Widening a congested highway seems like an obviously correct engineering response — more lanes, more capacity, less congestion. Decades of traffic data from cities around the world tell a more frustrating story: congestion on newly widened highways frequently returns to roughly its previous level within a few years, sometimes faster, even though the road now has considerably more physical capacity than before. The extra lanes didn't fail. They worked exactly as an economist would predict, given what happens when you make driving cheaper.
Travel time is a price, and lower prices increase demand
Traffic engineers describe this phenomenon as induced demand, and the underlying logic is the same basic supply-and-demand relationship that applies to any good with a price: the "price" of using a road includes the time cost of sitting in traffic, and when a widened highway temporarily reduces congestion, it effectively lowers that time-price of driving that particular route. Lower prices increase demand for almost any good, and travel is no exception — trips that people previously avoided taking during rush hour, or took via a different route, or skipped entirely, or made by transit instead, become newly attractive once the widened road makes driving faster, and that latent demand fills the new capacity, often to roughly the same congested equilibrium as before.
Where the extra traffic actually comes from
The added drivers filling a widened highway back up come from several overlapping sources: some are people shifting from other, less convenient routes onto the now-faster highway; some are people shifting from public transit, since driving has become relatively more attractive compared to the alternative; some are trips that get made at rush hour instead of being shifted to an off-peak time; and over a longer timescale, some reflect genuine changes in land use and development patterns, as easier access encourages more housing, retail, and commuting further from city centres than would have been practical before the widening. None of this means road capacity never matters — a road that's dramatically undersized for its area will still benefit from expansion — but it does mean that simply adding lanes to a congested urban highway, expecting congestion to permanently drop, routinely runs into the mechanism working directly against that expectation.
What we're still unsure about
The existence of induced demand as a real, measurable effect on urban highways is well supported by traffic studies across many cities and is not seriously disputed among transportation researchers. What remains more actively debated is the appropriate policy response: economists and planners disagree over how much to rely on alternative strategies like congestion pricing, transit investment, or land-use reform versus more targeted, selective road expansion, and predicting exactly how much induced demand a specific widening project in a specific city will generate involves enough local variability in land use, transit availability, and driver behaviour that precise forecasting remains genuinely difficult even with the general mechanism well understood.
This sits inside Transportation Engineering & Road Design, one of seven topics in Civil Engineering, one of four domains in Engineering, one of seventeen subjects the app can quiz you on.