Congestion can be perceived as an unavoidable consequence of the usage of scarce transport resources, particularly if they are not priced. The last decades have seen the extension of roads in urban areas, most of them free of access. Those infrastructures were designed for speed and high capacity, but the growth of urban circulation occurred at a rate higher than often expected. Investments came from diverse levels of government intending to provide accessibility to cities and regions. There were strong incentives for the expansion of road transportation by providing high levels of transport supply. This has created a vicious circle of congestion, which supports the construction of additional road capacity and automobile dependency. Urban congestion mainly concerns two domains of circulation, often sharing the same infrastructures:
· Passengers. In many world regions, incomes have significantly increased; one automobile per household or more is becoming common. Access to an automobile conveys flexibility in terms of the choice of origin, destination, and travel time. The automobile is favored for most trips, including commuting. For instance, automobiles account for the bulk of commuting trips in the United States. The majority of automobile-related congestion is the outcome of time preferences in the usage of vehicles (during commuting hours) as well as a substantial amount of space required to park vehicles. About 95% of the time, an automobile is idle, and each new automobile requires an additional footprint.
· Freight. Several industries have shifted their transport needs to trucking, thereby increasing the usage of road infrastructure. Since cities are the leading destinations for freight flows (either for consumption or transfer to other locations), trucking adds to urban congestion. The “last mile” problem remains particularly prevalent for freight distribution in urban areas. Congestion is commonly linked with a drop in the frequency of deliveries tying additional capacity to ensure a similar level of service. The growth of home deliveries due to e-commerce has placed additional pressures, particularly in higher density areas, on congestion in part because of more frequent parking.
Congestion in urban areas is dominantly caused by commuting patterns and little by truck movements. On average, infrastructure provision could not keep up with the growth in the number of vehicles, even more with the total number of vehicles-km. During infrastructure improvement and construction, capacity impairment (fewer available lanes, closed sections, etc.) favors congestion. Significant travel delays occur when the capacity limit is reached or exceeded, which is the case of almost all metropolitan areas. In the largest cities such as London, road traffic is slower than it was 100 years ago. Marginal delays are thus increasing, and driving speed becomes problematic as the level of population density increases. Once a population threshold of about 1 million is reached, cities start to experience recurring congestion problems. This observation must be nuanced by numerous factors related to the urban setting, modal preferences (share of public transit), and the quality of existing urban transport infrastructures.
Large cities have become congested most of the day, and congestion was getting more acute in the 1990s and 2000s and then leveled off in many cases. For instance, average car travel speeds have substantially declined in China, with many cities experiencing an average driving speed of less than 20 km/hr with car density exceeding 200 cars per km of road, a figure comparable to many developed economies. Another important consideration concerns parking, which consumes large amounts of space and provides a limited economic benefit if not monetized. In automobile-dependent cities, this can be very constraining as each land use has to provide an amount of parking space proportional to their level of activity. Parking has become a land use that significantly inflates the demand for urban land.
Urban mobility also reveals congestion patterns. Daily trips can be either mandatory (workplace-home) or voluntary (shopping, leisure, visits). The former is often performed within fixed schedules while the latter complies with variable and discretionary schedules. Correspondingly, congestion comes in two major forms:
· Recurrent congestion. The consequence of factors that cause regular demand surges on the transportation system, such as commuting, shopping, or weekend trips. However, even recurrent congestion can have unforeseen impacts in terms of its duration and severity. Mandatory trips are mainly responsible for the peaks in circulation flows, implying that about half the congestion in urban areas is recurring at specific times of the day and on specific segments of the transport system.
· Non-recurrent congestion. The other half of congestion is caused by random events such as accidents and unusual weather conditions (rain, snowstorms, etc.), which can be represented as a risk factor that can be expected to take place. Non-recurrent congestion is linked to the presence and effectiveness of incident response strategies. As far as accidents are concerned, their randomness is influenced by the level of traffic as the higher the traffic on specific road segments, the higher the probability of accidents.
Behavioral and response time effects are also important as in a system running close to capacity. For instance, braking suddenly while driving may trigger what can be known as a backward traveling wave. It implies that as vehicles are forced to stop, the bottleneck moves up the location it initially took place at, often leaving drivers puzzled about its cause. The spatial convergence of traffic causes a surcharge on transport infrastructures up to the point where congestion can lead to the total immobilization of traffic. Not only does the use of the automobile have an impact on traffic circulation and congestion, but it also leads to the decline in public transit efficiency when both are sharing the same road infrastructures.