Understanding metropolitan growth in German polycentric urban regions

ABSTRACT

The paper investigates the peculiarities of polycentric urban regions (PURs) in generating agglomeration economies, namely, the question of how urbanisation externalities – size, density and diversity – take effect in German PURs and influence the growth of metropolitan functions from 1995 to 2017. The metropolitan growth of PURs is compared with that of urban regions showing a dominant core. We use our findings to address questions about the forces of urbanization externalities on a regional scale. Finally, we outline what we see as emergent research topics in the interplay between metropolization and the development of PURs.

KEYWORDS:

• polycentric urban region
• polycentricity
• metropolitan function
• urban growth
• urban system
• Germany

JEL:

• R10
• R12

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. Formally, this is noted as:$\mathrm{H}\mathrm{H}\mathrm{I}={\sum }_{i=1}^n{a}_i^2,$

where a_i is the share of the population/employees for region i. The minimum HHI is 1/n in the number of the cities/districts (n = 140), and the maximum HHI is 1, corresponding to a purely monocentric structure.

2. DAYPOP reflects the urban size with which the population increases or decreases by the commuter balance to represent the number of persons who participate in the working life of the city.

3. The indicators are not collected annually; therefore, they are collected for the periods from 1995 to 1997 and from 2014 to 2017.

4. The sub-functions are split into eight sub-functions instead of the four functions because they show different locational systems. For instance, the control function refers to both (1) economic entities such as global headquarters and (2) national and international policies. National and international institutions are usually localized in certain cities on the basis of political decisions that are based on planning and structural policy objectives. The locations of private-sector firms are generally based on economic location decisions.

5. We are aware of the potential modifiable area unit problems (MAUPs) resulting from the small-scale spatial distribution of the phenomena behind the collected data. Thus, urban research in general is being sensitive towards the data structure, measurements and spatial delineation of the study units (Taubenböck et al., 2019). However, since some of the time-series data, which are essential to generate a reliable data set over 20 years, are only available for NUTS-3 administrative units, we cannot completely circumvent these problems in the analyses. However, the measurement concept is robust by testing several periods in between, and so the results seem to be stable over time (Volgmann, 2014).

6. A city is considered to be diversified if its sub-function structure in 1995/97 differs from the national structure:$RDI=1/\sum _{s=1}^m\left|{\displaystyle \frac{Esi}{{\sum }_{i=1}^{n}Esi}-{\displaystyle \frac{{\sum }_{s=1}^{n}Esi}{{\sum }_{s=1}^m{\sum }_{i=1}^{n}Esi}}}\right|$where E_si is the sub-function s in city i; and E_s is the total sum of the eight sub-functions in city i. Analogous to this, E is the total national metropolitan index (all cities and districts within the 13 urban regions); and E_s is the national index of sub-function s (NUTS-3 regions within the 13 urban regions). The greater the RDI, the more diversified is city i (Farhauer & Kröll, 2012).

7. Urban density is measured by:$\mathrm{H}\mathrm{H}{\mathrm{I}}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}}={\sum }_{i=1}^n{a}_i^2,$which provides an indication of the concentration or dispersion of the metropolitan functions within an urban region (see note 1).