Planning and the Complicated Causes and Effects of Congestion

If you're like me, you grew up waiting out traffic jams in the backseat of cars, wondering why the cars at the front of the traffic jam didn't just drive faster and let the rest of us through. 

I misunderstood the concept of congestion, but not entirely. Traffic isn't caused by the impossible scenario of a randomly assembled line of drivers simultaneously deciding to stop, in formation, on a giant freeway (until very recently, that is). But a single driver traveling much slower than the flow of traffic, or suddenly braking, can create a traffic jam in their wake.

The random slow driver is only one explanation among many for congestion, however. While my youthful conception of congestion kind of, accidentally, hit the mark, it still wildly underestimates the difficulty of solving congestion. I am far from alone in underestimating the complexity of traffic congestion, and the challenges of solving it. One of the most tacitly obvious solutions for traffic jams, adding more lanes to the road, is just as likely to end congestion as my imagined slow-drivers would be to suddenly speed up. 

Congestion is one of the most common complaints about modern life. In the post-automobile era, congestion is also one of the most common complaints about the outcomes of planning. The scrutiny is fair, to a certain extent: Few realities of contemporary living so broadly and conspicuously intersect with so many planning practices. In addition to transportation planning, which paves the way for freeways and bike lanes and bus stops, congestion is also a direct outcome of land use planning decisions.

But the public is just as guilty of contributing to the frustrations of congestion as the field of planning. Despite its consistent, direct impact on the experiences of people living in almost every developed corner of the world, congestion is still widely misunderstood by the public, heavily debated by leading scientists, and co-opted for loosely connected purposes by politicians and culture warriors. The expectation that planners might solve congestion with so many misconceptions and so much misunderstanding, cultural politics, and pseudoscience swirling around the issue is unfair  to a certain extent. 

This article, the first in a series, will begin separating fact from fiction, accounting for the unknowns and mixed messages of congestion and laying out a conceptual framework for a better understanding of the nature of congestion. The goal of this discussion is a more honest and rational discussion of the role of the field of planning in studying, creating, and responding to congestion. Part one will start by exploring the causes of congestion before future installments examine the many ways planning responds, effectively or not, to congestion.

What Is Congestion?

Congestion is caused by a supply and demand imbalance: more cars are on the road than space on the road allows. Depending on the time of day, and the surrounding population and workforce densities, and the conditions of the road and its intersections, demand for the roadway can increase up to and beyond the point of saturation—when the volume of cars using the road is greater than the capacity of the road. Congestion ensues.

We know congestion when we see it: the road is filled with cars at a complete stop or lurching, slowly, forward. Traffic congestion produces slower travel speeds, delayed arrival times, and increased fuel consumption—a recipe for widespread frustration. With so much at stake in transportation—returning to family, getting to work on time, making it to the hospital on time—it's understandable that congestion sometimes feels like the single most frustrating obstacle between humans and happiness.

Despite its familiarity in real life, congestion is notoriously difficult to model in a scientific setting. Scientists have been studying congestion—presenting models, refining models, suggesting further research, and going back to the drawing board—for as long as automobiles have been clogging roads. According to an article by James MacDonald for JSTOR Daily, scientists have attempted to model congestion as a fluid, a gas, or even as grains. 

"Many researchers have built complex mathematical models. The best models use real-life data, and incorporate such factors as traffic volume, velocity, and driver reaction time. The more lanes there are, the more complicated the model becomes." -James MacDonald

While advancements in computing and machine learning promise a more accurate science of congestion, the lack of scientific consensus around the fundamental nature of congestion contributes to the lack of real-world, scalable solutions for the problem. Problems without solutions, especially such a common problem, inspire public debate, informed or not. 

Congestion Classification

The Highway Capacity Manual, published by the Transportation Research Board (TRB) of the National Academies of Sciences, Engineering, and Medicine, uses a term called Level of Service (LOS), with letter grades A-F, to classify vehicle traffic as compared to roadway capacity. The LOS classification system therefore provides a highly systematic (and highly contested) manual for measuring congestion. 

Level of Service is also used in many state, local, and regional jurisdictions to measure the impacts of proposed developments or zoning changes, making it an essential term in the practice of both transportation and land use planning, and essential to an understanding of the role of planning as a response to congestion. In one prominent example of the importance of LOS, the California Environmental Quality Act (CEQA) for decades used LOS as its standard to measure the environmental impact of proposed developments and zoning changes. 

