Cities, Cities, Cities Housing, Infrastructure & Transportation
January 20th, 2022 16 Minute Read Issue Brief by Eric Kober

How Large Cities Can Grow Denser and Flourish: What the 2020 Census Reveals About Urban Sprawl


The latest decennial census revealed that most big cities become populous by spreading over a large land area. Only 12 of the 50 largest U.S. cities in 2020 had population densities as high as 7,500 per square mile—the U.S. big-city average in 1950, the last census before suburbanization, urban renewal, and the construction of the interstate highway system permanently redistributed the U.S. population away from dense, transit-rich urban areas.

This paper looks at those 12 cities, most of which have had significant population increases in the past two decades. Dense cities are important to the nation as focal points of economic and cultural innovation, as well as areas where residents voluntarily adopt housing prototypes and lifestyle patterns that have lower-than-average carbon emissions. As these cities grew, they raised U.S. labor productivity, slowed the U.S. contribution to climate change, and provided Americans with increased choices for how and where to live.

Overall, relatively dense U.S. cities have followed two distinct paths to growth. One is “grand bargain” planning: municipalities faced with the need for more housing but widespread opposition from neighborhood activists direct new housing (and thus population growth) to a small number of neighborhoods, in and near downtown, that do not have sizable preexisting populations that can object. The areas that can be redeveloped in these neighborhoods may include open parking lots, now-underutilized but once-industrial sites, public property no longer needed for its original purpose, or areas cleared of residents under long-ago urban renewal plans.

While pragmatic, “grand bargain” planning to achieve population growth suffers from several flaws. It fosters dependency on the costliest high-rise housing prototypes, requiring that most new units be targeted to high-income households. Most other households compete for the inadequate stock of older units, as well as the relatively few subsidized income-restricted new units that the city manages to construct. Furthermore, grand bargains lead to similarity of development, as new housing is built at much the same time, for much the same population, in a few locations. Such neighborhoods may adapt poorly to changing populations and lifestyles over time.

An alternative, more difficult path to growth tries to disperse new housing over a broad area of the city. Such “distributed growth” plans cause changes where sizable populations already live, potentially creating controversy over land-use decisions. Some strategies to disperse growth include Seattle’s “urban villages,” still leading to concentrated growth but affecting more parts of the city; Boston’s construction of new housing affordable to middle-income households, without subsidies, near dispersed transit stations; and Minneapolis’s designation of commercial corridors for growth, as well as permitting up to three units on all single-family lots. Dispersed growth is perhaps more easily accomplished through a comprehensive planning process to secure public buy-in, with upgraded transit and attention to the quality of other neighborhood infrastructure and amenities. The rewards for cities that succeed in this endeavor include more gradual and varied neighborhood change, new housing affordable at a wider range of incomes without subsidies, and greater perceived equity in the distribution of the burdens of population growth.

America’s Densest Cities in 2020

Figure 1 lists the 12 American cities among the top 50 in 2020 population [1] that had a population density of 7,500 persons or more per square mile [2] in 2020. Nearly all of them gained population and housing units over the 2000–10 and 2010–20 decades. Chicago, which lost population from 2000 to 2010 and remained below its 2000 population despite growth in the 2010–20 decade, is the exception. However, Chicago gained housing units in both decades.

Even within this unusual group of American cities, Seattle, Washington, D.C., and Miami stand out. The populations of these three cities grew over two decades by more than 20% (in Seattle’s case, more than 30%), and the number of housing units increased by 36.1%, 27.5%, and 43%, respectively. Four other cities gained more than 10% in population over the two decades—San Francisco, Boston, Oakland, and Minneapolis. New York City’s population grew 9.9%, but its numerical gain, about 796,000, was larger than the entire population of all but the 17 largest American cities.

