In the years leading up to the coronavirus pandemic, the intelligentsia came to a consensus that sprawling, car-dominated cities were doomed. The future, they said, was in dense, transit-dependent metropolises. The seeming success of compact cities such as San Francisco, Boston, and New York gave this theory credence. And the supposed dangers of sprawl to the climate gave it urgency.
Yet the facts show that sprawling and car-dependent cities have grown more rapidly than dense ones for decades and are far more affordable. The pandemic, meanwhile, showed they will expand even more rapidly in the future. By contrast, the climate-driven demands for density and transit are just the most recent version of a solution that has long been searching for a problem. Advocates will continue to search. In reality, sprawling cities are more environmentally sound than their dense counterparts and will become even more so as technology evolves.
Instead of warring against sprawl and cars, planners and environmentalists should recognize how the green spaces of suburbia, allied to autonomous electric vehicles and green single-family homes, can provide both the affordability and sustainability most Americans crave.
The Long Triumph of Sprawl
There has been much discussion of the benefits of density, of which there are many. If there weren’t, nobody would live in Manhattan or San Francisco. These cities allow many people, especially young, high-productivity singles and those who work in business services like finance or law, to congregate and learn from each other. Economists call these benefits “agglomeration effects.”
But too many advocates today ignore the other side of the coin, known to economists as “the demons of density.” These include things like congestion, crime, and, of course, pollution. Such problems explain why, as technology has evolved, people try to get more of the benefits of living near each other—the agglomeration effects—without the problems of living directly on top of one another. And so, despite periodic stories of a return to the city, America has been taking advantage of new technologies to become more and more sprawling: in other words, to live in places with lower density and more ease of driving by car.
Since 1950, the average density of the largest American cities has dropped from 6,000 people per square mile to 3,000. The fastest growing metros in terms of population have been the most sprawling ones. While places like New York or San Francisco have about 50 percent more people than in 1950, Houston, Dallas, and Jacksonville have 500 percent more. Atlanta, Phoenix, and Austin have almost 1000 percent more. Studies confirm that there is a direct correlation between how sprawling an area is and how fast it grows. This is not surprising, because housing prices in, for example, Texas are less than half the cost of those in California, which has some of the densest cities in the nation, and the cost of living in general is lower.
Complementing sprawl has also been a long-term trend toward people driving more and taking transit less. In 1960, about 12 percent of all Americans took transit to work. By 2020, it was about 5 percent, and the decline of non-work trips on transit was even faster. Meanwhile, the number of Americans who travel to work by car, especially those who travel alone in a car, continues to increase. The miles driven by car per capita have almost doubled. This is not because of supposed subsidies to cars; according to the Bureau of Transportation Statistics, cars pay for almost all of the cost of roads through gas taxes and other charges, while transit riders pay barely a third of the cost of their rides, and that ratio is dropping as transit use declines. For decades, in fact, more and more of the gas tax has been siphoned off to pay for increasingly expensive and increasingly empty buses and trains.
The COVID-19 pandemic has accelerated these trends, further increasing the demand for distant “exurban” development. In the past decade, exurbs grew at almost two times the national rate. But, during the pandemic year alone, construction in exurbs increased another 20 percent. A recent study for the National Bureau of Economic Research has shown that prices and rate of building have risen the fastest in the most far out parts of metro areas during the pandemic. That makes sense: people are moving farther away because, in a post-pandemic world, they won’t need to travel as often to work. The percentage of those who telework has gone from 5 percent of the workforce to what seems to be a long-term trend of 20 percent. Since this telework shift has been starkest for office workers, and since almost all transit systems are geared toward serving central business districts, the increase of telework has hit transit hard. Indeed, while passenger miles traveled by car have returned to pre-pandemic levels, bus travel is down 40 percent since 2019. Urban rail travel is down by half.
Some might long for a return to dense urban areas and a pre-automobile age. The Urban Land Institute says, “arguably, no tool is more important than increasing the density of existing and new communities.” As a recent National Geographic headline stated, “To Build the Cities of the Future, We Need to Get Out of Our Cars.” Yet those who hope for massive in-migration to transit-dependent urban areas have been disappointed for decades. And not just in the United States. In fact, the same trend toward more sprawling cities, more cars, more driving, and less transit has been observable all around the world. Densities in Paris, for instance, have dropped by half since 1950, even as the miles driven per person by car have doubled.
