The Price of a Parking Space

In 1954, the city of Peoria, Illinois adopted a zoning code that required every new restaurant to provide one parking space for every 100 square feet of dining area. The rule was not the product of careful transportation research. It was not derived from any study of how many cars actually arrived at Peoria restaurants or might plausibly arrive in the future. It was a number that seemed reasonable to the planners who wrote it down, informed primarily by the assumption that America in 1954 was a place where everyone drove everywhere, which was both a description of the present and a bet on the future.

Versions of that rule, adjusted upward and downward and subdivided by land use category, subsequently spread to nearly every American municipality that engaged in any zoning activity. By 1990, according to Donald Shoup’s exhaustive 2005 study The High Cost of Free Parking, minimum parking requirements covered roughly 90 percent of American urban and suburban land outside of a handful of older, pre-automobile city cores. The requirements were baked into the DNA of American development so thoroughly that most planners, developers, and municipal officials treated them as natural features of the built environment rather than policy choices that had been made by specific people at specific times for reasons that have since become archaeologically obscure.

The cost of those requirements is staggering and almost completely invisible in the conversations people have about why American cities are expensive, sprawling, and hostile to pedestrians. Shoup estimated in 2005 that the subsidy represented by required parking — the cost of construction, land, and maintenance that developers are forced to provide and then either absorb or pass on — amounted to somewhere between $127 billion and $374 billion per year in the United States alone. Those numbers are from 2005 dollars in a study that itself acknowledged the figures were conservative. The actual cost is probably higher in both absolute and relative terms today, after seventeen additional years of suburban sprawl development built around mandatory parking provision.

The mechanism is not subtle once you see it. A minimum parking requirement is a forced subsidy of car ownership by everyone who participates in the built environment, including people who do not own cars. When a developer is required to build 200 parking spaces to get approval for a 150-unit apartment building, those spaces cost roughly $30,000 to $60,000 each in a surface lot configuration and $40,000 to $80,000 each in a structured parking garage. The developer does not absorb this cost; they cannot afford to. The cost is incorporated into the rent charged to tenants, regardless of whether those tenants own cars, need parking, or have any interest whatsoever in subsidizing their neighbors’ vehicle storage.

The same logic applies to retail. The grocery store required to provide five parking spaces per 1,000 square feet passes that cost into the prices of everything it sells. The restaurant required to provide parking builds a surface lot that consumes land that might otherwise be a building or a park, spreads the development out to a scale that makes walking between destinations difficult, and charges everyone who eats there for the car storage whether or not they drove.

The walkability problem is at least as significant as the cost problem, and the two are deeply intertwined. Minimum parking requirements do not merely add cost to development; they change the shape of development in ways that make alternatives to driving structurally impossible. A strip mall with a 200-foot parking lot between the sidewalk and the storefronts is not merely inconvenient to walk to. It is built in a configuration that makes walking the rational choice only in a narrow range of circumstances (you happen to be going exactly there, from exactly here, with nothing to carry). When every destination in a city is surrounded by required parking, the cumulative effect is an urban form that can only be navigated by car, which validates the original assumption that everyone drives, which reinforces the case for requiring parking, in a feedback loop that took about forty years to close completely.

The geographer Peter Calthorpe documented this dynamic in his 1993 book The Next American Metropolis, pointing out that the minimum parking requirement was essentially a self-fulfilling prophecy: it created the conditions that made cars necessary, then cited the necessity of cars as justification for its own continuation. The cities that most aggressively required parking are now the cities most thoroughly dependent on cars, which makes rolling back the requirements politically and practically difficult because the alternative infrastructure (transit, bike lanes, walkable density) that would make lower parking provision viable has been systematically prevented from developing by the parking requirements themselves.

This is what economists call a coordination failure: everyone would be better off in a city designed differently, but no individual actor can capture the gains from building differently in a city designed this way. The developer who builds without the parking lot gets no tenants because the tenants have cars and need storage. The transit agency that wants to run more frequent service cannot justify the frequency because the density that would support ridership cannot be built because the parking requirements prevent the density.