Figure 1

America’s Densest Cities (in the Top 50 by Population) in 2020

Why Population Density Is Important

High-density cities represent an urban planning ideal that many less dense cities are trying to emulate. [3] Dense cities are at the center of many of the nation’s high-productivity metropolitan areas (particularly those not associated with the capital-intensive energy industry). [4] Urban density is associated with facilitating the exchange of information and innovation, particularly in industries requiring a high degree of skill and education. [5]

This chain of virtuous linkages—where urban density is associated with higher economic productivity, highly educated workers, and the nation’s most exciting and innovative industries— goes a long way toward explaining why many cities want the kinds of dense, mixed-use urban neighborhoods associated with social immersion, spontaneous interaction, and economic and cultural innovation. Such neighborhoods are thought to attract the highly educated professionals who power high-productivity economies. [6]

A second argument for the importance of America’s dense cities is that they point the way to a future of environmental sustainability. [7] Residents of dense cities drive shorter distances, use public transit more, and tend to have smaller housing units that are located in more energyefficient attached or multifamily buildings. The possibility of achieving significant reductions in carbon-dioxide emissions by constructing housing in dense urban locations, where at least a portion of the population wants to live, is attractive to city planners. [8] Unlike other emissionsreduction strategies that call for new subsidies, taxes, or regulations, urban densification can occur at low cost, particularly where cities have legacy transit and wastewater infrastructure with excess capacity, or new urban in-migrants come from the higher-income professional groups. Such households can afford newly constructed private-market housing and pay taxes to support their infrastructure and service needs, and still leave cities better off financially. [9]

Grand Bargain vs. Distributed Growth Planning

Each of America’s densest cities has unique qualities, but every city faces a similar dilemma in confronting growth pressures. All cities try to attract businesses and add jobs, and if they do that successfully, the labor force will also grow. New workers bring spouses and children, as well as more workers to provide them all with services. Thus does a population grow, and that growth runs up against a housing supply that is inelastic in the short term. High market rents and housing prices, well above the replacement cost of existing housing, indicate that the demand for housing outstrips supply in many neighborhoods. But the current residents of these neighborhoods will often fight any attempt to add to the housing supply by lifting zoning and other restrictions and allowing private developers to build.

City governments seem to respond to these conditions in one of two ways. [10] One is what the Canadian city planner Gordon Price calls the “Grand Bargain.” [11] In this bargain, new housing is concentrated at high densities in areas near downtown, on former industrial sites, and any other available land where there are few preexisting residents. Low-density residential neighborhoods are spared growth, preserving high home values and their desirable “neighborhood character.”

I have previously written about how New York City in the Michael Bloomberg mayoral administration pursued a version of “grand bargain” planning, negotiating with the city council for a small number of highly successful rezonings, mostly in nonresidential areas near the city’s core, while making the zoning more restrictive over broad swaths of the city’s residential neighborhoods. [12] The bargain powered housing growth in the 2010–20 decade, as the city’s population increased by 629,000. [13] The city’s population growth was concentrated in a handful of its 59 community districts, while many other districts with strong housing markets were protected from growth by restrictive zoning and widely mapped historic districts.

Washington, D.C., with its rapidly growing Sixth Ward forming a crescent north, east, and southwest of the Capitol, and its static, affluent, expensive Northwest, is another good example of grand bargain planning. San Francisco is another, with its growth concentrated in a handful of neighborhoods along the shoreline downtown and south of the Bay Bridge. Established residential neighborhoods in other parts of the city, in contrast, saw little growth. Miami, with high-rise development in a strip between I-95 and the shoreline, is yet another example.

This kind of planning has much to recommend: the city gets growth and benefits economically. Planners and mayors are pragmatic, and they focus on what is possible. However, there are downsides to grand bargains. One is that new housing is focused on the most expensive type of housing to build—large, often high-rise, apartment buildings. The housing types that are easier to finance, less costly to build, and potentially offer lower market rents and sales prices—such as small walkups, garden apartments, and infill apartment buildings—have limited opportunities to locate within the city. These housing types, sometimes labeled as “missing middle” housing,[14] were widely found in American cities before 1950, but as zoning became more specific and restrictive, they could no longer be built in many neighborhoods. Allowing such lower-cost housing once again expands the income range of households that can rent or buy new housing without public subsidies. Minneapolis’s legalization of up to three units per lot is an example of a city trying to revive this missing middle housing category.

