打印

作者

来源

URBAN GEOGRAPHY,Vol.42,Issue1,P.79-102

语言

英文

关键字

Monocentric city model,polycentricity,agglomeration economies,network economies,location decisions

作者单位

a Immobilien Research, Lidl Dienstleistung GmbH & Co. KG, Neckarsulm

摘要

The Munich Metropolitan Area (MMA) is embarking on becoming a more polycentric urban region. For long, the monocentric city model has provided strong explanatory power in understanding the location decisions of private households and firms. However, this model has been challenged by the emergence of large, polycentric city-regions. Increasing commute ranges and the spatial reorganization of economic activities show patterns of concentration and deconcentration at the same time but on different spatial scales. This paper gives an overview of and theorizing about the spatial driving forces and development within recent years. We use the interplay of agglomeration and network advantages to describe the spatial evolution of the MMA where new secondary centers emerge. Publicly available data and a large-scale web-survey provide the basis for descriptive and multivariate statistical analyses. We find – maybe not surprisingly that still the same driving forces of the monocentric city model – accessibility and economic competition – are at work: the MMA’s public transport rail network still forms a radial and monocentric system, while the present use of space by inhabitants shows an increasingly polycentric pattern with explicit tangential relations. We then discuss the future spatio-functional development path of the MMA by pitching the current strong population and employment growth with its tendency toward metropolisation against an alternative development path with competing driving forces. We conclude with visualizing and discussing the alternative spatial development that simultaneously projects concentration, deconcentration, and dispersion dynamics.KEYWORDS: Monocentric city modelpolycentricityagglomeration economiesnetwork economieslocation decisions1 IntroductionOver the last 100 years, we have witnessed the emergence of large-scale urban systems worldwide and observe an intensifying process of metropolisation of European city-regions. This development is accompanied by societal changes, such as smaller household sizes, older but more dynamic population, life-long learning and an ongoing process of up-scaling in which the spatial range of functional activities such as working, living, recreation, or shopping widens. New technologies in transportation and communication fundamentally affect the spatial range of individual activities and, hence, change the way private households make their location decisions.The process of metropolisation has general and specific features. At least for European city regions, general features include the upscaling of commuting distances, structural changes toward knowledge-intensive activities, and a re-concentration of people seeking urban environments. With the emergence of the knowledge economy, agglomeration and network economies have become important drivers of spatial development. The location decisions of private households and firms happen against the backdrop of agglomeration and network advantages (Bentlage et al., 2011; Van Meeteren, Neal et al., 2016; Zhao et al., 2017). Agglomeration advantages offer a multitude of options in close spatial proximity. This enables knowledge spill-overs between private households and firms as well as the opportunity to change jobs without relocation of residence (Parr, 2014; Rosenthal & Strange, 2003). Network advantages complement this spatial dimension and offer access to information, capital, and knowledge in remote locations (Castells, 2000). Cities and urban metropolitan regions with these advantages are demonstrably more competitive than others (Simmie, 2003; Van Oort et al., 2014). Agglomeration and network economies suggest that economic activities tend to concentrate in space.Metropolisation is also a specific process with specific features that differ between city regions as each region has its own history and geography. Many of these characteristics are based on past political decisions and thus produce development processes that are specific to that region. This includes the structure of the administrative body and the distribution of competences amongst these actors. Further specific features may be the morphology of the built environment, the transport system, or the distribution of functions in space.According to the general and specific features of the process of metropolisation, the outcome of a city region’s development is linked to the exploitation of agglomeration economies and network economies. This paper analyzes the location preferences of households within the Munich Metropolitan Area (MMA) and aims to understand to what extent do their location decisions support an emerging polycentric structure. We hypothesize that metropolisation in the MMA means a shift toward an increasingly polycentric morphological and functional structure by means of the same driving forces, which have evoked the monocentric city model – accessibility and economic competition (Clark, 2000, pp. 142, 145). Therefore, the transport infrastructure needs to be restructured in order to support this new polycentric structure. Subsequently, future development options shall address the provision and exploitation of agglomeration and network advantages.The spatial structure of the MMA represents an interesting spatial pattern for an analysis of future development options in a metropolitan context. Christaller showed in his empirical work on Central Places in the German South in the 1930s that the region is highly monocentric in terms of its central functions, which service a large hinterland (Christaller, 1968). Nowadays, this area encompasses a large-scale urban region including the City of Munich and the cities of Augsburg, Ingolstadt, Landshut and Rosenheim. The second biggest airport in Germany is located on the border of the two municipalities of Freising and Erding. The MMA is one of the most productive regions in Europe. It hosts a number of universities and several globally operating companies (Lüthi et al., 2010). The rail and motorway networks are radially directed toward the City of Munich. The public transport system dates from initial investments for the 1972 Summer Olympics and displays a radial system directed to the center of Munich. Within the last years the region experiences a restructuring of commuting flows around the City of Munich indicating a strong demand for new transport lines that enable tangential connections. Thus, the region’s current spatial organization does not correspond to people’s individual use of space.In addition, the MMA with its almost 5.9 million inhabitants experiences strong growth of population and employment. The forecast predicts an increase of population up to 6.4 million inhabitants by the year 2035 (Bayerisches Landesamt für Statistik, 2016). Such different circumstances among secondary cities and strong growth rates in the heart of the MMA is inducing increasing polycentricity in the region (Kinigadner et al., 2016). This transformation process revolves around adjustments in public transport infrastructure and a re-conceptualization of the region’s centers and peripheries.This paper is structured as follows: Section 2 outlines the theoretical background with reference to the monocentric city model and the extension to a polycentric city model. Section 3 introduces methods and data. Sections 4 discusses the extent to which the MMA represents a polycentric urban region and offers insights into the location decisions of private households within the MMA. Based on these analytical findings, new public transport projects and the subsequent emerging polycentric structure are considered. Section 5 presents conclusions drawn from the findings and discusses present planning strategies.2 Theoretical background: from monocentric to polycentric city modelsThe morphological and functional differentiation of city regions