Authors: Aurelie Lalanne*, University of Bordeaux
Topics: Economic Geography, Regional Geography, Urban Geography
Keywords: urban systems, gibrat's law, stochastic urban growth, urban growth, regional urban system
Session Type: Virtual Paper
Start / End Time: 4:40 PM / 5:55 PM
Room: Virtual 20
Presentation File: No File Uploaded
In urban economics, two laws summarize the functioning of national urban systems. Urban hierarchies converge towards a stable state in time (Zipf's Law) thanks to a stochastic process (Gibrat's Law) (Gabaix, 1999). These approaches of organization of urban hierarchies assumes thus that, for an urban system to be stable, all the cities that comprise it follow the same stochastic process without differentiation (Lalanne and Zumpe, 2020). It implies that the diversity of growth processes depending territorial and regional specificities are not integrated into this approach (Lalanne, 2014). The linearity depicted at aggregate national scales actually obscures variation observed in a multi-scale regional analysis (Bessey, 2002). We assume in this paper that cities are complex systems structured by key processes that occur at distinct spatial and temporal scales, and that regional scale as national sub-system is relevant to better apprehend different urban growth processes. Considering the Gibrat’s law mentioned above, we expect that big, medium and small sized cities within an urban system don’t respond to the same urban growth processes and to the same stochastic processes. This paper explores the different growth processes within a French regional urban system among big, medium and small sized cities. For this purpose, we use French censuses data from 1800 to 2016 for a french south-west region, the Nouvelle-Aquitaine. Moreover, we use methodological improvements to detect stochastic processes (Lalanne and Zumpe, 2020) as the unit root tests and the cointegration tests.