Another critique is that “grand bargains” encourage neighborhoods built up at the same time and based on the pursuit of a similar market segment—in the last decade, often young, affluent urban professionals. Such uniform neighborhoods may not adapt well to social and demographic changes over time. Washington urbanist Payton Chung cites Jane Jacobs’s celebration, in The Death and Life of Great American Cities, of the incremental change that takes place in urban neighborhoods, permitting diversity and adaptation to new conditions. [15] Jacobs was critiquing the uniformity of urban renewal. But privately initiated large-scale real-estate developments, influenced by zoning constraints and design guidelines imposed by the local planning commission and city council, can have the same flaws.

A third problem with grand bargain planning is that growth makes continued adherence to the grand bargain impossible. As the politically easier sites are used, the neighborhoods once exempted from growth must share in the burden of housing a growing population. Cities do not have an unlimited supply of no-longer-needed industrial sites or disused public facilities to redevelop with housing. Yet they will continue to grow. Even if a city succeeds in building all the housing that the comprehensive plan projects will be needed in the foreseeable future, another future looms, in which yet more housing needs to be built, and some other category of sites needs to be identified.

Moreover, political pressures will mount to enact bad housing policies that ostensibly address affordability issues. Most of the cities discussed in this paper have high levels of “rent burden,” the term that applies to households that pay more than 30% of their income for housing (Figure 2). The Minneapolis rent-control referendum, authorizing the city to enact such a law, is one such example of a potentially damaging policy. Voters in the neighboring city of St. Paul concurrently approved the nation’s strictest rent-control ordinance. [16] While the details of any rent-control law matter, rent controls are fundamentally intended to keep existing tenants in place, exacerbating shortages and placing an even greater burden on new construction to accommodate demand.

Figure 2

Rent-Burdened Households in the U.S. and Selected U.S. Cities, 2019









California cities are also pressured by the Regional Housing Needs Allocation (RHNA) / Housing Element process mandated by state law, which requires cities to plan for a specified housing construction goal. [17] All these factors create an impetus for cities to move away from grand bargains.

The alternative to “grand bargain” planning is distributed growth, which is much more difficult to achieve politically. By spreading growth to multiple neighborhoods throughout the city with a range of urban densities, this approach can be fairer in distributing the burdens of growth and better able to respond to changing conditions as the redevelopment of any one neighborhood is spread over a longer period. Seattle, for example, has pursued an “urban village” strategy in its comprehensive plan that has resulted in significant multifamily housing growth in far-flung neighborhoods. [18] The city’s latest tracking report indicates that from the beginning of 2016 to June 30, 2021, housing units increased by 28.9% in Urban Centers—the densest areas, including neighborhoods near downtown. Housing units also increased by 21.7% in Hub Urban Villages and 20.1% in Residential Urban Villages. [19] However, the city is already closing in on its 2035 housing goals for these areas and will need to expand their boundaries or select new growth locations Boston’s comprehensive housing plan takes a different, but relatively effective, approach to distributing growth. By setting a goal for market-generated middle-income housing, [20] the plan induces the government to issue construction permits in areas farther from downtown, where land costs are lower and lower-cost housing prototypes are appropriate. Boston’s latest progress report, for 2020, indicates that from 2011 through 2020, the city issued permits for 7,584 market-rate middle-income units, compared with 35,955 units permitted overall. [21] The Minneapolis 2040 plan’s focus on transit corridors and higher densities, even in single-family areas, is another way of achieving distributed growth. [22]

Getting from grand bargains to distributed growth requires determined and enlightened leadership and a credible public consultation so that the final result is seen to be fair. In many cities, this takes the form of a citywide comprehensive plan that considers the typical concerns that are raised by citizens and elected officials in the context of proposed housing growth. One such concern is that transit services be commensurate with the added population burdens. Older cities such as Boston have legacy rail transit systems that can serve a growing population, particularly with upgraded signaling and modern rolling stock. Boston’s comprehensive transportation plan, Go Boston 2030, includes a list of proposed improvements, particularly to speed up bus service. [23] Cities without a legacy rail transit system, like Seattle, have built, and are expanding, light rail. [24] The Minneapolis 2040 plan also calls for bus upgrades, including priority bus lanes and signal timing priority. [25]