A few American cities have started to recognize this and act on it. Minneapolis eliminated single-family-only zoning in 2018 and has been incrementally removing parking minimums in certain zones since then. Buffalo, New York became the first major American city to eliminate all minimum parking requirements citywide in 2017, motivated primarily by the observation that the requirements were suppressing redevelopment in a city that desperately needed it. Oregon passed state legislation in 2021 eliminating parking requirements near transit stops for cities above a certain population threshold. California went further in 2022, passing SB 6 and other legislation that essentially prohibited cities from requiring parking for certain housing types near transit.

These are real changes, and the early evidence is encouraging. Studies of Buffalo’s reform found that development activity in affected areas increased substantially and that the feared parking apocalypse — the nightmare scenario where new residents circle endlessly looking for spaces, destroying neighborhood character and making driving impossible — did not materialize. People adjust. The demand for parking in dense urban environments, it turns out, is not fixed; it responds to price and proximity and the quality of alternatives, just like demand for everything else.

What the reforms have not yet produced is a wave of genuinely walkable urban development, because removing the parking requirement is necessary but not sufficient. You also need to allow the height and density that makes walkable development economically viable, which runs into a different set of zoning constraints. You need transit that is frequent and reliable enough to make car ownership genuinely optional, which requires capital investment and political will that most American cities have not sustained. You need a built environment redesigned for pedestrians and cyclists, which takes decades even if you start immediately.

The international comparison is instructive and humbling. Japanese cities, which operate under a completely different regulatory framework, allow development of extraordinary density and mix without minimum parking requirements in most urban zones. The result is cities like Tokyo and Osaka that are simultaneously among the most expensive real estate markets in the world and meaningfully more affordable than comparable American cities, largely because the regulatory constraints on supply are substantially lower. The transit that Japanese cities built starting in the 1950s was the product of policy choice, not geography or culture. The walkability of central Tokyo is the product of zoning that permitted dense mixed use, not of any inherent Japanese preference for walking.

The Netherlands, which spent most of the postwar period building car-dependent suburbs on the American model, made a deliberate policy pivot in the 1970s and 1980s toward cycling infrastructure and transit and has spent forty years rebuilding the urban form that the car had displaced. Amsterdam and Utrecht are genuinely bikeable cities today. They were not bikeable cities in 1970. The change required sustained political will and continued investment across multiple electoral cycles, but it happened, and the result is cities that score consistently high on quality-of-life measures and low on per-capita transportation cost.

The lesson of minimum parking requirements is not that car ownership is bad or that parking is inherently wrong. It is that when you make something mandatory, you subsidize it in ways that distort every decision made downstream. American cities mandated parking for seventy years and got cities built around cars, with all the costs that entails in land consumption, housing affordability, transportation spending, and the quality of daily life for people who cannot or do not want to drive.

Unwinding that is going to take longer than most reformers would like to admit. The parking is already built. The sprawl is already there. The transit networks have not been built in most places. The political economy of any given parking reform effort runs up against the organized interests of people who currently have free or subsidized parking and experience the proposal to price it or reduce it as an attack on something they have come to regard as a right.

Shoup spent his career at UCLA arguing that the solution is simple: charge market prices for parking, eliminate the minimums, and let the market sort out how much parking actually gets built. The solution is simple in the sense that its logic is airtight. It is not simple in the sense that it requires a political majority to agree that something that currently feels free should cost what it actually costs, which is a very difficult sell in any democratic context.

The Peoria restaurant from 1954 probably still has a parking lot, if it is still a restaurant. The lot was required when the building was built, and now it is just there, occupying land that in a denser city would be something else, adding to the per-mile driving necessity of everyone in the neighborhood, costing something that nobody directly pays.

That is how seventy years of parking policy has worked. The costs are real and they are paid. They are just paid in forms that are hard to see and impossible to trace to the decision that caused them.