Polycentricity, Performance and Planning: Concepts, Evidence and Policy in Barcelona, Catalonia

More than half of the world’s population currently lives in urban settlements, a proportion that is expected to increase to more than 65 percent by 2050 (UN, 2014). The larger agglomerations are a complex spatial configuration of places and flows that are polycentric by nature, or at least they demonstrate a certain development of a multi-center structure. Recently, the focus on agglomerations’ polycentric structure has attracted a great deal of attention from both researchers and policymakers, who must manage the economic, social, and environmental challenges that the population of these metropolitan agglomerations will experience in the coming decades.

In research, a considerable portion of the study of polycentric agglomerations has focused on the conceptualization of polycentricity and the empirical analysis of its economic, social, and environmental dis(advantages). Although academics have made a strong effort both to clarify the concept of polycentricity and to empirically explore its dis(advantages)–see, e.g., the special issues of journals such as European Planning Studies (1998; 2015), Urban Studies (2001) and Regional Studies (2014)–two major issues remain in the literature.

First, various approaches to polycentricity co-exist without a high level of integration. One approach refers to polycentricity on the intra-urban (Davoudi, 2003) or intra-metropolitan scale (Brezzi and Veneri, 2015; Limtanakool, 2006), whereas another refers to polycentricity on the inter-urban (Davoudi, 2003) or regional scale (Brezzi and Veneri, 2015; Veneri and Burgalassi, 2012). Moreover, when these approaches are integrated, they are often conflated, at least to an extent (Van Meeteren et al., 2015). Second, empirical examinations of the economic, social and environmental advantages of polycentricity have not yet led to conclusive findings (see, e.g., Burger, 2011; Lee, 2006a; Meijers, 2007a).

In the policy realm, polycentric development appears to be the main hallmark of spatial plans for metropolitan areas worldwide. Indeed, more than 75 percent of recent spatial plans developed for large metropolitan areas in OECD countries consider polycentric development as the best strategy for managing urban development. Some of the key policy objectives that polycentric development is expected to fulfill include offering an economical, efficient transportation system and a sustainable environment, along with extending access to education, jobs, amenities, and decent housing to a large number of people. Policy experts’ current interest in polycentricity is rooted in the early 1990s, when after two decades of focusing on local urban development projects and land-use regulations, planning practice refocused its attention on producing strategic frameworks and visions for territorial development in cities and metropolitan regions, strongly emphasizing their relationship with sustainable development (Albrechts et al., 2003). Polycentric development therefore re-entered planning practice as a bridging concept between sustainable development (broadly interpreted as fulfilling economic, social, and environmental objectives) and territorial development. However, the understanding of polycentric development in current planning policies appears largely disconnected from the ongoing polycentricity debate in research.

This lack of connection between the understanding of polycentricity in research (evidence) and in policy (spatial plans) becomes apparent when considering the issue of how polycentric development can be conceptualized in spatial plans and how the assumed benefits of polycentricity can be realized in planning practice. This issue is of great importance to facilitating a more evidence-informed planning in which polycentricity appears as a bridge-building tool between research (evidence) and policy (spatial plans) with the aim of improving the feasibility and effectiveness of spatial plans’ economic, social, and environmental objectives.

It is necessary to conduct a further exploration of the three aforementioned major issues related to (1) the conceptualization of polycentricity, (2) the empirical analysis of the dis(advantages) of polycentricity, and (3) how to interpret the relationship between polycentricity in research and polycentricity in policy. That is the key motivation for this thesis: to link the knowledge of polycentric constellations and their economic, social, and environmental effects to planning practice and policy in metropolitan areas.

General aims and questions

The overarching research goal of this dissertation is to contribute to the debate on polycentricity in the three interrelated issues mentioned above. First, it aims to renew the conceptualization of polycentricity by bringing together two distinct literatures, namely, the literature on intra-urban polycentricity and the literature on inter-urban polycentricity. Second, it aims to empirically substantiate the relationship between polycentricity and performance in metropolitan areas. Third, it aims to understand how the makers of spatial plans have addressed polycentric development and how the assumed benefits of polycentricity can be realized in planning practice. To accomplish these goals, this thesis addresses three general research questions:

• How has the conceptualization of polycentric development in spatial plans evolved over time, and what can be learned from this evolution?
• How has polycentricity been conceptualized in research, and how can it inform planning practice?
• To what extent does polycentricity foster better performance in a metropolitan area, and how can its effects be realized in planning practice?

