs of transportation infrastructure to developers, thereby disincentivizing the kind of far-out exurban housing that leads to so much driving. Mobile phone data can also help make the case for imposing demand-based pricing on parking where its free availability may be leading more people to drive distances easily traveled on foot or by bike.
“This data should really create a new sense of urgency behind adopting many of the public policies that have had some friction politically to become adopted,” Tomer says. With more precise data explaining where and how far people are driving, policymakers and developers may be able to start building places that require less of it.