Airports and population density: where benefits outweigh costs

ABSTRACT

This study investigates the impact of airports on population density distribution in the surrounding areas. Using data for 197 European airports, results show that a 10% increase in airport activity leads to a growth in population density of 3.8%. While moving farther from the city centre leads to a decline in population density, we demonstrate that, on average, population density increases up to 7.5 km from the airport, then starts to decrease beyond that threshold. This threshold can be interpreted as the point in space where positive effects start outweighing negative externalities.

KEYWORDS:

• population density
• economic development
• spatial distribution
• infrastructure

JEL:

• L93
• O18
• R41

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. See the section dedicated to causality issues, dealing with the chicken-and-egg question of which comes first, transport infrastructure or socio-economic development?

2. Recently, scholars have been debating the spatial structures of cities (Angel & Blei, 2016; Salvati et al., 2016). Considering that there is still no agreement on the spatial structure in the European context and the well-known role of airport as CBD (e.g., McMillen, 2004), this study relies on a polycentric model of population density (e.g., Anas et al., 1998; Heikkila et al., 1989; Huang et al., 2017; McDonald & Prather, 1994; Small & Song, 1994).

3. The literature recognizes three main forms of the polycentric density function (e.g., Anas et al., 1998; Heikkila et al., 1989; Small & Song, 1994), which vary according to the roles of the subcentres in influencing population density. To this extent, we assume that the centres taken into consideration (i.e., the closest big city and the airport) cannot be considered as substitutes; rather, they perform different functions and people living in a single cell may require access to both.

4. In absence of information on airport-specific departure and arrival procedures, orientation to the runway represents a good proxy to account for the distribution of aircraft negative externalities.

5. For airports with more than one runway, we consider the longest and thus most impacting airstrip.

6. This value represents the sum of the theoretical number of people/km² living at $x_A=0$ and $x_C=0$.

7. WLU are the sum of moved passengers and cargo tons, where a passenger corresponds to 100 kg.

8. Regression analyses pooled two years for more precise estimators, while including a year dummy to allow the intercept to differ across periods (Wooldridge, 2016).

9. To overcome the potential issue related to the influence of flight distance of an airport on its local economy, we consider the same IV computed in similar markets as performed in former studies (e.g., Morlotti et al., 2017; Mumbower et al., 2014).

10. Source: http://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/population-distribution-demography/geostat/.

11. The most updated dataset today.

12. The centre of the closest big city represents the city geographical centre.

13. We rely on previous studies to fix the distance of 15 km from the airport (Jud & Winkler, 2006; McMillen, 2004; Mense & Kholodilin, 2014; Tomkins et al., 1998).