Single case study: the Barcelona metropolitan region

The case study of this thesis is the Barcelona metropolitan region. With approximately 5 million people, the Barcelona metropolitan region is the primary urban agglomeration of Catalonia, an autonomous region of 7.5 million inhabitants that is located in Spain. The study on the multiple links among polycentricity, performance, and planning within the Barcelona metropolitan region yields learning potential for other metropolitan regions because there exists, for example, a strong historical planning tradition in Catalonia and ideas on polycentric development have been around for many decades. This enables the study of transition patterns in the conceptualization of polycentric development in planning over time. Even the most recent plan for the Barcelona metropolitan region, the 2010 Barcelona Metropolitan Territorial Plan, is influenced by a planning vision of polycentricity that was coined by the 1966 Director Scheme of the Barcelona Metropolitan Area. The latter plan was one of the first to break with the then-popular concentric model of green belts and satellite cities and to propose networked, polycentric spatial configurations to resolve the pressure of urbanization on metropolitan regions’ central cities.

Research methods

This dissertation employs several research methods to explore how the multiple relationships among polycentricity, performance, and planning manifest themselves in the Barcelona metropolitan region. The methods used include qualitative methods such as policy/discourse analysis to answer the first general research question about how the conceptualization of polycentric development in spatial plans has evolved over time and what can be learned from this evolution. Additionally, this thesis employs quantitative methods such as descriptive statistics, correspondence analysis, simple regression models, and advanced regression models (in which both spatial autocorrelation and endogeneity issues are controlled to avoid biased estimation results) to address the second general question, which refers to how polycentricity has been identified and measured in research and how this identification and measurement of polycentricity can inform planning practice. Finally, this research uses advanced statistical methods to answer the third general question of the extent to which polycentricity fosters better performance in a metropolitan area and how the effects of polycentricity can be realized in planning practice. These methods include both multilevel multinomial logit models and multilevel structural equation models. Because of the use of these models, this dissertation can explain the estimated effects of the link between polycentricity and performance to architects, planners, and policymakers in an evidence-informed form.

Contributions to the literature

In fulfilling the threefold goal of this thesis to contribute to the debate on polycentricity with respect to the three interrelated issues mentioned above, this thesis has also made two other main contributions.

First, this dissertation has proposed a novel methodology to identify centers in metropolitan areas by considering the different pathways through which centers, and thus a polycentric configuration, may emerge, namely, the decentralization and the incorporation-fusion trajectories. This required the integration of two quite separate literatures. What also added to the novelty of this methodology was the introduction of the concept of ‘agglomeration shadows’, which has received little attention in the literature, when evaluating this identification method against its fit with the theoretical and empirical (polycentric) models adopted in the economics literature. More specifically, this thesis has also proposed a new, theory-informed conceptualization of centers as not only places with the highest level of agglomeration economies in a metropolitan area but also places that cast the most wide-ranging (spatially), powerful agglomeration shadows over their surroundings. Therefore, the center’s concept proposed in this dissertation is not exclusively static; instead, it is also placed into a dynamic perspective: a center in a metropolitan area must cast an ‘agglomeration shadow’ (growth shadow effects) over its surrounding areas, meaning that the number of firms and the amount of urban development (growth) in areas near a center will be limited because of fierce competition effects.

Second, this thesis has proposed a conceptual framework for exploring the link between polycentricity (on the intra-urban scale) and metropolitan performance aimed at enabling broad testing of the effects of polycentricity. Building upon the relationship between theories of agglomeration and polycentricity in the literature, this thesis argues that the consideration of three distinct dimensions of a polycentric spatial structure that play a role in the development of agglomeration economies in a metropolitan area–namely, (1) the size of centers, (2) the (geographic) proximity to centers, and (3) the aggregate size of centers through their integration–allows scholars to arrive at broader conclusions about the effects of polycentricity. The translation of these three dimensions of a polycentric metropolitan structure into a more comprehensive, systematic empirical framework has required an examination of the effects (1) of being located in or oriented toward centers, (2) of being located close to centers, and (3) of interaction patterns among centers.

Conclusions

Below are the main conclusions regarding the three general research questions.

How has the conceptualization of polycentric development in spatial plans evolved over time, and what can be learned from this evolution?