has been the subject of ongoing research for many decades. This section provides an explanation of the spatial-functional differentiation of city regions based on Alonso’s monocentric city model (Alonso, 1960). We then reflect upon the transformation into polycentric city regions. Finally, we discuss theoretical assumptions about the effects of growth on monocentric cities.2.1. The monocentric city modelThe monocentric city model as proposed by Alonso explains the spatial-functional differentiation of cities by means of the willingness to pay for a location and the distance from the urban center. A location in general represents a piece of land that offers certain opportunities at certain costs. An early conceptualization of the value of land and its usage is given by Von Thünen (1826). Alonso’s theory of the urban land market elaborates on the ideas of von Thünen and shows the interrelationship between what people bid for land and what kind of spatial usages result from this willingness to pay (Alonso, 1960). “When a purchaser acquires land, he acquires two goods (land and location) in only one transaction” (Alonso, 1960, p. 149). This transaction represents a trade-off between the size of the land and the location, which is given by closeness to the central business district (CBD), service supply, or other factors that enable a certain spatial usage. A bidder makes his bid based on what he is expecting to earn at a certain location.In this regard, we have to differentiate between the value of the transaction itself and the value of the expected utility that this piece of land provides (Lloyd & Dicken, 1978, p. 202). The transaction value represents the real costs that are paid for rent or purchase of land. The expected utility also comprises a non-monetary value that includes career prospects, ability to maintain social networks, or to organize mobility with a specific mode. In other words, to achieve optimal convenience and future prospects. Lloyd and Dicken (1978) use the term economic rent, which “is the surplus income which can be obtained from the unit of land above that which can be obtained from an inferior unit of land” (Lloyd & Dicken, 1978, p. 33). Within advanced economies, advantages from agglomeration and networks increase the value of land as individuals seek proximity to others.Alonso’s monocentric city model is centered on the CBD, in which employment is located at a single point (Figure 1(a)). The monocentric city model predicts a denser concentration of high order business activities in the center of the city. Such usages require locations with high centrality and frequent encounter with other actors. This results in high land prices at central locations.Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 1. Spatial-structural clusters (own illustration based on Thierstein et al. (2016, p. 26)Display full sizeFigure 1. Spatial-structural clusters (own illustration based on Thierstein et al. (2016, p. 26)The bid rent for such usages shows a very steep decline with increasing distance from the center. As the utility of a location too far away from the city center becomes too low, the probability of encounter decreases, and other competing functions acquires these locations. The bid rent curve of those functions is less steep and the willingness to pay does not decrease as strongly as for the former functions. However, these functions also have unacceptable locations that are too far away from the city center. In general, we observe decreasing building heights and lower densities the greater the distance from the center. Thus, we notice immediately that the functional structure and the morphology of a city region are interrelated.2.2. From monocentric models to polycentric city regionsThe transformation of monocentric city regions into polycentric regions has been investigated within spatial disciplines for decades (Arribas-Bel & Sanz-Gracia, 2014; Van Meeteren, Poorthuis et al., 2016). Different forces affect and change the spatial-functional pattern of a monocentric city model. These changes include, firstly, an up-scaling process due to declining transport costs and more efficient transport. Commuting distances increase and the activity ranges of households and firms widen. This results in an extension of the bid rent curves. Secondly, there is spatial concentration resulting from firms’ need to exploit the agglomeration advantages and preference of private households for urban lifestyles. Such development leads to higher densities in city centers and steeper bid rent curves.Both processes, up-scaling and concentration of activities trigger effects on other spatial entities. Firstly, as the hinterlands of cities expand geographically, bid rent curves of neighboring cities overlap in the in-between-areas. This may drive certain functions away from centers because of intertwined spatial usages between these two cities. From a morphological perspective, cities merge into one another, and boundaries blurr. From the functional perspective, land values may be higher between both cities as the in-between-area offers proximity to two centers. This attracts functions with higher value-added.A second implication has to do with the process of concentration. Concentration results from new companies locating in urban centers and from relocation of established firms to urban centers. Thus, concentration in one place often means loss for other places. As the term “agglomeration shadow” suggests, large-scale urban regions may experience winner-takes-it-all situations, in which major cities attract highest order functions and secondary cities fall behind (Burger et al., 2014). For bid rent curves, this implies that of the primary city becomes steeper as concentration fosters competition for land. The bid rent curve for the secondary city flattens due to drain of functions.By contrast, the concept of “borrowed size” predicts that a secondary city may attract functions that would have been expected to locate in primary cities. An important mechanism in profiting from the size of another place is provided by network economies (Alonso, 1973; Burger et al., 2014). Thus, a secondary city exploits the agglomeration advantages of the primary city with the help of good accessibility and connectivity to that primary city.The functional and morphological structure of a city region seems to be interrelated (Rauhut, 2017). The monocentric city model explains the functional-spatial structure. This model meets an emerging polycentric urban structure, in which functions such as employment, supply of daily goods and services, and residential use are located in multiple but interlinked centers. This also affects the morphology of those city regions as the form and density of settlement structures, the built environment, or transport infrastructure differ. Overall, these overlapping developments come together in order to merge in a process of metropolisation, which is defined as a process of functional integration (Boussauw et al., 2018, p. 2).2.3. The emergence of new centersThe aforementioned analyses of concentration and up-scaling show that city regions form interrelated nodes. Nevertheless, this does not yet qualify as a polycentric city model but the overlapping of monocentric city-regions, in which each city hosts a full set of functions – namely: business, commerce, residence, and agriculture and has the potential for further integration (Boussauw et al., 2018, p. 2). On the other hand, a polycentric city region may include centers with narrowly specific functions, such as business, commerce, residence, or agriculture. However, Meijers (2008) suggests that polycentric city regions are more than just the sum of small cities. Fujita and Ogawa (1982) formulate a polycentric city model in which employment is located in multiple centers. As agglomeration economies gain in importance, firms tend to concentrate where other firms are already located. The authors use the term locational potential which “varies among locations depending on the degree of concentration or