Continuous vs. Discrete Urban Ranks: Explaining the Evolution in the Italian Urban Hierarchy over Five Decades

Abstract

The reasons for changes in ranking within urban systems are a matter of a wide and long debate. Some focus on a continuous and smooth ordering of cities by their size within the urban system, in the tradition of Zipf’s law. Others focus on discrete, discontinuous ordering, as cities take on functions at different levels, such as specialized market places or high-level education, in the tradition of Christaller. We enter the debate by empirically evaluating whether the same determinants explain continuous or discrete changes in urban ranks in the evolution of the Italian urban hierarchy over the years 1971 to 2011. We empirically show that small, continuous changes of cities’ ranks have different drivers than large, discontinuous leaps. The presence of high-level functions in a city predicts major leaps across discrete ranks. Results are robust to the use of an instrumental variable strategy based on a shift–share argument.

Key words:

• central place theory
• agglomeration economies
• high-level functions
• functional upgrading

JEL codes:

• O18
• R11
• R12

Acknowledgments

We would like to thank Livia Fay Lucianetti (ISTAT) for kindly providing Functional Urban Areas shape files and Carlo Pisano (Bicocca University) for the revised 2011 census data with ISCO professions at the municipality level. We would also like to thank participants at the 59th Congress of the European Regional Science Association (ERSA), Lyon (France), August 27–30, 2019; the Resilient Built Environment for Sustainable Mediterranean Countries conference, Milan (Italy), September 4–5, 2019; the 66th Annual North American Meeting of the Regional Science Association International, Pittsburgh (PA), November 13–16, 2019; and handling editor Andrés Rodríguez-Pose, and two anonymous referees for helpful comments and suggestions. All remaining errors are our own.

Notes

1 Un bon réseau urbain hiérarchisé peut contribuer à favoriser un dévéloppement régional équilibré [a good hierachical urban network can contribute to favoring a balanced regional development] (Beguin, 1988, 242), authors’ translation.

2 Technical Appendix A.1 in the online material zooms in on the US and European urban system, highlighting that indeed Zipf’s law only holds when limiting the analysis to the subsample of medium and large cities. Technical Appendix A.2 in the online material presents the rank evolution in Italian cities in the period 1971–2011.

3 It is beyond our scope to fully discuss the spatial inequalities that arise from urban systems with persistence in top ranks and little movement in the right-hand side tail of the Zipf distribution. However, it is worth stressing that in the field of economic geography a lively debate took place on this issue. The latter also focused on the policies needed to correct the most severe consequences of these long-run trends, typically led by market forces that delivered less equal development than initially promised (see, e.g., Iammarino, Rodriguez-Pose, and Storper [2019 273] who speak of “place-sensitive distributed development policy”).

4 The number of two centers was not taken into consideration since it would have not overcome the problem of too few centers in a rank.

5 The NUTS classification (nomenclature of territorial units for statistics) is EUROSTAT’s system for classifying subnational statistics. It ranges from level 0 (country data) to level 5 (local administrative units, or LAUs).

6 Technical Appendix A.4 in the online material presents a number of robustness checks to further support this choice, as well as a test of the sensitivity of our findings to the choice of alternative geographies.

7 ISCO stands for “International Standard Classification of Occupations” and has been released by the International Labor Organization for classifying jobs according to their skill intensity.

8 For future work, the role played by changing transportation infrastructure in minimizing physical distance among selected nodes of, for example, fast railway connections and flights could be analyzed, and the robustness of our findings concerning borrowed size and functions tested in the light of the relevant evolution of the Italian network over the past half century.

9 “The Framework programmes have been the main financial tools through which the European Union supports research and development activities covering almost all scientific disciplines. FPs are proposed by the European Commission and adopted by Council and the European Parliament following a co-decision procedure” (European Commission 2022)

10 Additional robustness checks focusing on the role of these same explanatory variables in periods of generalized economic growth, and for different layers of the urban hierarchy, are presented in Appendix A.3 in the online material.