Envisioning polycentric development in spatial plans has become a hallmark of planning practice in Catalonia. The first vision of polycentric development appeared in the 1930s as a response to the debate about the urban-rural opposition between Barcelona (city) and Catalonia (countryside) that resulted from increasing demands to address the (negative) challenges posed by cities’ industrialization. Since then, the vision of polycentric development in spatial plans evolved, showing two transitions in its conceptualization in successive plans. The first transition was that although polycentricity was first conceptualized as a decentralization strategy aimed at restricting Barcelona’s growth, it later changed into a territorial model to organize and canalize future urban development building on the urban dynamics themselves. The second transition involved the addition of a network perception to the vision on polycentric development. This network perception on polycentricity made a definitive contribution to overcoming the antagonism between Barcelona and Catalonia because it integrated the capital city of Barcelona into a polycentric territorial model for the entire territory of Catalonia.

The applications of polycentric development in various spatial plans in Catalonia also exposed some shortcomings stemming from spatial plans’ prescriptive or normative approaches to defining polycentric development in which the empirical evidence related to existing territory was overlooked. However, the simultaneous consideration of all of the applications of polycentric development in spatial plans–and therefore, when the role played by factors other than evidence, such as interests and institutional policy traditions can be better disentangled–noted that some shortcomings in the definition of a polycentric development strategy can be explained by the fact that to a certain extent, plans are indeed politicized. This posed the challenge of building an understanding of polycentric development that was more closely connected to the ongoing academic debate on polycentricity and thus, a call for a more evidence-informed planning based on an improved knowledge of polycentricity, primarily respect for its conceptualization (identification and measurement) and effects on the economic, social, and environmental performance of metropolitan areas. Public and private actors influencing policy, for example, through their ideology or their own interests, would occupy a crucial role in the implementation of this understanding of polycentric development, based on considering (or not) the policy guidelines/recommendations that resulted from empirical evidence and aimed to improve the effectiveness and feasibility of spatial plans.

How has polycentricity been conceptualized in research, and how can it inform planning practice?

A better integration between the literatures on the conceptualization of polycentricity potentially informs spatial plans about the effectiveness and feasibility of polycentric development strategies. This integration revealed which method (empirical or non-empirical) of identifying centers most accurately defines the polycentric model in the Barcelona metropolitan region, which is an essential step in empirically substantiating the link between polycentricity and performance in a metropolitan area because differences in the identification of centers could lead to different conclusions on the understanding of the costs and benefits of a polycentric metropolitan structure. The main advantage of the novel method of identifying centers that is proposed and tested here is that it considers the various pathways through which centers may emerge, namely, the decentralization and the incorporation-fusion trajectories. This method was better able to identify as centers those cities that have the highest level of agglomeration economies and cast the most severe agglomeration shadows over their surroundings.

In addition, the incorporation of the functional and morphological dimensions of polycentricity–as traditionally coined by the inter-urban polycentricity literature–into the measurement of the degree of polycentricity on the intra-urban scale has contributed to building more sound arguments either for or against supporting a polycentric development strategy in a metropolitan area. Additionally, it has provided planners with valuable insights into not only how to address issues related to the understanding, governance implications, and expectations of polycentric development but also how to monitor the implementation of a polycentric development strategy.

To what extent does polycentricity foster better performance in a metropolitan area, and how can its effects be realized in planning practice?

A polycentric metropolitan structure exerts a considerable influence–both active and passive–on enhancing performance in a metropolitan area through individuals’ travel behavior. The effects of polycentricity–i.e., (1) of being located in or oriented toward centers, (2) of being located close to centers, and (3) of interaction patterns among centers–appear to be generally larger than the effects of individual-specific characteristics (i.e., sociodemographic characteristics and travel-related attitudes) and built environment attributes with respect to encouraging people to use more intensely sustainable mode choices (public transit and non-motorized modes) and reducing travel behavior externalities (i.e., trip distance, trip time, and transportation-related CO2 emissions). More specifically, the most important dimension of a polycentric metropolitan structure in fostering a more sustainable mobility pattern is generally the type of interaction, followed by the type of city, which in turn is more important than the distance to centers.