These tendencies represent more than the effect of one or another policy, and more than a short-term trend. Rather, they indicate a clear global and long-term preference. The pandemic has only made the shift toward the modern, sprawling city more rapid and obvious.
The Environmental Costs of Density
People choose more sprawling areas because they limit many of the downsides of urban living, from cost to congestion. But we learned in the pandemic one of the most important environmental costs of density: disease. For most of human history, cities were “demographic sinks” because births did not make up for high death rates due to infectious disease. Cities had to be continuously replenished by people moving from the countryside.
In the early 20th century, the triumph of public health against smallpox, typhoid, cholera, and other infectious diseases allowed for more urban living. Yet cities never entirely overcame the dangers of density for disease transmission. We saw this again during the COVID-19 pandemic. One study found population density accounted for up to 76 percent of the difference between infection rates in different parts of the United States. There was also a correlation between infection and density in the 1918 flu and doubtless will be in the next pandemic.
A more prosaic concern with density is air quality. The more concentrated humans are, the more likely they will be breathing each other’s pollution, whether that comes from congested streets, local industry, or simple heating and cooling devices. The larger the urban area, by contrast, the more space there is for air pollution to spread out and disperse, and this is true around the world. A study released this year in Regional Science and Urban Economics using data from Germany found what dozens of other studies have found. Simply, that “higher population density worsens local air quality.” Another recent study of U.S. cities showed “denser cities are linked with worse air quality,” and that dangerous particulate matter especially tended to increase in denser areas. By its authors’ calculations, a dense American city has dozens more pollution deaths a year, merely because of its density, than a more sprawling city of similar population. While pollution everywhere is lower than it was in the smog-choked cities of the Industrial Revolution, the move to the suburbs has been an important part of why more people breathe clean air today.
There are other local environmental costs to density as well. Urban residents have to endure a “heat island” effect, where urban concrete and asphalt amplify temperatures. The United States Environmental Protection Agency (EPA) notes that such effects can make daytime temperatures up to 7 degrees Fahrenheit hotter in cities than in rural and suburban areas, and nighttime temperatures up to 5 degrees hotter. (For perspective, the Intergovernmental Panel on Climate Change’s estimate for the total impact of global warming by 2100 is about 5 degrees Fahrenheit.)
For all the talk of urban planning for “climate resilience,” and pitches for green roofs, cooler building colors, and so forth, there has been limited discussion about how spread-out suburban areas could reduce the heat island effect even more substantially. Yet, as a recent article in Nature Communications argues, “sprawling development will lead to a better thermal environment” by creating more green space and lowering the heat island effect by several degrees. By contrast, the article notes that other strategies, such as green roofs, tend to have much more localized and limited effects. The heat island effect of dense areas in fact exacerbates the problems of local air pollution. Heat breaks down nitrogen dioxide and volatile organic compounds into ozone, which irritates eyes and lungs. This explains why air pollution is far worse in summer months and why local air pollution will be amplified by density in the future.
Perhaps the most important tool for reducing the heat island effect is trees, which provide shade and absorb solar radiation. Suburbs are, almost by definition, more verdant than cities. Famously sprawling cities like Atlanta or Houston have tree cover on more than 30 percent of their landmass, while older, denser cities such as Philadelphia, San Francisco, and Chicago have under 20 percent. Besides reducing the heat island effect, nearby trees have been shown to intercept particulate matter and absorb ozone, sulfur dioxide, and nitrogen dioxide, thus reducing local air pollution as well. The personal and psychological effects of trees are real too. Studies have shown that the presence of trees decreases stress, increases attentiveness and sense of safety and comfort, and reduces the likelihood that pregnant women will have low-birthweight babies. There is no way to have the same access to trees in dense urban areas.
Since the publication of Ian McHarg’s justly famous work Design with Nature (1969), urban planners and developers have become aware of the need to include natural landscapes in new communities. They now build depressed swards to absorb rainwater, retain hills and creeks as contributors to natural beauty, and create green pathways to allow migration of animals. Yet these features all tend to spread out development and lower density, as McHarg himself noted. His book included maps of Philadelphia that showed how social and physical ills increased with density. As he said, it was “not poverty, but density” that “bears a remarkable correspondence to the pattern of pathology” in urban living, in everything from crime to chronic disease. It should be no surprise that McHarg helped design the very livable, but sprawling, master-planned community of The Woodlands outside Houston.
