What’s wrong with Euclidean Zoning?
Examining how zoning and agglomeration effects may hinder growth in the United States
Zoning, the regulation of how different spaces can – or, more importantly, cannot – be used, is becoming an increasingly politically salient issue. Zoning is so politically relevant because it is the main cause of the current affordable housing crisis: currently, half of Americans spend more than 30% of their income on rent, with average incomes having grown half as quickly as rental prices since 1999. By artificially restricting the supply of homes, zoning causes demand for housing to outweigh the available supply, driving up prices and increasing the difficulty for individuals and families to afford suitable housing. Zoning strictness is therefore both highly positively correlated and even causally linked with high house prices, with places facing the most severe housing shortages, like San Francisco and New York City (NYC), having the most restrictions on building new houses.
Yet, coverage needs to reference another, potentially equally important, consequence of zoning: lower economic growth. Chicago economists Chang-Tai Hsieh and Enrico Moretti estimate that inefficient spatial allocation due to zoning has cost the US 0.3% of annual GDP growth from 1964 to 2009. Even more strikingly, they estimate that if zoning restrictions in the most heavily zoned US cities, like NYC and San Francisco, were as regulated as the median US city, total US output would increase by 9.5%. Assuming standard 2% GDP growth, this increase would equal five years’ average growth. Reforming zoning could be a key under-discussed intervention to combat the great stagnation and flatten productivity growth. This is beginning to happen, with politicians and policymakers across the ideological spectrum calling for intervention. For example, last year, the White House announced the $85 million Pathways to Removing Obstacles to Housing Program, aiming to weaken legal barriers to producing new homes. However, reform is justified by the housing crisis, not growth.
Such estimates raise several questions: how did such a suboptimal equilibrium, with such high house prices and little possibility of change, come to be? Why is the spatial allocation of labour closely tied to productivity and economic growth? Further, why has zoning regulation yet to be significantly reformed to allow greater housing supply despite such large potential returns? To analyse this link between zoning and growth, I first explain the origins and effects of the current ‘Euclidean zoning’ regime, and how it has impeded urban density. I then link this lost density to GDP growth through the ‘agglomeration economies’ mechanism.
Why was Euclidean zoning introduced?
The modern approach to regulating land use commenced with the US Supreme Court’s 1926 Euclid v. Amble decision; before the ruling, Americans had near complete discretion over what they built, as land was relatively abundant. Such freedoms applied even to America’s largest cities. For example, there were no regulations on how high buildings could be in NYC before 1916.
Crowded tenement buildings in East Harlem, NYC. (Source: Skyscraper.org)
Yet, large-scale immigration from Europe, which primarily settled in urban areas, in the 1890s led two-thirds of New York’s population to live in packed tenement buildings by 1900; as many as 30 people could occupy one room, normally without windows, sewage or running water. Simultaneously, East Coast cities began to build upwards, as innovative deep foundations and iron-framed structures combined with Gilded Age capital to enable the first skyscrapers, depriving surrounding structures of sunlight.
Despite this, contemporary housing supply was still unable to accommodate such a population increase, creating urban density unprecedented in America, and the many problems that came with it; few restrictions on building density and materials heightened fire risks, culminating in the 1871 Great Chicago Fire, which left a third of the city's population homeless. This increased density also led to terrible public health outcomes, with the most dense area of Manhattan, the Lower East Side, having a 20% lower disease mortality relative to the New York City mean. These impacts of increased density on fire risks, public health and general quality of life were all exacerbated by limited regulation of industrial facility location, as apartments were often next to factories. The resultant air pollution was egregious, with Pittsburgh being described as ‘hell with the lid taken off’ due to the smog.
This context created a progressive housing reform movement, which attempted to resolve these spillover effects associated with urban density, culminating in Euclidean zoning.
Thomas Moran, Communipaw, New Jersey, 1884. (Source: Gilcrease Museum)
What is Euclidean zoning?
‘Euclidean zoning’ is named after the city in Ohio, not the Ancient Greek geometrist. The 1926 US Supreme Court ruling upheld the legitimacy of government regulation of land use, leading to the creation of Euclidean zoning (named after the city in Ohio, not the Ancient Greek geometrist), a legislative framework determining what can be built where. The ruling stated that land use regulation should aim to reduce fire risks, noise, and create a ‘favourable environment in which to rear children with …. free circulation of air and no monopolising of the rays of the sun’. The creation of Euclidean zoning attempts to resolve the many aforementioned social maladies of high-density cities.
To achieve these goals, Euclidean zoning uses the following zoning structure: each area of a city is separated into districts, determined by the local government. Each district is allocated a zoning code, which determines what may be built there. For example, currently, in Washington DC, there are 15 different types of residential zones, falling into four zoning categories: R-1A, R-1B, R-2, and R-3:
R-1A and R-1B zones are the least dense, allowing only detached dwellings, and aiming to ‘promote a suitable environment for family life’.
R-2 zones are denser, allowing both detached and semi-detached homes, but still aim to ‘protect these areas from invasion by denser types of (housing)’.
R-3 zones allow for terrace housing and are intended for young professionals rather than families.
Yet, all R (residential) zones only permit single-family dwellings, which prevents the possibility of high-density neighbourhoods. For example, none of the residential zones in DC allow apartments, as they ‘intend to discourage multi-household development’. Euclidean zoning therefore promotes low-density neighbourhoods by explicitly encouraging single-family, detached or semi-detached residential housing, the lowest density forms of housing.