dispersion of business firms” (Fujita & Ogawa, 1982, p. 163). As a result, the model of the city region becomes a multipolar pattern of locations offering different potentials.Agglomeration economies play a crucial role in the emerging polycentric urban region but the question as to where these new centers occur remains (Duranton & Puga, 2015, p. 503). If “new potential sites for cities are available, then agglomeration economies are essential to understand why cities exist at all.” (Duranton & Puga, 2014, p. 805) Knowledge-intensive activities in particular tend to concentrate in space thanks to the increasing importance of spatial proximity, which offers advantages and positive externalities for these activities. A further contribution in explaining polycentric city rests upon network economies. Jacobs’ assessment of the economy of cities offers an understanding in which cities evolve at intersections of trade, information, or capital (Jacobs, 1969). Following this argumentation, Taylor (2007) suggests that “city-ness” is one process that contributes to the spatial-functional differentiation of cities (Taylor, 2007, p. 292). City-ness refers to the network function of places as being interlinked with complementary nodes. Subsequently, locations that allow access to networks may have higher potential than others.It is widely acknowledged that agglomeration and network economies represent important assets for spatial development (Duranton & Puga, 2015; Fujita & Ogawa, 1982; Henderson, 1974; Mills, 1967). Recent studies have shown that agglomeration and network economies are synergistically interwoven (Rozenblat, 2010; Van Meeteren, Neal, et al., 2016). These externalities induce cumulative causation and polarization of spatial development. Cities with higher externalities of agglomeration and networks are more likely to attract further economic activities and highly qualified labor. Thus, the emergence of the knowledge economy is closely linked to the growth of urban regions (Glaeser, 2011). However, city regions that experience high-pressure growth have to deal with problems such as congestion or insufficient supply of housing. Often, these issues reach beyond city limits and across disciplines.(Mills, 1967, p. 199) discusses the outcome of growth in a monocentric city region and states that doubling the population of such city by higher buildings will increase the demand for transportation in that area over-proportionally. In order to build additional transport lines, buildings will have to be replaced and the population will move to the edge of the city. This will result in additional transport provision as distances increase. Such a situation may also be called the diseconomies of scale (Parr, 2004; Richardson, 1995).Recent advances provide evidence that congestion indeed increases as monocentric cities grow and become denser. Ferreira and Batey (2011, p. 245) show that the concentration of people results in unmanageable transportation in such urban patterns. In addition, Maat et al. (2005) found that “mode shifts due to slow mode friendly designs contribute more to sustainable travel patterns than do merely denser cities.” (Maat et al., 2005, p. 44) The reason for such unmanageable situations is the product of a lack of coordination: “each worker would prefer more cities of a smaller size each in their sector but a worker alone cannot create a new city” (Duranton & Puga, 2014, p. 810). Accordingly, as growth puts pressure on monocentric city regions, institutions are needed to reorganize spatial structure in order to keep the level of agglomeration and network economies high without the risk of experiencing diseconomies of scale. New centers function as nuclei for growth.At the same time, commuting trips and individual movements become more and more polycentric, demanding further tangential connections. We see that urban systems can at some point in time reach a level of congestion, infrastructural stress, hikes in real estate prices, environmental concerns, and such. Strong growth offers a window of opportunity for radically altering the logic of the urban system. Batten (2001) coins the term “spatial avalanches” to indicate that a multitude of individual decisions is able to trigger a radical change in an urban system, which we term “moment of change.” This represents a window of opportunity where responsible authorities as well as proactive firms and individuals have the chance to redirect existing strong pressure of growth toward transforming the monocentric region into a polycentric urban structure. This helps to sustain agglomeration advantages due to high density in close vicinity to public transport accessibility.Finally, this leads to our hypothesis: the MMA represents a monocentric city region with a radial transportation system. The densities of jobs, population, and services decrease from the city center to its surroundings. Further growth of population will lead to further congestions. To solve this problem, the region has to establish other centers that help to redirect and reconcentrate commutes.3 Methods and dataOur analysis follows two avenues. Firstly, we use indicators from official statistics to describe and assess the degree of polycentricity within the MMA. Secondly, we use data from our own data collection on private households that help to explain why people reorganize their spatial conditions. To this end, we set up a web survey, which is explained in detail in this section.Individuals aim at optimizing living and mobility costs and therefore the trade-off costs and benefits of individual features that generate their overall utility. Since housing cost in Germany on average amounts to 35% and transport to 14% of gross household budget, such trade-offs tend to be between more spacious dwellings and longer commutes from place of residence to place of work (Statistisches Bundesamt, 2018). We assume that each household aims to reduce costs for living and transport. This includes changes in mobility behavior and mode choices. Regarding residence or workplace locations, modifications may result in changes of domicile sizes, the centrality of each location, or the wish to own property. Accordingly, the location decisions of private households take place against the backdrop of complex and intertwining conditions that frame trade-offs between mobility and living costs.The location decisions of individuals are also taken on the basis of a bundle of different factors. This means that, for example, an apartment is not evaluated solely by its construction, but also by its spatial location, its proximity to the workplace, its closeness to retail and other amenities or accessibility to points of interest, and different modes of transport. All these factors represent one bundle and thus describe the supply side. As a result, location alternatives often differ not only according to one but a number of the factors mentioned, and an improvement in a factor in the bundle is often accompanied by a cut-back in other factors.Our analysis is based on an online questionnaire distributed to 6,962 households that met the following requirements at time when completing the web survey: firstly, the questionnaire addressed respondents that had worked or participated in higher education and training within the last three years. Secondly, respondents were required to have relocated either workplace or residence, or both, within the period of the last three years. We then focussed on private households that had undertaken a spatial change and accordingly dealt with the question of what spatial organization is best for them.The questionnaire was only available online and the survey was carried out from November 2014 to April 2015. It was announced several times on social media platforms, in newsletters of municipalities or companies. A press conference in January 2015 enabled strong coverage by means of regional newspapers and broadcasting in television and radio. The structure included three points in time: current situation, the time of search or relocation, and the