There is also an assumption across much of the popular environmental literature that single-family homes are environmentally destructive, while tall buildings are green. Advocates claim that building up instead of out limits buildings’ footprints and that tall buildings require less energy to heat or cool, thus reducing greenhouse gas and other emissions. We now know this is false.
For one, taller buildings rely on steel and concrete to support themselves. These materials take five times more energy and carbon dioxide to produce than wood, which predominates in single-family or smaller homes and is a renewable carbon sink. Beyond the materials used, each additional story of a tall building requires more support beams and structures on every story of that building, which increases the ratio of material to livable space. These tendencies help explain why building taller is more expensive. Going from two to four stories increases the cost of each square foot of a building by 25 percent. Going from five to ten stories increases the cost of each square foot by over 50 percent. Those costs are the result of more—and more energy-hungry—materials.
Large buildings also use more energy to function. Tall buildings require fans to push and pull air through their HVAC systems, as well as energy-hungry pumps to lift water to the top floors. For very large buildings, elevators use up to 10 percent of all energy. Common areas such as stairwells and lobbies need to be heated, cooled, and lit, adding to both environmental and economic costs without contributing to anyone’s living space. One recent study found that “each additional story in a building is associated with a 2.4 percent increase in electricity use and 2.9 percent increase in fossil fuel use.” Taller buildings specifically tend to absorb more heat and then give it off, exacerbating the heat island effect, even as they cast shadows that limit natural light. Taller buildings, in short, create more burdens on both the local and global environment than small ones.
The Environmental Costs of Transportation
Today, much of the discussion around the environmental impact of density revolves around a single metric: vehicle miles traveled (VMT), the number of miles cars travel on the roads. The assumption, embodied in many state and local regulations, is that denser cities tend to bring destinations closer and therefore reduce VMT, which therefore reduces air pollutants and greenhouse gas emissions from burning gasoline in cars.
There are several faulty assumptions behind the attempt to reduce VMT. First and foremost, there is little correlation between density and VMT. A 2009 metasurvey of the literature by the National Research Council found that doubling residential density in an urban area, which in fact has never happened for any major city in modern history, would reduce VMT by only 5 to 12 percent. The massive costs of such doubling, and the minor reduction in car-related emissions, would make such an effort one of the least cost-efficient means to reduce carbon emissions imaginable.
Any benefits of density-related reductions in VMT are offset by other factors. For one, denser areas tend to have more congestion (think 14th Street in New York City), so traveling a mile in a dense city will require more starts and stops, and will therefore burn more gasoline, than traveling on a less- congested suburban highway or arterial road. The CO2 emissions when traveling at 5 mph are 300 percent higher than when traveling at 55 mph, and this ratio is even worse when the slower speed is due to congestion. Focusing on miles traveled, rather than actual greenhouse gas emissions, is a blunt and inopportune metric that tends to bias planners against suburban and exurban development.
More importantly, we know that the relationship between VMT and all kinds of pollution, including greenhouse gases, has weakened over time. Much of the rage against cars in previous decades came from their supposed impact on local air quality, so much so that the 1991 federal transportation act gave grants to cities to reduce driving, with the stated purpose of improving air quality Thankfully, those efforts to reduce driving failed, and, despite massive increases in VMT, the prevalence of the six major air pollutants measured by the EPA has dropped by 70 percent since 1980. The most important reason is increasingly efficient and environmentally sound cars. According to the EPA, new passenger vehicles now emit 99 percent less air pollution than they did five decades ago. The efficiency improvements explain why the concern with local automobile emissions, once the foundation for the anti-automobile movement, has almost been forgotten.
The reduction in greenhouse gas emissions from driving has been more limited since these cannot be scrubbed away through physical or chemical processes. Yet, since 1970, the average miles per gallon of the US vehicle fleet, which closely approximates gas emissions, has more than doubled, from 10.3 mpg to 24.9 mpg in 2019. This explains why despite recent increases in VMT, total greenhouse gas emissions from transportation, the majority of which is from cars and trucks, have declined since 2006. Under new federal requirements, the Biden administration aims to increase general fleet efficiency to 52 mpg by 2026, which means total emissions from cars will drop even further.