The low density created by Euclidean Zoning is further evident when comparing American zoning to, for example, German zoning law. In Germany, no distinction is made between single and multi-family housing, using a ‘non-hierarchical’ zoning model. This has led to much higher urban density, but without compromising on problems Euclidean zoning aimed to solve, like air or noise pollution from factories; instead, these are targeted via direct regulation.
Is Euclidean Zoning Justified?
Euclidean zoning was an understandable reaction to the public health and quality of life challenges of early 20th-century urban living, aiming to prevent exploitative and dangerous tenements and protect homeowners choosing to live in quiet areas from industrial expansion. However, Euclidean zoning is no longer suitable for this purpose. The number of Americans working in manufacturing has decreased from 32% to 8% since 1900, while regulatory agencies and legislation like the EPA and Clean Air Act have been established.
Zoning therefore no longer needs to be the primary method of reducing pollution. Similarly, the public health issues associated with urban density, at least in America, are mostly resolved; rural individuals now have worse physical and mental health outcomes than those in urban areas and the countries with the highest life expectancy (Monaco, Macau and Singapore) are all city-states.
There are still reasons to have zoning: there is value in preserving the character of a neighbourhood, whether for historical reasons or to ensure there are houses for those that prefer less dense areas. It can also be an important tool for mitigating negative externalities, such as the noise a car factory would create if it were built next to your house.
Yet, these current justifications are much less consequential than Euclidean zoning’s historical impetuses. Furthermore, zoning’s promotion of low-density neighbourhoods has had significant negative consequences: most prominently, it has caused housing density to be artificially low. Euclidean zoning regulations designate many US cities for single-family, semi-detached housing. Even if the population of a city increases, it is impossible to respond to the increased demand for housing, as developers are legally barred from building multi-family, or even terraced housing, let alone apartment complexes. Harvard economics professor Edward Glaeser and colleagues corroborate this, arguing that ‘man-made scarcity of housing in the sense that the housing supply has been constrained by government regulation as opposed to fundamental geographic limitations’.
The adverse effects of the resultant artificially low housing supply on homeownership are well documented and justifiably drawn attention to: according to the Urban Institute, the ‘real homeownership rate’, the rate when changes in household composition like marriage and divorce are controlled for, has decreased by 12% when compared to 1990. The Federal Statistical System estimates the ‘real homeownership rate’ to be at its lowest point since 1965.
Furthermore, the cost of zoning regulation has been estimated to be 40% of home values. Glaeser estimates this cost by separating the value of a house into three elements: construction costs, land value and the value of the right to build on that land. The third component, ‘regulatory tax’, is not a tax paid to the government and instead refers to the additional costs resulting from zoning rules reducing supply, thereby increasing price. The regulatory tax is estimated to compose more than 50% of house values in San Francisco and Manhattan, indicating that absent the restrictions on housing supply due to Euclidean zoning, house prices in these areas would halve. Restrictive zoning is therefore the main contributor to our current housing crisis by greatly limiting how dense cities can become.
The Cost of Euclidean Zoning for Growth
Yet, although the role of Euclidean zoning in the housing crisis is undoubtedly huge, there is a second, perhaps even more important adverse effect of Euclidean zoning which has received much less attention: the cost of Euclidean zoning to productivity and GDP growth.
Costs for productivity are much less obvious. But recent work on agglomeration economies brings them into stark relief. Agglomeration economies refer to the simple idea that denser places are more productive. According to Glaeser, there are two mechanisms which cause density to be positively related to productivity: (1) the lower cost of moving things and (2) the higher likelihood of knowledge spillovers. The initial intuition behind (1) is simple. Nobel Prize-winning economist Paul Krugman’s model of agglomeration shows that firms benefit from decreases in the cost of moving goods since they are closer to suppliers and consumers. The same is also true for transporting people, i.e. commuting; transportation costs decrease with density, since people have to travel a shorter distance, and also benefit from faster public transportation. This increases the available labour supply and productivity: some individuals join the labour supply, and those already working can work more.
The second mechanism contributing to agglomeration effects is (2), knowledge spillovers: having other individuals proximate to you will make you more knowledgeable as a worker, thereby improving your productivity. A canonical example of knowledge spillovers is the collaborative environment of Silicon Valley, where having so many workers and firms in technology increases overall productivity. This was recently demonstrated by MIT economist David Atkin and colleagues, finding that in Silicon Valley, spontaneous face-to-face interactions due to so many technologists being in one place result in productive collaboration among firms.
Empirically, the power of agglomeration economies and therefore density for accelerating economic output are evident from measuring ‘agglomeration elasticities’. Agglomeration elasticities estimate the effect on firm and labour productivity of changing the population density where they are located, holding all else equal. A recent review by the Centre for Cities estimated that in the US, the agglomeration elasticity was 0.04, meaning a 1% increase in urban density leads to a full 4.6% increase in productivity. Corroborating this, economists Robert Hall and Antonio Ciccone find an individual moving to a city of double the density yields a 6% increase in their labour productivity. Furthermore, the effects of increased density were found to be even larger for high-skilled workers, which can be explained by knowledge spillovers, since there is more ‘knowledge’ to spillover.
It is therefore evident that the dominance of Euclidean zoning is preventing economic growth: Euclidean zoning's main effect — and historical purpose — is to ensure that most neighbourhoods remain low-density. However, agglomeration economics demonstrates that this low density has large costs for economic output and productivity, increasing transportation costs and reducing knowledge spillovers. This cost is technical and mostly hidden, meaning debates about zoning understandably focus on how the housing crisis affects people’s rents and ability to live where they want. Yet, to fully evaluate Euclidean zoning and the structure of the American housing market, it is essential to acknowledge the negative effects of Euclidean zoning on growth.