FAR regulations and the spatial size of Brazilian cities

ABSTRACT

This paper evaluates the influence of the maximum allowed floor area ratio (FAR) on the spatial size of Brazilian cities. We built a novel database on building height restrictions for the 325 largest cities in the country and combined it with recent satellite data. Our estimations show that, as predicted by the monocentric model, tighter constraints are associated with larger urban areas, and this result is robust to several specifications, sensitivity tests and different estimation strategies. Additionally, back-of-the-envelope calculations show that a higher stringency of the building-height regulation generates an annual cost of about US$9.67 million per average Brazilian city.

KEYWORDS:

• floor area ratio (FAR) regulations
• urban sprawl
• land-use restrictions
• Brazil

JEL:

• H11
• H77
• R11

ACKNOWLEDGEMENTS

The authors thank Cláudio Shikida, Rafael Pereira, Vanessa Nadalin, Bernardo Furtado, Daniel da Mata, and the participants at the UFPel, IPEA, ENABER and ENAP seminars for comments and suggestions. Eduardo Cenci generously shared data on the roughness of the Brazilian territory. The authors also thank Pedro Borges for his excellent research assistance.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. UGBs establish artificial boundaries in the city where lands inside the border can have urban use, and outside land can only have agricultural use or remain preserved.

2 The maximum FAR is a number multiplied by the area of a lot. It indicates the maximum amount of built-up area that developers can construct on the lot. For example, on a plot of 100 m²and maximum FAR = 4, the maximum building area must be equivalent to 400 m².

3. The authors defined the planning restrictiveness variable as the ratio between the quota planned to be converted into urban constructions and the amount effectively converted.

4. Although only plausibly exogenous, we believe this instrument brings advances in the identification compared with previous studies closer to ours that investigate the association between land-use regulations and urban area size. For example, Zhou et al. (2017) and Geshkov and DeSalvo (2012) do not tackle any strategy capable of limiting or solving the endogeneity issues. On the other hand, despite using an instrumental variable/two-stage least squares (IV/2SLS) approach, Brueckner and Sridhar (2012) adopt two instruments for the FAR regulation (an indicator of a city localised in a state controlled by the Indian National Congress (INC) party and population growth between 1950 and 2000) where the validity of the exogeneity condition is unclear and little investigated throughout the paper.

5. It is possible to demonstrate that the housing price falls with increasing distance to the CBD, that is, $\mathrm{\partial p}/\mathrm{\partial x}<0$. Lower prices discourage capital-intensive buildings in peripheral areas, so that $\mathrm{\partial S}/\mathrm{\partial x}<0$.

6. Usually, the local zoning law or the municipal master plan defines the building-height limits.

7. It is not feasible to build a panel database with the maximum allowed FAR due to the impossibility of tracking the historical evolution of the maximum FAR adopted in each city.

8. This database is freely available from the authors upon request.

9. Of the 42 cities that do not adopt the maximum allowed FAR regulation, only two implement regulations that impose restrictions on the building height or number of floors.

10. This measure is calculated by the ratio between the number of high-income homeowners and the total number of high-income households.

11. We did not include controls associated with the housing market because these tend to be directly influenced by our variable of interest (maximum allowed FAR), which generates the problem of bad controls. Furthermore, according to the monocentric urban model, the mechanism by which FAR regulation affects the spatial size of the city occurs through changes in housing construction patterns (lesser construction in the CBD accompanied by more construction on the urban edge).

12. Brazil has five macro-regions: North, Northeast, Midwest, Southeast and South.

13. Fine particulate air pollution is a known health problem and our estimates predict an unequivocal increase in particulate emissions due to the rise in commuting times. However, the impact of this on the health of urban residents is less clear, because a larger area of cities provides more airspace for dispersing these particulates. Therefore, we have chosen not to include in our estimates the potential health costs associated with this type of pollution.

14. Brazil has two types of property taxes collected by municipalities: the Imposto de Transmissão de Bens Imóveis (ITBI) and the Imposto Predial Territorial Urbano (IPTU). We use the average amounts collected in 2019 as a reference in this calculation.

Additional information

Funding

This study was mainly financed by the Escola Nacional de Administração Pública – Brazil (ENAP) and in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES) [Finance Code 001] and CAPES-PROEX. Ricardo Lima thanks the Conselho Nacional de Desenvolvimento Científico e Tecnológico – Brazil (CNPq) for financial support [grant number 303229/2022-5].