Over even a medium-term time frame, the increasing adoption of hybrid, electric, and autonomous vehicles will almost completely sever the connection between VMT and greenhouse gases. Those who are attempting to redesign cities, projects that will take decades or even centuries, merely to reduce the use of gasoline-powered cars are thus engaged in a futile exercise that will only become more futile with time. It would be like attempting to redesign cities in 1900 to reduce horse manure. The technology will change faster than the city will.
Meanwhile, anti-automobile policies can have negative environmental effects right now. For instance, in 2007, amendments to the California Environmental Quality Act required that all new housing developments show how they mitigate global warming. Most importantly, new developments had to show they would reduce VMT relative to current California standards. Yet the typical Californian today produces only nine tons of carbon dioxide a year, about half the national average. These lower emissions are not because of lower VMT, which are close to the national average, but because of California’s balmy climate and green electric grid. Yet the law, and some of California’s other supposed environmental acts, has been used to prohibit “sprawling” development and thus push people out of an otherwise climate-friendly state. Every home not built, no matter where it is located, is keeping at least two more people out of California, which is effectively doubling those persons’ carbon emissions. It is difficult to imagine many laws with such a deleterious climate impact, made worse because it exacerbates what attorney Jennifer Hernandez has called California’s “Green Jim Crow.”
Even today, mass transit is not an environmental improvement over cars. As cars have been getting more efficient, the buses that make up the majority of US public transit have been getting less so. One reason is that although total bus VMT keeps increasing, the number of bus passengers has been declining for years. In other words, each bus is carrying fewer people. The increasing subsidies thrown at transit systems mean that they burn ever more fuel to carry ever-more empty seats. In 2018, passenger cars and light trucks in US cities used up to 3,400 British thermal units (an energy measure) per passenger mile traveled, while transit buses used over 4,500. Although rail energy costs are lower than both cars and buses, that is largely a result of the New York City Subway, which itself transports the vast majority of US rail passengers. Yet, even before the pandemic this system was in trouble due to falling ridership, flooding, and an assortment of ills. New rail projects have even more empty seats than buses.
Throwing more subsidies at new buses or trains that travel routes humans don’t want to take will only lead to more empty seats and more burned energy, just as cars are getting more efficient and greener. The transit catastrophe of the pandemic means that the environmental costs of moving passengers by bus or rail has become even greater. By all indications, the future of the city, and the future of the environment, will be based on energy-efficient cars and increasingly distant homes.
The Future Is Spread Out
Just like density, sprawl has costs as well as benefits. For instance, sprawl can result in the loss of species’ habitats and natural landscapes. But these problems can be accommodated. The fact that only 2 percent of the American landmass is urbanized, and that not even the most sprawling projections of the future would imagine that figure going over 5 percent, means Americans can protect species and environmentally sensitive areas as we expand. We can, as McHarg noted, design with nature.
Sprawl isn’t for everybody, and just as we shouldn’t force everybody to live in dense metropolises, we shouldn’t force anybody to live in sprawling ones either. For many people, Manhattan or its equivalents will be the best and most exciting option, and American cities need to accommodate those preferences. Good policy and increasingly green technology—especially when it comes to building materials and transportation—can ameliorate the demons of density in those areas. But it cannot eliminate them.
The future of the American city will not be a growing number of Manhattans. It will be more Dallases and Atlantas and Nashvilles and Columbuses. These are the types of cities that most Americans have moved to in recent years, and all evidence is that, after the pandemic, they will grow even more rapidly.
These cities already represent a future that is more environmentally sound and economically affordable than the dense metropolises of the past. But we can keep working to improve them, by accelerating the move to electric vehicles, by improving energy efficiency in homes, and by changing our energy mix. Scientists, engineers, and entrepreneurs have been engaged in this task for decades and will continue doing so. Instead of destroying the sprawling city, they are improving it.
Increasingly irrelevant attempts by environmentalists to fight these sprawling cities and the cars that allow them to exist are counterproductive. Instead, environmentalists should embrace the same future that most Americans have already chosen.