Another public concern that city officials typically hear is a commitment to achieve a share of new housing as affordable. City housing plans often establish affordable housing goals. In New York City, the only housing goal that local government tracks is for affordable housing [26]—which results in distorted priorities as enormous public resources are funneled to meet income-level targets without any consideration of overall population growth. Boston, by contrast, tracks affordable housing as a subcategory of all housing permits. Seattle tracks overall housing construction and separately produces an annual report on affordable housing production. [27]

Affordable housing is an added expense that needs, in most cases, to be borne to a large extent by the public sector so that private developers can still achieve a threshold rate of return. Since public resources are limited, many cities resort to zoning-based inclusionary housing programs, in which developments are incentivized or required to include affordable units. A poorly designed mandatory program, like that of New York City, acts as a brake on housing production. [28] New York’s policy applies to any development of more than 10 units seeking a zoning change. Boston’s Inclusionary Development Program (IDP), which applies to any development of more than 10 units seeking a variance—a common occurrence in that city for new multifamily housing—is more flexible than New York’s. It requires a smaller percentage of the development to be affordable, establishes generally higher income limits for those units, and allows offsite construction and cash-out payments in lieu of providing affordable units within new buildings where other units are market-rate. [29] Boston’s planning agency explicitly states that “the City sets IDP requirements at a level that will assure the creation of income restricted units, without discouraging developers from building the market rate housing that Boston also needs.” [30]

In recent years, Seattle has implemented a Mandatory Housing Affordability program in areas rezoned to permit more development capacity. [31] While this program is too recent to have affected the 2020 census results, the city reports that it was widely used in 2020, the first full year of operation. Like Boston’s, but unlike New York’s, Seattle’s program has a relatively low requirement for the percentage of new units that must be affordable in an otherwise market-rate development, although income-level affordability requirements are more stringent. Seattle also has an allowance for a cash contribution in lieu of providing on-site affordable units.

A plethora of additional demands—for parks, schools, and other infrastructure—also need to be addressed by a city’s planning process. A thorough analysis is needed of zoning controls that potentially thwart theoretically permitted higher-density development—parking requirements, floor area controls, height and setback controls, unit-size controls, and so forth. Minneapolis, for example, eliminated minimum parking requirements citywide in May 2021.[32] San Francisco did so in 2019. [33] New York City, by contrast, continues to require off-street parking for residences and ground-floor retail in many areas of the city well served by public transit. [34]

Procedural impediments, including environmental reviews and discretionary reviews by planning boards and city councils, need to be replaced with predictable standards. In Boston, for example, the Zoning Board of Appeals variance process has become a common mechanism for approving new housing. While this may be practical and effective, the lack of clear rules about future development makes this procedure controversial. [35]

For cities that achieve their housing goals, there are many rewards, both local and societal. Distributed growth is good for choice—new, denser housing in cities is marketable because some people want that lifestyle. It’s good for economic and cultural innovation as well. And not least, because people who live in dense cities drive less, walk, and use transit more, and need less energy to heat and cool their more compact homes, it’s good for mitigating the effects of climate change. Undertaken on a broad enough basis, without instituting other regulations that undercut its effectiveness, distributed growth offers the only hope of mitigating America’s urban affordability problems. As more cities move toward accepting distributed growth, the ranks of America’s dense cities may be less thin in 2030.


PDF icon See Appendix.


See Endnotes in PDF.

About the Author

Eric Kober is a senior fellow at the Manhattan Institute. He retired in 2017 as director of housing, economic and infrastructure planning at the New York City Department of City Planning. He was Visiting Scholar at NYU Wagner School of Public Service and senior research scholar at the Rudin Center for Transportation Policy and Management at the Wagner School from January 2018 through August 2019.  He has master's degrees in business administration from the Stern School of Business at NYU and in public and international affairs from the School of Public and International Affairs at Princeton University.


Are you interested in supporting the Manhattan Institute’s public-interest research and journalism? As a 501(c)(3) nonprofit, donations in support of MI and its scholars’ work are fully tax-deductible as provided by law (EIN #13-2912529).