What Causes Congestion?

Here's where it starts to get tricky. Despite all their direct experience with congestion, most people can't list all the reasons why congestion exists. Some explanations for congestion are counterintuitive and some are inconspicuous—like most other complex features of modern living. But given the ubiquity of congestion, many explanations for congestion are also far too convenient to be accurate. Congestion is often the lie we tell ourselves.

Driving in Cars

Every time drivers angle cars out onto the road, they make a choice to create and wait in congestion. The more people drive, the more congestion there will be, in more locations, and at more times of day. Take all the cars away and there would be no traffic jams, that much is obvious. "You aren't stuck in congestion," says a famous altruism, "you are congestion." 

Many people have few alternatives to their driving habits. Much of the world's built environment has been planned and designed to be driven, which is especially true in the United States. Proponents of the car-centric planning status quo say cars are necessary to participate in society and the economy. The car culture of the United States adds extra dimensions of pressure to adopt the car as the primary mode of transportation. New roadway capacity induces driving. Parking induces driving, and drivers add to driving distances and congestion while searching for parking. The separation of uses (as a result of a century's worth of Euclidean zoning codes) induces driving. Even when Americans aren't driving, they are bombarded with ads imploring them to drive instead of bike or take transit, with promises of magical cars, capable of flying around and even over traffic jams. The car is one of a few status symbols, along with a house, that signifies the "American Dream."

Americans are stuck with congestion, just as much as they are stuck in congestion.

Collisions

Driving is terribly unsafe. You've probably heard the saying that you're much more likely to die in the car on the way to the airport than you are to die in an airplane crash. In 2021, 40,000 people died in traffic collisions in the United States. That figure is relatively small compared to the total number of collisions every year. More than 12 million vehicles were involved in crashes, fatal or not, in 2019. The point of that data is not to say, "Go ahead and drive; you'll probably survive a collision." The point is that crashes happen. Often. Choosing to drive is choosing to accept that risk and its consequences.

Congestion is one of the consequences of collisions. Crashes show a repeated, demonstrated ability to slow traffic around the crash. Many crashes require the reduction of roadway capacity, to make room for the vehicles involved in the crash or for first responders responding to the crash. When there's less capacity on the road, the saturation point for congestion occurs much faster. Crashes also create congestion when passing drivers slow to safely pass the crash site or to get a look at the carnage. The practice of "rubbernecking" to see a crash frequently create cascading congestion effects.

Unfortunately, over the many decades since automobiles became household items in the United States, many of the engineering techniques used to reduce automobile collisions and mitigate the congestions effects of collisions—straightening roads, adding buffer zones, and adding roadway capacity, to name a few examples—have also contributed to higher vehicle speeds and automobile dependence that perpetuates the cycle of traffic violence. The book, Confessions of a Recovering Engineer, written by Charles Marohn, offers detailed, specific examples of how the transportation engineering profession perpetuates this cycle. 

Unfortunately, planners often have very little say about the details of street design and the deployment of traffic safety infrastructure—politicians and street and engineering departments often hold the keys, so to speak, to road safety. Road safety plans, complete streets plans, and Vision Zero goals almost always face significant technocratic and political hurdles to actual traffic safety improvements that could reduce the amount of congestion caused by collisions.

The engineering know-how and technology exists right now to eliminate the worst traffic collisions from the roads. The world, collectively, is making a deliberate choice to allow the carnage and congestion continue.

Slow Moving Vehicles

It's come up a few times so far, but slow moving vehicles can cause congestion through a phenomenon identified by researchers as "shockwaves" or "butterfly effects." One car suddenly slows in the normal flow of traffic, causing other cars to slow or merge, causing yet more cars to slow or merge, and so on. The phenomenon has even been duplicated in a research setting.

Changing lanes and failing to stop for an obstruction are also the frequent causes of collisions, so shockwaves can contribute to traffic congestion in more than one scenario of cascading effects. 

Road Capacity and Induced Demand

A lack of road capacity is the most common explanation for persistent congestion, and thus more roadway capacity is the most commonly proposed solution. It seems self-evident that more road—more lanes and fewer obstacles like intersections and traffic controls—solves traffic congestion. One way to solve a supply-demand imbalance is to add supply, according to the basic lessons of economics. 