situation before moving or relocating job location. Additionally, respondents gave information on their socio-demographic characteristics including size of household, age of household members, income, qualifications, and profession.The data collection used a Google maps plug-in to obtain georeferenced information on workplace and residence. Respondents were able to mark their current locations of work and residence, alternative locations for residence during search, and their former situation, again for both residence and work. Respondents with multiple workplaces were asked to indicate their main location of work at the moment. This provided fine-grained and precise address data. To this was added further data about a spatial structure such as supply facilities, accessibility, or demographic indicators in order to describe these locations by means of their spatial integration and quality. Apart from the georeferenced information the web survey also collected descriptions of residence locations in terms of size, rental, or purchasing costs as well whether the accommodation was rented or privately owned.An evaluation of the representativity of the questionnaire is not possible as information of relocation processes of private household below the municipality level does not exist. Nevertheless, we did some comparison of socio-economic factors of the respondents with the entire population. In general, the questionnaire includes households with rather higher education level and income.For our analysis, we included subjective data, which came from the georeferenced information we gathered from our web survey, with statistical spatial-structural data from our MMA. The following section describes this two-fold approach in detail.4 How polycentric is the MMA?This section firstly analyzes the functional and morphological development of the MMA within the last years. Secondly, it shows the location decisions of private households with regard to spatial accessibility. Based on these findings, it thirdly discusses future development options with a closer look at existing public transport infrastructure and potentials for new connections.The spatial structure of the MMA represents an interesting pattern for an analysis of future development options in a metropolitan context. The core of the region conveys the idea of an emerging, spatially discontinuous polycentric form, with a dense agglomeration of municipalities around the City of Munich. A fair number of “urban, peripheral” centers – regional sub-centers – are spread across the MMA. The question arises whether these places will be able to generate agglomeration economies that foster positive economic performance. The existence of urban catchment areas and peripheral locations points to another striking challenge for the MMA. The low accessibility engenders car traffic, which in turn triggers congestion in the main cities and central workplaces. It becomes obvious that mobility, housing, and spatial structure are interdependent factors that overstrain single municipalities and call for development strategies on at least a supra-local scale.4.1. Morphological and functional developmentThe spatial scale of this analysis is given by the definition of the Munich Metropolitan region. This scale encompasses the City of Munich and the next larger cities around Munich. Thus, we cannot provide a functional delamination of the region, but one that contains the main centers and covers a large area around these centers. Figure 2 Figure 1 shows the spatial-functional structure of the MMA based on a cluster analysis with 17 indicators such as population and workplace density, urban functions such as service infrastructure, shopping facilities and cultural facilities, accessibility by public transport and car, housing costs and building types the share of dwellings rented out as touristic accommodation.Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 2. Population concentrations in the MMA from 1990 to 2014 (own illustration based on Bayerisches Landesamt für Statistik (2018))Display full sizeFigure 2. Population concentrations in the MMA from 1990 to 2014 (own illustration based on Bayerisches Landesamt für Statistik (2018))The heart of the region is labeled “urban, central” and forms a discontinuous polycentric urban structure with highest density of employment, population, and supply facilities. This type includes the City of Munich and surrounding municipalities, with Freising and the airport. Municipalities in the cluster “urban, peripheral” are mainly cities that provide services and centrality to their regional context. The term peripheral points toward an area-wide distribution of these spatial units throughout the MMA. “Urban catchment area” represents municipalities that have a negative commuting balance while accommodation costs are relatively low. These places are home to people that purchase their own property in vicinity to the biggest cities such as Munich, Augsburg, or Ingolstadt. Thus, its spatial structure forms a first concentric area around these main cities. “Accommodation locations in tourist environment” represents a specificity to the MMA. This cluster includes municipalities that are located at the lakes and close to the Alps. Although accessibility and job density is lower than in cluster “urban, decentral,” accommodation costs are relatively high. The number of long-term shopping facilities is also high since these places host shops serving sports and touristic demand. Finally, the cluster “peripheral locations” shows the lowest values in accessibility, accommodation costs, and provision of services. It spatially connects to the urban catchment area in the sense that people with lower financial budgets decrease their living costs at the price of longer commutes and less access to services.The core of the region conveys the idea of an emerging, spatially discontinuous polycentric form, with a dense agglomeration of municipalities around the City of Munich. A fair number of “urban, peripheral” centers – regional sub-centers – are spread across the MMA. The question arises whether these places will be able to generate agglomeration economies that foster positive economic performance. The existence of urban catchment areas and peripheral locations points to another striking challenge for the MMA. The low accessibility engenders car traffic, which in turn triggers congestion in the main cities and central workplaces. It becomes obvious that mobility, housing, and spatial structure are interdependent factors that overstrain single municipalities and call for development strategies on at least a supra-local scale.Figure 2 shows the development of the concentration of population for each year between 1995 and 2018 for both the MMA and the labor market region of Munich. The latter is defined by intense commuting between the districts around the city of Munich (BBSR, Bundesinstitut für Bau-, Stadt- und Raumforschung, 2012). We use the Gini coefficient on the level of municipalities to assess the concentration of population. A Gini coefficient of a value of 1 indicates that the whole population is concentrated in one municipality; a value of 0 shows that the whole population is equally distributed across all spatial units. In our case, the Gini coefficient reaches values between 0.645 and 0.67 for the entire MMA and indicates that population is concentrated in the largest cities of Munich, Augsburg, and Ingolstadt. The labor market region of Munich shows even higher values of between 0.74 and 0.79. The population thus is mainly concentrated in the City of Munich. Although the dynamics are relatively stable throughout that period, we observe a clear trend of de-concentration until the year 2004 followed by an era of concentration for both spatial scales. The analysis of population development provides an index of morphological change within the region as larger cities tend to grow faster than smaller ones. Figure 4 Figure 3 shows the population development between 2013 and 2018 combined with the accessibility of jobs