Based on these effects, polycentric development fosters better performance in the Barcelona metropolitan region because it has influenced individuals’ travel behavior through three different dimensions. First, people living in centers or doing their daily activities in these centers use more public transit or slow modes, and their trips are shorter, take less time, and cause less transportation-related CO2 emissions than if they do not live in centers or are not carrying out their activities in these centers. Second, people living close to centers exhibit a more sustainable pattern of travel behavior than those living further away. Third, people traveling among centers are more likely to use public transportation, to experience shorter-length or -duration trips and to make greater reductions in the environmental impact of their travel than people traveling among peripheral areas. In short, agglomeration benefits in a polycentric metropolitan region explain these three aforementioned findings. Therefore, the translation of the benefits of polycentricity into planning policies requires the simultaneous consideration of (1) the size of centers, (2) the size of and proximity to centers, and (3) the size of and interaction among centers.

Evidence-informed guidelines for planning policies

The estimated effects of polycentricity on individuals’ travel behavior have led to a set of policy recommendations on urban and transportation developments that will enhance the performance of the Barcelona metropolitan region. These policies inform the plans’ makers about how the benefits of polycentricity can be realized in planning practice and therefore, provide them with an improved understanding of polycentric development to more effectively fulfill spatial plans’ economic, social, and environmental objectives.

Essentially, the translation of the benefits of polycentricity into evidence-informed guidelines for planning policies has required the consideration of the various dimensions of a polycentric spatial structure that play a role in the development of agglomeration benefits in a metropolitan area: (1) the size of centers, (2) the proximity to centers, and (3) the aggregate size of centers through their integration. Seven policy recommendations have been elaborated to improve the effectiveness of the planning objectives of the 2010 Barcelona Metropolitan Territorial Plan in terms of individuals’ travel costs and the environmental impact of travel. 

Aggregate size of centers through their integration

• Support new, more efficient public transportation networks among centers to allow those centers to better exploit their aggregate urban size, leading to a greater development of agglomeration economies.
• Enhance the complementarity among centers on the metropolitan scale in terms of economic sectors, occupations, and urban functions through promoting compact-city/transit-oriented development.
• Support new, more efficient public transportation networks between centers and their neighboring areas to stimulate interactions toward centers and increase nearby residents’ access to the agglomeration benefits of centers that are integrated with their nearest center.
• Support new, more efficient road networks among secondary centers to mitigate congestion along the radial transportation axes oriented toward the central city of Barcelona.

Size of centers

• Promote compact-city/transit-oriented development in existing centers (central city and secondary centers) to encourage more residents of centers to access their agglomeration benefits.

Proximity to centers

• Promote compact-city/transit-oriented development in larger places near centers to allow more residents of these centers’ neighboring areas to benefit from their proximity to the agglomeration benefits of one or more centers.
• Limit growth in areas located further away from centers both to mitigate (as much as possible) the high travel costs (trip distance and time) incurred by the residents of these peripheral areas and to decrease the transportation-related CO2 emissions that they cause.

Agenda for research and policy

Despite providing new insights and conceptual and empirical frameworks to analyze the multiple relationships between polycentricity, performance, and planning in metropolitan regions, further research is needed to address a range of challenges and research gaps that this dissertation could not cover in their entirety. These challenges and research gaps refer both to the Barcelona metropolitan region case and to more general advances that are needed in the reciprocal relationships among polycentricity, performance, and planning.

The focus on individuals’ travel behavior in this dissertation’s empirical analyses needs to be extended to achieve broader conclusions about the effects of polycentricity on the performance of the Barcelona metropolitan region. Moreover, this dissertation’s empirical analyses must be extended to elaborate more comprehensive evidence-informed guidelines for planning policies that address all of the planning objectives of the 2010 Barcelona Metropolitan Territorial Plan. Two additional research perspectives can be distinguished to address these demands. First, the object of analysis could be extended from people to firms and their spatial behavior. Second, a wider range of externalities could be considered. It would be particularly interesting to conduct additional research into the link between polycentricity and other indicators of performance such as labor productivity, unemployment, housing and land prices, income per capita, household-related CO2 emissions, and land consumption.

The type of exploration performed in this thesis, a single case study, calls for further research into whether its findings can be corroborated in other metropolitan areas. Many perspectives on new research can be distinguished, but the following two are probably the most important. The first perspective would involve carrying out a multi-case study research aimed at examining the effects of polycentricity on metropolitan performance (using the indicators of performance mentioned above) by considering–and extending, if possible–the conceptual framework of this thesis mentioned above. The second perspective would involve conducting a multi-case study research aimed at testing the novel method of identifying centers proposed in this thesis against other identification methods.