The problem with that nice little solution, however, is a phenomenon known as induced demand. Writing for Curbed in 2018, Patrick Sisson explained induced demand as follows: "there is always an underlying demand for driving, which exceeds the road space available, so building more roads induces more traffic. Congestion soon returns but with more vehicles affected than before." So, more road capacity can actually cause more congestion by creating vehicle trips that otherwise wouldn't have happened. 

In 2011, The American Economic Review published a study that quantified how roadway capacity induces driving and, thus, congestion. According to the study, driving, as measured in the study as the number of vehicle-kilometers traveled (VKT), increases in direct proportion to the available lane-kilometers of roadways. Counter-intuitive as it might seem that building road capacity creates traffic, rather than functioning as an effective response to congestion, the researchers referred to their calculation of induced demand as "the fundamental law of congestion." That study, and the phrase, is cited in related studies and media coverage to the present day.

Advocates have called out the transportation engineering and planning fields for willfully misrepresenting or disregarding the effects of induced demand when proposing new automobile infrastructure projects. Despite all of the evidence that new roadway capacity will not solve congestion, departments of transportation and politicians haven't gotten the message. Powerful economic interests still control the politics of congestion relief. 

Urbanization and Sprawl 

The connection between roadway capacity and congestion calculated in the "fundamental law of congestion" raises another fundamental implication: congestion is created by cities and settlements. Even before the advent of the automobile, streets and roads were an essential part of cities and towns—the connections between buildings are a necessity of modern urban environments. When a settlement includes roads, and people need to drive to navigate between destinations, congestion will be included in the deal. 

How roads and land uses are arranged, the primary concern of the planning field, has a major impact on how much congestion occurs. Researchers have long attempted to calculate the connections between population density and congestion, and have recently made advancements in connecting land use to transportation outcomes, but the science of congestion as a result of planning and development is still in development. (More on the evolving understanding of the connections between land use and transportation congestion can be found here.)

Some of the congestion in developed areas can be attributed to planning decisions about how the street grid is laid out. Some of the congestion in developed areas is created by how the various land uses are laid out—residential, commercial, retail, open space, and industrial.

Suburban sprawl, which in most cases in the United States, is built on a car-centric planning model that ensures large numbers of car trips—both to and from the urban core during work commutes and for cultural or social reasons and in an around the suburbs and exurbs on the periphery of metropolitan areas. An example of how sprawl induces vehicle traffic has been visible during the Covid-19 pandemic, when congestion shifted to different locations as well as different times during the day because of people moving to the suburbs and working from home.

Affluence and Economic Growth

The most successful cities and regions are full of congestion. The correlation between economic growth and increased driving (as measured by Vehicle Miles Traveled, or VMT) is well documented. The U.S. Census Bureau even calculates an expected 14 percent increase in VMT through 2030, 500 billion miles in all, if the country hits its expected economic growth targets over the same period. There is also evidence that wealthier people drive more, as reported by researchers in 2021 via Transfers magazine.

The increase in driving and congestion associated with economic growth and prosperity implies its opposite: that economic contraction, otherwise known as recessions, will reduce car trips and congestion. Evidence from the Great Recession, in 2009, supports that implication.

The connections between economic prosperity and the affluence of drivers to congestion will be relevant to the discussion about congestion pricing—one of the planning efforts proposed, and debated, as a solution for congestion—later in this series.

Traffic Controls

Traffic jams are often found around traffic controls like lights, stop signs, and toll collection. While traffic controls are necessary to ensure the safe navigation of intersections (always one of the most dangerous places to drive a car), traffic lights and stop signs can be deployed without consideration of the traffic volumes that use the street. Traffic lights can be poorly timed or unresponsive to conditions on the intersecting streets, creating large clusters of vehicles where previously there had been free flowing traffic—sometimes while waiting for a light to turn for an empty intersection. Stop signs also stop the flow of traffic to one vehicle at a time, when stop lights might be a more effective tool for moving large numbers of vehicles through an intersection while minimizing delay. New technological applications, like cloud-based computing and machine learning, are being sold, with obvious conflicts of interest, as a potential game-changing addition to traffic control.

Like so many of the factors that cause congestion, traffic controls are both blamed for congestion and proposed as a solution for congestion. We'll examine how effective these tools, and many others, are for reducing congestion in the forthcoming installments of this series.

Part two of this Planetizen series on planning and congestion was published the following week.