Regional transport research documents the induced demand phenomenon across 34 highway expansion projects, finding the pattern consistent enough to function as a predictive model whose predictions transport ministries find inconvenient
Asia-Pacific Traffic Study: New Road Solved Congestion for 18 Months Then Created Its Own Congestion
KUALA LUMPUR — The Asia-Pacific Transport Research Centre released a study Wednesday tracking 34 highway expansion projects across nine markets over an average of seven years post-completion, finding that all 34 projects reduced congestion at their target location in the period immediately following completion, that the average congestion reduction lasted 18.3 months, and that after 18.3 months all 34 projects had generated traffic volumes at the expanded capacity that produced congestion equivalent to or greater than the congestion that motivated the expansion. The study’s lead researcher, Dr. Rahmat Bin Ishak, described the pattern as “the clearest example of induced demand in the transport infrastructure literature,” a term referring to the well-documented phenomenon in which road capacity expansion generates new vehicle trips that fill the expanded capacity, rather than distributing existing trips more comfortably across the new capacity. He confirmed that induced demand is “not a new or contested finding,” that it has been “consistently documented in transport research since the 1960s,” and that “the 34 projects in this study proceeded despite this documentation, which is itself a finding of a different kind.”
“The research is not ambiguous,” Dr. Rahmat said at a Kuala Lumpur briefing attended by transport ministry officials who had been invited, had confirmed attendance, and had sent three junior officers to attend in their place. “Road capacity expansion in congested urban environments generates new vehicle trips at rates that fill the new capacity within predictable timeframes. This is what the literature says. This is what our 34 cases confirm. The 18.3-month figure in our data is consistent with the 12-to-24-month timeframe documented in equivalent studies in the United States, Europe, and Australia. The finding is stable across contexts. What varies is not the phenomenon but the degree to which transport planners acknowledge it in their project justifications, which in our 34 cases was, on average, limited.”
The 34 Projects and What Happened to Them
The study documents each project’s congestion reduction period, the mechanisms by which new trips were generated, and the resulting return to congestion. The new trip generation mechanisms were consistent across cases: some new trips came from drivers who had previously avoided the congested route and now used it; some came from route changes by drivers using alternative roads that now shifted to the expanded highway; some came from mode shifts in which drivers who had previously used public transport chose to drive given the improved road conditions; and the largest category in all 34 cases came from what the study calls “latent demand activation” — trips that had not been taken because congestion made them unattractive and that were taken once congestion was reduced. The latent demand category’s size in each project predicted the speed of the congestion return: projects in high-density markets with significant latent demand returned to congestion faster than projects in lower-density markets where latent demand was smaller.
The fastest congestion return in the study was 7 months, at a highway widening in a Southeast Asian capital city where the study found that commercial vehicle use of the expanded capacity had been substantially underestimated in the project’s traffic modelling. The slowest was 34 months, at a bypass road in a smaller city where the project genuinely redistributed traffic from a congested inner route and where the population growth that would have eventually generated equivalent congestion on the new route had not yet occurred within the study window. Dr. Rahmat described the 34-month case as “the study’s most optimistic example, in which the project achieved something closer to its stated objective,” while noting that “the traffic modelling for that project projected that the congestion benefit would be permanent, which is a different claim from 34 months, and one the study’s follow-up period could not assess.”
The Transport Ministry Response
The three junior officers who attended on behalf of their ministry confirmed they would “share the findings with the relevant teams” and asked whether the study had “recommendations for highway expansion projects.” Dr. Rahmat said the study’s recommendations covered: accurate traffic modelling that incorporates induced demand rather than assuming static trip numbers; project justification processes that acknowledge the likely congestion timeline; parallel investment in public transport and active travel infrastructure to provide alternatives that reduce latent demand activation; and “mode shift programs that change the conditions that make driving the rational choice in the first place, which is the only intervention that addresses the root cause rather than the symptom.” The three junior officers noted the recommendations. Dr. Rahmat confirmed he did not expect all four recommendations to be implemented before the next highway expansion project in any of the nine surveyed markets was approved. He said he was “aiming for one, in the most receptive market, in the most favourable political cycle.” He said he had been working on which one for the past four years.
What Commuters Know and What It Changes
The study included a commuter survey asking 4,200 regular users of the 34 expanded highways whether they expected the congestion improvement to be permanent when the project was announced. Twenty-three percent said yes, they expected the improvement to be permanent. Sixty-one percent said they expected improvement but were uncertain how long it would last. Sixteen percent said they expected the improvement to be temporary. The 16 percent were, in all 34 cases, correct. The study’s longitudinal tracking found that the 16 percent who expected temporary improvement changed their commute behaviour more quickly once congestion returned — they were more likely to adjust their travel time, explore alternative routes, or use public transport — than the 23 percent who had expected a permanent improvement, who showed a pattern of persistent route loyalty despite the return of congestion that the study’s behavioural analysis team described as “consistent with the emotional investment of a commuter who has been promised a specific outcome and is reluctant to conclude that the promise has expired.” Dr. Rahmat said this was “one of the study’s more human findings” and that “the transport minister who announced the project at the ribbon-cutting was almost certainly in this group.”
