By Chris McCahill
The concept of induced demand is now widely recognized in transportation. But we often treat it as a technical modeling issue rather than a basic principle of how people respond to the world around them. Build for cars, and you’ll get more driving; build for transit or biking, and you’ll get more of those too. Shifting the conversation in that direction can improve near-term decisions and strengthen communication between transportation professionals and the public.
We have plenty of evidence for induced demand from new highways. A common rule of thumb says that for every 10% increase in highway capacity, we see roughly a 10% increase in traffic, or vehicle miles traveled (VMT), within a few years. A study out of New Jersey this year showed a similar effect on local roads—every 10% increase in lane miles led to a 1% to 3% increase in VMT.
These numbers are helpful, but they often end up being used mainly to push back on highway expansions—less often to make the case for other kinds of investments. One exception is California’s SB 743, which explicitly encourages multimodal strategies to reduce VMT.
A new study adds more depth by looking at induced demand across modes. Using 19 years of data from 11 regions in the UK, the researchers examined how infrastructure investments shape travel behavior. Not every finding was statistically significant, but the key takeaways are compelling. For example, a 10% increase in light rail track is linked to an 8% to 9% increase in light rail use—an even larger response than they found for highways. They also found a 10% increase in cycle track is linked to about 0.5% more biking.
For planners and modelers, quantifying these effects is useful. But the larger point is straightforward: our transportation investments shape how people travel.
Where models fall short
Some induced demand effects are easier to anticipate, like how people will adjust their routes or departure times when capacity changes. Some advanced models are better at accounting for this than others. Other effects, especially development patterns, are more complicated and typically get baked into baseline assumptions.
A recent Caltrans report highlights this challenge. Many regions in California rely on a single future land-use scenario, which essentially presumes a certain pattern of growth before any modeling even begins. For example, if an outlying suburban area is assumed to grow with low-density housing, that assumption drives future travel demand estimates, and often justifies widening the highways that would enable that growth.
But what if agencies tested a different vision? Suppose planners modeled a scenario anchored by light rail. The UK study suggests that kind of investment could drive major increases in ridership. And new research from Dijon, France, shows that light rail can catalyze infill and mixed-use development on certain corridors, especially when bolstered by local land use planning and development efforts.
Right now, though, the system is stacked against these multimodal alternatives. As we’ve written before, forecasts tend to overestimate traffic more often than they underestimate it, signaling that induced demand is already built into many of our assumptions. That, in turn, locks in past development patterns and keeps us on an auto-dependent path.
Someday, we may have AI-powered tools that can quickly model hundreds of transportation and land-use scenarios, giving decision makers a clearer understanding of tradeoffs and benefits. Until then, we should focus on investments that deliver the outcomes people actually want. A recent national survey found that 18% of Americans would prefer to live car-free, and another 40% are open to it.
Tweaking conventional models around the edges won’t get us there fast enough. Experience shows we need to be clearer about the outcomes we’re aiming for. It helps to keep in mind that what we build influences how people get around and ultimately, what we value as a society.
Photo credit: Mikechie Esparagoza via Pexels, cropped. License.