Historical activity-transition rules can misallocate travel demand: Nationwide evidence from US
This study demonstrates that relying on historical activity-transition rules for travel demand forecasting can significantly misallocate demand across purposes, as evidenced by a nationwide weakening of activity dependencies between 2009 and 2017 that led to an 8.9% underestimation of work and school-related transitions despite stable total travel volumes.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Every day, millions of people move through their cities and towns, weaving together a complex tapestry of errands, work, and family life. To plan the roads, trains, and buses that keep these movements flowing, transportation engineers rely on forecasts. These forecasts are not crystal balls; they are built on a simple, practical assumption: that the way people connect their daily activities today will look much the same in the future. If a person typically goes from work to the grocery store, planners assume that pattern will hold true decades from now, allowing them to calculate how many cars will need to pass a certain intersection or how many seats a bus line requires. This reliance on historical patterns is the bedrock of long-term infrastructure investment, guiding decisions that commit public resources for generations.
However, a fundamental question has lingered in the background of this planning process: do these connections between activities actually stay the same over time? It is one thing to know that people go to work and shop; it is another to know if the specific sequence of those events remains predictable. If the rules that govern how one activity follows another change, then the forecasts built on old rules could be sending planners down the wrong path. They might build a highway for a type of traffic that no longer exists, or underfund a service that has suddenly become essential. Understanding whether these daily habits are truly stable, or if they are quietly shifting beneath our feet, is crucial for ensuring that the transportation systems of tomorrow can actually serve the people of tomorrow.
A team of researchers from universities across the United States and abroad set out to test this assumption using a massive collection of travel data. They examined nearly 623,000 days of travel recorded by the U.S. National Household Travel Survey in 2001, 2009, and 2017. By looking at the specific order in which people performed their daily tasks—such as going from home to work, then to a child's school, and finally to the grocery store—the team could measure how predictable these sequences had become over nearly two decades. They were not just counting trips; they were looking for the invisible threads that link one activity to the next, asking whether knowing what a person is doing right now still tells us what they will do next.
The researchers discovered that while the overall number of trips people took and the variety of activities they did changed gradually, the predictability of their daily sequences underwent a dramatic shift. Between 2001 and 2009, the rules governing how one activity followed another remained remarkably stable. But by 2017, that stability had eroded significantly. The connection between consecutive activities had weakened so much that knowing a person's current activity provided far less information about their next one than it had just eight years prior. This change was not limited to a specific group of people or a single region; it was a nationwide phenomenon observed in 49 out of 51 states and across every demographic group studied, from urban workers to rural residents, and from young families to retirees.
Crucially, the team found that this shift could not be explained simply by changes in the population itself. One might assume that as the country's demographics changed—perhaps with more people working from home or different family structures—the mix of daily activities would naturally shift. To test this, the researchers created a scenario where they kept the actual activity patterns from the past but applied them to the population makeup of 2017. The result showed that even with the correct population numbers, the old rules failed to reproduce the new reality. The change was driven by a fundamental reorganization of how people were choosing to link their activities, a shift that occurred within groups of people who shared the same age, income, and job status.
The consequences of ignoring this shift are tangible and substantial. When the researchers applied the old rules from 2009 to the actual travel data from 2017, they found a significant mismatch in where people were actually going. Specifically, the historical rules underestimated the number of trips ending at work, school, daycare, or religious activities. In reality, there were 8.9 percent more trips to these essential destinations than the old rules predicted. On an average day in 2017, this discrepancy amounted to approximately 12.4 million trips that were misallocated. This means that planners relying on the old models would have been systematically underestimating the demand for services and infrastructure related to work and education, while potentially overestimating demand for other types of travel.
This study serves as a vital warning for the future of transportation planning. It demonstrates that the relationships between daily activities are not permanent fixtures; they can evolve quietly and broadly without a major global disruption like a pandemic to signal the change. The findings suggest that relying on historical data to forecast future needs is no longer sufficient if that data is not regularly checked against current behavior. For the engineers and policymakers who design the systems that move our world, the lesson is clear: the rules of the road for human behavior are changing, and the maps we use to navigate the future must be redrawn to reflect the reality of how people are actually living their lives today.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.