by public transport for each municipality. Accessibility was measured by a gravitational accessibility approach provided by Hansen (1959). Our analysis uses results for the accessibility of jobs by public transport from a calculation from Büttner et al. (2010). This calculation uses driving time within the network of public transport between all pairs of stations. These driving times are weighted by the population of origin and destination. We obtain an accessibility measure for each station that depends on the time required to get to centers of employment and the size of employment that works there. In order to compare results, we used values below or above the average accessibility of the entire region. Above-average population growth and at the same time above-average availability of jobs is mainly evident in Munich, Freising, Ingolstadt, and Augsburg. These are also the larger work centers of the region, which are engines of population growth. The municipalities around Ingolstadt, however, show above-average population development with below-average accessibility. This implies the population increases at locations with poorly developed public transport. Around the Augsburg area, a different effect can be seen. The population growth is below average, although the availability of jobs is relatively good. This suggests that the potential for population growth has not yet been exhausted at suitable locations. Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 3. Population development from 2013 to 2018 and accessibility (own illustration based on Thierstein et al. (2016))Display full sizeFigure 3. Population development from 2013 to 2018 and accessibility (own illustration based on Thierstein et al. (2016))Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 4. Home-to-work trips in Munich and its surrounding cities for the years 2009 and 2014 (own illustration based on Bundesagentur für Arbeit (2018))Display full sizeFigure 4. Home-to-work trips in Munich and its surrounding cities for the years 2009 and 2014 (own illustration based on Bundesagentur für Arbeit (2018))The MMA features a radial public transport network that is directed toward the City of Munich and lacks tangential connections. At the same time, we observe an increase of tangential commuting flows. Figure 5 Figure 4 shows commuting trips for the years 2009 and 2014 with a distance lower than 30 km. This filter was set because commuting to Munich – by far the largest concentration of workplaces – overshadows relations in the close vicinity around the city. In the year 2009 commuting trips are strongly directed toward Munich. In the year 2014, we observe an emerging polycentric structure with centers in the surroundings of Munich and in particular in the north of the city. These lateral connections are mainly accessed by individual transport. Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 5. Change of gravitation-based accessibility of population before and after moving residence (own illustration based on Thierstein et al. (2016))Display full sizeFigure 5. Change of gravitation-based accessibility of population before and after moving residence (own illustration based on Thierstein et al. (2016))One of these emerging centers is the city of Garching, which hosts a large campus of the Technical University of Munich and a large site of employment in Garching-Hochbrück. The city of Unterföhring southwest of Garching hosts the insurance company Allianz and a number of media and broadcasting companies. The City of Munich clearly remains the main employment center within the region, but the largest flow of workers commutes outwards to the city of Unterföhring: on a daily basis, 12,298 commuters travel from Munich to Unterföhring. This may indicate that employment locations are the main driver for an emerging polycentric structure around Munich.In summary, by using objective spatial-structural data our analysis shows a heterogeneous picture of the region. The inner part of the MMA forms an emerging polycentric urban space. Other regional centers appear to form an area-wide provision of services and infrastructures. Based on the concentration of population in the largest cities within the MMA and the changes in commuting flows around Munich, we observe a transformation of the region toward a polycentric urban mainly between the city of Munich and the airport. Nevertheless, this development is supported by a monocentric and radial public transport structure. In order to better understand how private households use their daily space, the next section reports results from our survey and casts light on the motives of households when choosing locations for residence and workplace as well as the interlinking choice of mobility.4.2. Location choices of private householdsThe relocation processes of private households’ trigger concentration or dispersal depending on the existing spatial structure and the motives involved. As public transport acts as an urban catalyst, we aim to understand how peoples´ motives for relocation interact with accessibility by public transport. The analysis combines motives for relation with gravitation-based accessibility of population to represent a measure of centrality based on the method shown above (Büttner et al., 2010; Hansen, 1959).Figure 5 shows the change of gravitation-based accessibility of population before and after moving residence broken down by motives for moving. A deterioration of gravitational accessibility in public transport is equivalent to a greater distance from main centers or a movement away from well-connected transport stops. A gain in accessibility indicates a move into larger centers or to transport stations with shorter driving times to these centers.This analysis shows that some of the reasons for relocation, like “residential costs too high” or “more persons in household” are related to a decrease in accessibility, while others such as “unsatisfactory public transport,” “broadband Internet not available” cause an increase in accessibility. For example, respondents who say they have relocated due to “entry into a job, start to work, beginning of education/study” have improved the accessibility on average by 20.0%. The relocation of workplace shows similar spatial patterns of concentration and dispersion. Interestingly, access to social infrastructure such as schools and childcare is one of the strongest drivers of deconcentration. This search for close proximity to metropolitan functions brings advantages for the daily life and produces a new organization of metropolitan life (Leven, 1978, p. 104). Another striking result is in the response given to the reason “journey to workplace too long” which does not show any difference with regard to accessibility. Two potential explanations may apply here. Firstly, reducing travel time to workplace is either realized by moving residence closer to workplace, as captured in Figure 5 Figure 6, or it is realized by relocating workplace closer to residences that are peripheral. A second potential explanation that one should take into account is that knowledge-intensive occupations tend to concentrate more densely in space than other occupations (Zhao et al., 2017). Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 6. Change of gravitation-based accessibility of population before and after moving residence (own illustration based on Thierstein et al. (2016))Display full sizeFigure 6. Change of gravitation-based accessibility of population before and after moving residence (own illustration based on Thierstein et al. (2016))Considerations of relocation constitute a multifactorial decision-making process. This means that some of the reasons for decisions overlap with others. For example, some respondents that moved residence because of high mobility costs also mentioned that journey to workplace is too long.The analysis of motives within the existing spatial structure of the MMA reveals a strong tendency of private households to either concentrate in the most central municipalities or in locations such as city catchment areas that are still close to these urban and central areas. The overall goal of this paper is to demonstrate agglomeration and network economies at work, which, in the context of the MMA, tend to create more polycentric structures. In the light of the analysis discussed, we see how accessibility is an important driver of development. However, due to high residential costs in the MMA, households tend to leave agglomeration centers in order to be able to afford larger dwelling sizes.4.3. Toward a polycentric city-region?The third part of the analysis discusses the empirical findings and takes planned and discussed changes in the spatial structure of the MMA into consideration. The theoretical survey and our empirical findings suggest that agglomeration and network advantages play a crucial role in the development process of a metropolitan region such as the MMA. Public transport may help to generate development forces that induce a polycentric transformation (Bertolini, 1999; Curtis et al., 2009). Such infrastructure generates “accessibility advantages” as a key service, which we have shown above to be decisive for location choices of households and firms (Bentlage et al., 2013). Thus, we consider accessibility as a tool for strategic development of agglomeration and network economies on different spatial scales.Figure 8 Figure 7 shows plans for future changes in Munich’s public rail network, including light rail, underground, and tram lines. Future changes originate from several planning documents (Landeshauptstadt München, Referat für Stadtplanung und Bauordnung, 2015, 2017; MVV, 2012; Niedermaier et al., 2017). Figure 7 Although single routes may change slightly, there is a clear indication of an intensifying polycentric structure within the City of Munich.Becoming more polycentric: public transport and location choices in the Munich Metropolitan AreaAll authorsMichael Bentlage , Christiane Müller & Alain Thierstein https://doi.org/10.1080/02723638.2020.1826729Published online:22 December 2020Figure 7. Existing and planned public transport railway lines (own illustration based on Landeshauptstadt München, Referat für Stadtplanung und Bauordnung (2015, 2017)); (MVV, 2012; Niedermaier et al., 2017))Display full sizeFigure 7. Existing and planned public transport railway lines (own illustration based on Landeshauptstadt München, Referat für Stadtplanung und Bauordnung (2015, 2017)); (MVV, 2012; Niedermaier et al., 2017))The first major change is located in the north of the city, where existing freight rails will be converted into a northern ring of the suburban train network. This area is currently an industrial area with large companies and a huge number of workplaces. Among others, BMW, MAN, and Knorr Bremse have headquarters or subsidiaries there. This intervention will fundamentally change the transport structure. A second major change is located in the south of the City of Munich. A number of new connections will link the biotechnology cluster in Martinsried with the western part of the city. These interventions will allow a stronger integration of employment centers within the city and enable structural changes in home-to-work trips. Next to these major changes, we expect a number of arterial linkages within the boundaries of the City of Munich, mainly provided by underground and tram. Thanks to these interventions, the suburban train system that provides service on a regional level will be complemented with better accessibility at the level of neighborhoods and city districts. As a result, the City of Munich will continue to grow strongly as a core city and this will continue to put pressure on the housing market.As indicated in Section 4, the MMA exhibits a discontinuous functional polycentric space. Neighboring municipalities such as Fürstenfeldbruck, Dachau, and Garching are connected to the City of Munich by light rail and underground. The main track through the inner City of Munich forms a monocentric transport structure and represents the backbone of this regional transport system. In fact, the entire regional transport system is very vulnerable and running over its limit capacity. Technical problems on the main track will affect the surrounding areas and a large number of commuters.In the following, we will present and discuss different development scenarios for the regional transport system: The “Erdinger Ring” adds a circular-like transport network to the region, which links the airport of Munich, the cities of Freising and Erding more closely. The transformation of these centers into inter-connected urban areas supports an emerging polycentric structure. Another major change may be realized with an extension of the underground line from Garching to the northern suburban train line. This intervention will connect the research center, the university, and the business campus of Garching directly with the airport.Reviewing all the planned and discussed changes, the City of Dachau is linked by a radial suburban rail track directing to the City of Munich. A future transformation of the public rail network with lateral linkages to the north east of Munich and the airport may face the challenge of relatively large distances between Dachau and other centers such as Unterschleißheim. This may be solved by deploying a Bus-Rapid-Transit (BRT) system of electrified buses with dedicated lanes or other transportation modes that run on an existing street network without calling for high investments in tracks and stations.A third scale represents inter-urban development options including the regional centers of Augsburg, Ingolstadt, Landshut, and Rosenheim. The development in these city regions depends strongly on agglomeration and network economies. The question remains as to how these cities will be able to exploit such advantages. Regarding agglomeration advantages, these cities will need to accelerate urban transformation in close proximity to public transport stops. This may include reuse of brownfields around stops or establishing dense and mixed-use spaces. This enables cities to make use of the centralizing forces of accessibility in which functions, population, and workplace concentrate at transport nodes. Further action addresses the regional transport network. Although it is very unlikely to happen, new connections between these regional cities may help to foster the potential of an inter-urban polycentric region. In this regard, Munich airport offers the potential to bring global accessibility to the entire MMA by means of a high-speed rail connection between the airport and the larger cities. Such an alternative future is very unlikely to happen as the investments required to link the secondary cities of the region are too large.Comparing these options on different scales reveals a strong component of competing forces in location development. On the first scale, the city Munich benefits strongly from new tangential lines on the city’s territory. This contributes to urban transformation process within the city and comes at the price of rising real estate prices. On the second scale, development of the metropolitan core represents a strong interplay of changes in surrounding cities of Munich and new tangential transport lines. This induces a process of urbanization in smaller cities and villages. This perspective highlights the cumulative development of agglomeration and network advantages. The third scale with its inter-urban development shows opposing forces for the development of the metropolitan core. This development option shows that regional cities have to foster endogenous potential for urban renewal to attract people and companies who demand mixed-use areas, geographical proximity, and high-quality public space. If cities such as Augsburg, Ingolstadt, or Landshut are not able to implement their own efficient supra-local transport networks and improve their urban qualities, they will not be able to compete on a level playing field with the capital City of Munich in offering complementary division of labor for knowledge creation, employment, and