Everyone who follows debates about urban planning already knows that sprawling cities build more housing and have lower housing costs. Yet last week Issi Romem, an economic analyst at BuildZoom, a company that helps people find and hire contractors, published an analysis of this phenomenon that sent urbanists reeling. It should not have done so. Romem’s data was not new and his analysis was flawed and misleading.

Romem also shows that cities have not been increasing in population density as much as the past. Again, this isn’t new. If anything, Romem understates what has happened: the largest cities in the United States in 1950 all began to experience population decline due to suburbanization in that decade and only New York City has exceeded its previous peak, while expensive, growing coastal cities like Boston and Washington, DC are still far off the densities and populations they had in the past.

While Romem’s data is indisputable, it doesn’t tell the whole story. Sprawl isn’t really as cheap as it seems. A network of tax breaks, financial guarantees, subsidies, and other chicanery keep parts of suburbia relatively inexpensive. Most notably, transportation costs are often excluded from the discussion of housing affordability, even though it’s hard to live anywhere without a way to get to work. For example, Chuck Marohn at Strong Towns has shown that the low density, car-dependent development that has typified American cities since World War II does not produce enough tax revenue to service the debt that cities took out to build the infrastructure needed for sprawl.

Romem compares the San Francisco and Atlanta metropolitan areas on affordability and the sprawling Peach City naturally beats out the more compact City by the Bay on housing costs, but if one includes transportation costs as a percentage of income, then Atlanta becomes less of a bargain, according to the Center for Neighborhood Technology’s H+T Index. A typical household in San Francisco, according to the measure, spends 50 percent of its income on housing and transportation, but a typical Atlanta household spends 54 percent of its income on the same costs.

This leads to what Romem calls the “land-use trilemma,” which presents the perceived trade-offs between more sprawl, doing nothing while letting expensive cities get more expensive, or liberalizing land use laws to allow more density. Much like C.S. Lewis’s trilemma, which was useful for Christian apologetics (but is usually ignored by more serious theologians and Biblical scholars), Romem’s trilemma is useful for the apologists of sprawl but falls apart upon examining his assumptions.

For instance, tucked away in a footnote, Romem writes that “Shifting from single family to multifamily housing involves a sacrifice in terms of living standards. The current wave of interest in micro-units takes the sacrifice of living standards to an extreme.” This is a load of tosh: last year a CBS affiliate in San Francisco did a story on a house in the Mission District where the rent was $1,800 a month (far less than the area’s average rent), but tenants were in bunk beds eight to a room and the house ultimately had 30 people in it. When compared with this living situation, even a 140 square foot micro-apartment one has to one’s self is going to be a major step up. Nor does an increase in density imply a shift from single family homes to large apartment buildings: the attached houses of the English Midlands are single family and have yards. Tokyo also achieves very high densities with detached single family houses, thanks to small sizes, small lots, and narrow streets.

A similar view of development underlines and undermines the trilemma. For example, Romem implies that development is a major intervention that happens on a neighborhood scale. He told Bloomberg‘s Patrick Clark, “No one is really thinking about tearing down single-family neighborhoods and putting up apartment buildings.”

But no urbanists really think that kind of demolition is necessary. There’s a wide spectrum between the Levittowns, archetypes of postwar sprawl, and Hong Kong’s former Kowloon Walled City, once the densest neighborhood on the planet. Density can be added, as in the albeit extreme Mission bunkhouse example, by turning a single family house into a multifamily home. Apartments can be carved out in basements, second stories, and above garages. The principles of the New Urbanism suggest that such incremental development is healthier for cities, since property remains in the hands of smallholders instead of assembled into giant lots.

Some cities are improving the quality of life for seniors and housing young people with homeshares. Seniors in many cities face increased difficulties in staying in their homes because they have more house than they can take care of and are hit harder by increases in property taxes. These programs match young people with seniors and they pay a small rent while taking care of the house and keeping the older person engaged and active.

New Urbanism also presents an alternative to the trilemma by supporting the regeneration of small towns affected by deindustrialization. Many cities have older, traditional small towns with existing but dilapidated multifamily housing stock or downtown commercial or industrial blocks capable of being attractively renovated. New England in particular is home to many of them—towns like Southbridge, Mass and Woonsocket, R.I., or even Bridgeport, Conn. New Urbanist designs can also be applied to new development in expansive cities. Some big cities, especially outside the northeast and west coast, are surrounded by unincorporated land not subject to municipal zoning laws. A developer could build a denser, more urban neighborhood in these areas—which is exactly what’s happening in Toronto’s suburbs, according to Stephen J. Smith. The financing might be difficult, thanks to federal rules that, according to the Regional Plan Association, discourage the construction of small mixed-use buildings by capping how much commercial space they can have and promoting larger buildings.

Just as importantly, the land-use trilemma falls apart because realistically there is no alternative to allowing greater density. There are hard limits to the development pattern of American suburbia. The most discussed is commute time across metropolitan areas. According to Slate, longer commutes are associated with divorce, isolation, obesity, stress, neck and back pain, sleeplessness, and unhappiness. The conventional wisdom for dealing with long commutes has been to build bigger and faster streets and highways, but in the long term, this does not work. Several years ago Texas spent $2.8 billion to widen the congested Katy Freeway connecting Houston to its suburbs. At 23 lanes it is the widest highway in the world and Houston commute times have still increased, according to City Observatory. Metropolitan areas need not follow the standard pattern of a dense core—not that most American cities are dense by global standards—and dispersed suburbs, but could become more decentralized with pockets of higher density at certain points throughout a region, along the lines of Joel Garreau’s neglected “edge city” concept.

American cities need not face an ugly choice or a hard choice, but their leaders and citizens do need to make decisions based on the realities of market demand, high levels of public and personal indebtedness, and climate change.

Matthew M. Robare is a freelance journalist based in Boston who writes about urbanism and history. This article was supported by a grant from the Richard H. Driehaus Foundation.