population. The result will be that the inner metropolitan core continues to grow faster than secondary cities.5 ConclusionThe driving forces of concentration and dispersion affect the morphological and functional structure of the region. Firstly, we see that ongoing strong growth rates and high pressure on central locations require an adjustment of both the regions’ spatial structure and transport system, which might eventually lead to a more polycentric spatial strategy. However, the region follows a development path that is strongly induced by a radial transport system, which probably would be labeled “path dependency” within an evolutionary economic framework (Martin & Sunley, 2006). Certainly, the realization of the second express rail track manifests a certain stickiness of path development. Our paper but argues with the interplay of agglomeration and network economies and thus a change in spatial structure and the creation of a new development path offered by new transport lines, new centers, or employment locations may offer alternatives to certain demands and triggers a polycentric re-conceptualization that transforms dispersion of population into decentralized concentration and multiple cores within the region. Therefore, regional development strategies should consider agglomeration and network advantages more explicitly. Accessibility to public transport can be used as one strategic tool to develop locations in terms of centrality. This increase in accessibility provides the base for further spatial development around stations.Secondly, the decisions of individuals are made in the context of the existing urban system. Changes in spatial structure such as new railway lines may also induce varying individual preferences as locational potential rises at those stations with better accessibility. Private households seek to realize their preferences in a given spatial structure. Due to budgetary constraints, not all preferences will be satisfied. The decision-makers then need to consider alternatives that reflect a trade-off between their preferences and reality. In many cases, this results in spatial dispersion with spread-out spatial development. In such a spatial setting, people want to minimize distances but are not able to do so. This calls for integrative approaches in planning in order to balance residential costs and mobility costs (Boussauw et al., 2018, p. 3).Thirdly, the MMA represents an example of a monocentric region that might embark on a more polycentric future. This development seems to depend strongly on the inner part of the MMA with the international Munich airport, the City of Munich, and its inter-connected surrounding municipalities. Although this will remain the dominant core of the region, we assume that if it is not able to transform to a polycentric region, the MMA might experience diseconomies of scale in the long run. The question arises as to whether the secondary cities of Augsburg, Ingolstadt, Landshut and Rosenheim may benefit from the transformation within the core of the MMA, and may thus borrow size from there or experience agglomeration shadows. Such development depends strongly on the speed of the transformation within these secondary cities. As shown above for the case of Ingolstadt, the growth of population at locations with low accessibility increases the car dependency and causes additional congestion. A structural improvement requires an upgrading of the regional public transport infrastructure and a concentration of settlement development at these locations. Thus, key drivers are urban functions and regional public transport systems, which both help to create agglomeration and network advantages. If these secondary cities are able to improve their urban character, they will be able to compete with the urban, central core, which reduces pressure of growth in this area.Disclosure statementNo potential conflict of interest was reported by the authors.ReferencesAlonso, William. (1960). A theory of the urban land market. Papers in Regional Science, 6(1), 149–157. https://doi.org/10.1111/j.1435-5597.1960.tb01710.x [Crossref], [Google Scholar]Alonso, William. (1973). Urban zero population growth. Daedalus, 102(4), 191–206. [Google Scholar]Arribas-Bel, Daniel, & Sanz-Gracia, Fernando. (2014). The validity of the monocentric city model in a polycentric age: US metropolitan areas in 1990, 2000 and 2010. Urban Geography, 35(7), 980–997. https://doi.org/10.1080/02723638.2014.940693 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Batten, David F. (2001). Complex landscapes of spatial interaction. The Annals of Regional Science, 35(1), 81–111. https://doi.org/10.1007/s001680000032 [Crossref], [Web of Science ®], [Google Scholar]Bayerisches Landesamt für Statistik. (2016). Genesis-online - Statistisches informationssystem bayern. https://www.statistikdaten.bayern.de/genesis/online/logon [Google Scholar]Bayerisches Landesamt für Statistik. (2018). Genesis-online - Statistisches informationssystem bayern. https://www.statistikdaten.bayern.de/genesis/online/logon [Google Scholar]BBSR, Bundesinstitut für Bau-, Stadt- und Raumforschung. (2012). Laufende Raumbeobachtung - Raumabgrenzungen. Arbeitsmarktregionen. http://www.bbsr.bund.de/BBSR/DE/Raumbeobachtung/Raumabgrenzungen/AMR/Arbeitsmarktregionen.html?nn=443270 [Google Scholar]Bentlage, Michael, Lüthi, Stefan, & Thierstein, Alain. (2013). Knowledge creation in German agglomerations and accessibility – An approach involving non-physical connectivity. Cities, 30(1), 47–58. https://doi.org/10.1016/j.cities.2012.07.003 [Crossref], [Google Scholar]Bentlage, Michael, Thierstein, Alain, & Lüthi, Stefan. (2011). Intra firm networks in the German knowledge economy. Economic performance of German agglomerations from a relational perspective. Paper presented at the European Regional Science Association, Barcelona. [Google Scholar]Bertolini, Luca. (1999). Spatial development patterns and public transport: The application of an analytical model in the Netherlands. Planning Practice & Research, 14(2), 199–210. https://doi.org/10.1080/02697459915724 [Taylor & Francis Online], [Google Scholar]Boussauw, Kobe, Van Meeteren, Michiel, Sansen, Joren, Meijers, Evert, Storme, Tom, Louw, Erik, Derudder, Ben, & Witlox, Frank. (2018). Planning for agglomeration economies in a polycentric region: Envisioning an efficient metropolitan core area in Flanders. European Journal of Spatial Development, 69, 1–20. [Google Scholar]Bundesagentur für Arbeit. (2018). Pendler in den Gemeinden Bayerns. [Google Scholar]Burger, Martijn J., Meijers, Evert J., Hoogerbrugge, Marloes M., & Tresserra, Jaume Masip. (2014). Borrowed size, agglomeration shadows and cultural amenities in North-West Europe. European Planning Studies, 23(6), 1090–1109. https://doi.org/10.1080/09654313.2014.905002 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Büttner, Benjamin, Keller, Johannes, & Wulfhorst, Gebhard. (2010). Ein Erreichbarkeitsatlas für die Europäische Metropolregion München. Retrieved from München. [Google Scholar]Castells, Manuel. (2000). The rise of the network society. The information age: Economy, society and culture (2nd ed.). Blackwell. [Google Scholar]Christaller, Walter. (1968). Die zentralen Orte in Süddeutschland. Eine ökonomisch-geographische Untersuchung über die Gesetzmäßigkeit der Verbreitung und Entwicklung der Siedlungen mit städtischen Funktionen (2 ed.). Wissenschaftliche Buchgesellschaft. [Google Scholar]Clark, William A. V. (2000). Monocentric to policentric: new urban forms and old paradigms. In Gary Bridge & Sophie Watson (Eds.), A companion to the city (pp. 141–154). Wiley-Blackwell. [Google Scholar]Curtis, Carey, Renne, John L., & Bertolini, Luca (Eds.). (2009). Transit oriented development. Make it happen. Ashgate. [Google Scholar]Duranton, Gilles, & Puga, Diego. (2014). The growth of cities. In Philippe Aghion & Steven N. Durlauf (Eds.), Handbook of economic growth (Vol. 2B, pp. 781–853). Elsevier. [Crossref], [Google Scholar]Duranton, Gilles, & Puga, Diego. (2015). Urban land use. In Gilles Duranton, John Vernon Henderson, & William C. Strange (Eds.), Handbook of regional and urban economics (Vol. 5A, pp. 467–560). Elsevier B.V. [Google Scholar]Ferreira, António, & Batey, Peter. (2011). On why planning should not reinforce self-reinforcing trends: A cautionary analysis of the compact-city proposal applied to large cities. Environment and Planning B: Planning and Design, 38(2), 231–247. https://doi.org/10.1068/b36102 [Crossref], [Google Scholar]Fujita, Masahisa, & Ogawa, Hideaki. (1982). Multiple equilibria and structural transition of non-monocentric urban configurations. Regional Science and Urban Economics, 12(2), 161–196. https://doi.org/10.1016/0166-0462(82)90031-X [Crossref], [Web of Science ®], [Google Scholar]Glaeser, Edward. (2011). Triumph of the city: How our greatest invention makes us richer, smarter, greener, healthier, and happier. The Penguin Press. [Google Scholar]Hansen, Walter G. (1959). How accessibility shapes land use. Journal of the American Institute of Planners, 25(2), 73–76. https://doi.org/10.1080/01944365908978307 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Henderson, J. Vernon. (1974). The sizes and types of cities. The American Economic Review, 64(4), 640–656. [Web of Science ®], [Google Scholar]Jacobs, Jane. (1969). Cities first, rural development later. In Jane Jacobs (Ed.), The economy of cities (pp. 13–54). Random House. [Google Scholar]Kinigadner, Julia, Wenner, Fabian, Bentlage, Michael, Klug, Stefan, Wulfhorst, Gebhard, & Thierstein, Alain. (2016). Future perspectives for the Munich Metropolitan Region – An integrated mobility approach. Transportation Research Procedia, 19, 94–108. https://doi.org/10.1016/j.trpro.2016.12.071 [Crossref], [Google Scholar]Landeshauptstadt München, Referat für Stadtplanung und Bauordnung. (2015). Nahverkehrsplan der Landeshauptstadt München. Anlage 3. Landeshauptstadt München. https://www.ris-muenchen.de/RII/RII/ris_vorlagen_dokumente.jsp?risid=3703936 [Google Scholar]Landeshauptstadt München, Referat für Stadtplanung und Bauordnung. (2017). Verkehrskonzept Münchner Norden Öffentlicher Personennahverkehr/Schienenpresonennahverkehr. Referat für Stadtplanung und Bauordnung. [Google Scholar]Leven, Charles L. (1978). Growth and nongrowth in metropolitan areas and the emergence of polycentric metropolitan form. Papers of the Regional Science Association, 41(1), 101–112. https://doi.org/10.1111/j.1435-5597.1978.tb01040.x [Crossref], [Google Scholar]Lloyd, Peter E., & Dicken, Peter. (1978). Location in space. A theoretical approach to economic geography (2nd ed.). Harper & Row. [Google Scholar]Lüthi, Stefan, Thierstein, Alain, & Goebel, Viktor. (2010). Intra-firm and extra-firm linkages in the knowledge economy: The case of the emerging mega-city region of Munich. Global Networks, 10(1), 114–137. https://doi.org/10.1111/j.1471-0374.2010.00277.x [Crossref], [Web of Science ®], [Google Scholar]Maat, Kees, Van Wee, Bert, & Stead, Dominic. (2005). Land use and travel behaviour: Expected effects from the perspective of utility theory and activity-based theories. Environment and Planning B: Planning and Design, 32(1), 33–46. https://doi.org/10.1068/b31106 [Crossref], [Web of Science ®], [Google Scholar]Martin, Ron, & Sunley, Peter. (2006). Path dependence and regional economic evolution. Journal of Economic Geography, 6(4), 395–437. https://doi.org/10.1093/jeg/lbl012 [Crossref], [Web of Science ®], [Google Scholar]Meijers, Evert. (2008). Summing small cities does not make a large city: Polycentric urban regions and the provision of cultural, leisure and sports amenities. Urban Studies, 45(11), 2323–2342. https://doi.org/10.1177/0042098008095870 [Crossref], [Web of Science ®], [Google Scholar]Mills, Edwin S. (1967). An aggregative model of resource allocation in a metropolitan area. The American Economic Review, 57(2), 197–210. [Web of Science ®], [Google Scholar]MVV. (2012). Regionaler Nahverkehrsplan für das Gebiet des Münchner Verkehrs- und Tarifverbundes. [Google Scholar]Niedermaier, Josef, Löwl, Stefan, Niedergesäß, Robert, Bayerstorfer, Martin, Hauner, Josef, Karmasin, Thomas, & Roth, Karl. (2017). Positionspapier der Verbundlandkreise im MVV Zukunftsperspektiven für die S-Bahn München aus Sicht der Verbundlandkreise. Verbundlandkreise im MVV. [Google Scholar]Parr, J B. (2014). The regional economy, spatial structure and regional urban systems. Regional Studies, 48(12), 1926–1938. https://doi.org/10.1080/00343404.2013.799759 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Parr, John B. (2004). Economies of scope and economies of agglomeration: The Goldstein-Gronberg contribution revisited. The Annals of Regional Science, 38(1), 1–11. https://doi.org/10.1007/s00168-003-0142-0 [Crossref], [Web of Science ®], [Google Scholar]Rauhut, Danie. (2017). Polycentricity – One concept or many? European Planning Studies, 25(2), 332–348. https://doi.org/10.1080/09654313.2016.1276157 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Richardson, Harry W. (1995). Economies and diseconomies of agglomeration. In H Giersch (Ed.), Urban agglomeration and economic growth (pp. 123–155). Springer-Verlag. [Crossref], [Google Scholar]Rosenthal, Stuart S., & Strange, William C. (2003). Geography, industrial organization, and agglomeration. The Review of Economics and Statistics, 85(2), 377–393. https://doi.org/10.1162/003465303765299882 [Crossref], [Web of Science ®], [Google Scholar]Rozenblat, Céline. (2010). Opening the black box of agglomeration economies for measuring cities’ competitiveness through international firm networks. Urban Studies, 47(13), 2841–2865. https://doi.org/10.1177/0042098010377369 [Crossref], [Web of Science ®], [Google Scholar]Simmie, James. (2003). Innovation and urban regions as national and international nodes for the transfer and sharing of knowledge. Regional Studies, 37(6), 607–620. https://doi.org/10.1080/0034340032000108714 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Statistisches Bundesamt. (2018). Private Konsumausgaben (Lebenshaltungskosten) - Deutschland 2016. https://www.destatis.de/DE/ZahlenFakten/GesellschaftStaat/EinkommenKonsumLebensbedingungen/Konsumausgaben/Tabellen/Gebietsstaende.html [Google Scholar]Taylor, Peter J. (2007). Cities within spaces of flows: Theses for a materialist understanding of the external relations of cities. In Peter Taylor, Ben Derudder, Pieter Saey, & Frank Witlox (Eds.), Cities in globalization. Practices, policies and theories (pp. 287–297). Routledge. [Google Scholar]Thierstein, Alain, Wulfhorst, Gebhard, Bentlage, Michael, Klug, Stefan, Gilliard, Lukas, Ji, Chenyi, & Zhao, Juanjuan. (2016). WAM Wohnen Arbeiten Mobilität. Veränderungsdynamiken und Entwicklungsoptionen für die Metropolregion München. Retrieved from München. [Google Scholar]Van Meeteren, Michiel, Neal, Zachary, & Derudder, Ben. (2016). Disentangling agglomeration and network externalities: A conceptual typology. Papers in Regional Science, 95(1), 61–80. https://doi.org/10.1111/pirs.12214 [Crossref], [Web of Science ®], [Google Scholar]Van Meeteren, Michiel, Poorthuis, Ate, Derudder, Ben, & Witlox, Frank. (2016). Pacifying Babel’s Tower: A scientometric analysis of polycentricity in urban research. Urban Studies, 53(6), 1278–1298. https://doi.org/10.1177/0042098015573455 [Crossref], [Web of Science ®], [Google Scholar]Van Oort, Frank, de Geus, Stefan, & Dogaru, Teodora. (2014). Related variety and regional economic growth in a cross-section of European urban regions. European Planning Studies, 1–18. https://doi.org/10.1080/09654313.2014.905003 [Web of Science ®], [Google Scholar]Von Thünen, Johann Heinrich. (1826). Der isolierte Staat in Beziehung auf Landwirtschaft und Nationalökonomie. Nachdruck 1990. Akademie. [Google Scholar]Zhao, Juanjuan, Bentlage, Michael, & Thierstein, Alain. (2017). Residence, workplace and commute: Interrelated spatial choices of knowledge workers in the metropolitan region of Munich. Journal of Transport Geography, 62, 197–212. https://doi.org/10.1016/j.jtrangeo.2017.05.012 [Crossref], [Web of Science ®], [Google Scholar]